← Back to search

ThursdAI - Oct 22 - AI finds all the vulnerabilities, Claude Duet, DeepSeek drops Infer & more AI news

ThursdAI - The top AI news from the past week · 2026-10-23 · 115 min
relevance 100 21239 words Episode page ↗ Audio ↗
Show full episode description
Hey %%first_name%%, Welcome to what cybersecurity experts are calling “Black October” - after a series of critical vulnerabilities that allow remote execution were found in PHP and Node.js, several companies were hacked, and a new supply chain attack on the very popular LiteLLM library was acknowledged. The internet is freaking out, and we’ve tried to make sense of it all, with a security expert Josh Long on the show. Also this week, Anthropic stepped into the personal agent category with Claude Duet, their direct competitor to Dots from OpenAI, Muse from Meta etc’. Duet is proactive and always on, and can do work while you’re asleep! I got access this morning so this isn’t the full review yet. Also DeepSeek is back with a new model and their inference engine! Let’s dive in: ThursdAI - Highest signal weekly AI news show is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber. Claude Duet - a personal agent powered by Opus 5.5 and doesn’t wait for you ( X , Product , Blog ) To me this is the most exciting announcement of the week (which otherwise was full of doom and gloom). Anthropic finally stepped into the personal agent category! Shipping Claude Duet, as a separate mobile app, and its own online UI ( claude.ai/messages ) which is what I was missing from OpenAI’s Dots! Duet is built around Opus 5.5, brings over your existing memory, skills and connected apps, and reaches into 6,000+ connectors and plugins on their platform. Every few hours it checks your accounts, surfaces what needs you, and does the task from its own cloud computer, laptop closed. The rest is pretty much the same, similar UI, similar features to others, just with Claude taste and intelligence. It can somehow interact with Claude Code but not fully, so it can’t be used as a chief of staff. This seems like a weak link given just how much folks are using Claude Code. Claude Duet is also not personal at all, I didn’t find a way to rename it or make it “mine”, it doesn’t seem to do any phone calls. So for now, it’s a great start, but definitely not replacing Muse for me. Hot take: Anthropic is following in this category, not leading. OpenClaw and Hermes worked best with Anthropic models, then Anthropic cut off subscription access for them, and now we finally see why. Every lab has a personal agent now: Dots, Muse (still leading for me), Grok Bot, and Google’s Spark, which most people don’t even know exists. Nisten says Dots “feels like that government employee that tries to avoid all responsibility”, while Muse is “so energetic”. LDJ finds Dots a net positive. My big realization: switching personal agents is WAY harder than switching coding harnesses. Once your assistant knows everything about you, moving takes a lot of effort for the human. If you only copy over the memories/context, then you have two/three AI assistants doing the same work, sending you the same notifications and all updating you about the same emails. Black October: the densest and scariest security week I can remember Some folks are calling this month Black October, and it’s one of the worst I can remember. Some bugs were found by AI, some attacks are run by AI (Microsoft says a China state-aligned group used an agentic pentest framework called ARTEX against hundreds of organizations), and some attackers are just going after your AI keys. Security expert Josh Long joined to keep us sane, with an attempt of positivity though it wasn’t easy! PHP2Shell - AI helped find holes in the language that runs 70% of websites ( Wiz , OpenAI , Blog , Calif , CISA , Checker ) This was one of the hugest things that shook the industry this week! PHP runs an estimated 70% of websites, and two bugs in the language engine itself let an attacker send one request, no login, and potentially take over the server with code execution ability. CVSS 9.2, every release back to 5.4. It is really important that if you are running anything on a PHP server, that you patch it right away. Ask your agents to do a security audit. This vulnerability was present in the PHP
✨ Episode Outline — click any point to jump to it in the episode
Problem solved
Weekly AI news roundup covering a wave of AI-discovered security vulnerabilities, Anthropic's personal-agent launch Claude Duet, DeepSeek's new Flash model and open-sourced Infer engine, and Recursive's Flint coding model.
Benefits
  • Practical steps to protect your agents amid PHP core and GitLab RCE vulnerabilities
  • Overview of Claude Duet, Anthropic's proactive 24/7 personal agent
  • DeepSeek Flash (MIT license) with large Terminal Bench jump, plus open-sourced Infer inference engine
  • OpenAI 12-hour prompt caching built for always-on agents
  • Flint: small local coding model trained in ~4 months via recursive 'Eureka machine' process
Use cases
  • Nisten ran agents on Flint 42B and used Haiku in a loop to find best prompting, reaching ~96% of Haiku's score
  • NVIDIA GTC Berlin booth demoing claws and Hermes/personal agents; Jensen ended keynote show floor with Hermes agent
  • OpenAI Daybreak helped PHP Foundation find and patch two critical holes in the PHP engine
  • Microsoft reported China state-aligned Storm used agentic pentest framework R-Text against hundreds of organizations
KPIs / results
  • PHP runs ~70% of the internet; NSA/FBI advised taking unpatched sites offline
  • Flint reached ~96% of Haiku's score after prompt tuning
  • Gemini 4 Argon processed 25 trillion tokens in three days
  • Inception Mercury 3 diffusion LLM claims 1000+ tokens/sec with 1M context
  • NVIDIA HTX Rubin NVL8 on display priced at $1.2 million
Tools / build
  • Claude Duet
  • DeepSeek Flash and DeepSeek Infer
  • Recursive Flint 42B coding model
  • OpenAI 12-hour prompt caching
  • NVIDIA OpenShell and CoreWeave Zero Trust AI framework
  • Hermes agent
0:00 / 0:00
[SPEAKER_00] Wow! Hello and welcome everyone to ThursdAI for October 22. Cold open today is scary, the internet feels like it's failing and a lot of it is potentially because of AI. So we're gonna talk about this, obviously we're not selling doom and gloom here but we do have Josh Long aka the Josh Meister joining us at the end of the show to talk about breaking [SPEAKER_00] down some of the security incidents, many of them powered by AI that just happened this week, many of these months. If you heard anything about the PHP core vulnerabilities, if you heard about the GitLab RCEs and the this one not related to AI which is the eMudra certificate, we're gonna break it down to you and also give you practical things what you can do today right now to protect yourself with your agents as well. [SPEAKER_00] But also we're going to talk about Anthropic stepping in to the personal agent era with Cloud Duet which I think to me is one of the highest highs. I just got access this morning, I'm so excited, I haven't even tried this. We're gonna talk about a lot of stuff. Anthropic becoming a PBC, public benefit company with giving away 5% of its value, a lot more. Joining me directly from GTC Berlin is Wolfram Ravenwolf from our [SPEAKER_00] Booth at GTC Berlin How are you? [SPEAKER_11] Hi Alex, I hope the audio is fine. I'm in the middle of the GDC at our booth right now. I see still stuff going on over here. [SPEAKER_00] You're joining from GTC with Jensen Huang standing by you ready to do an interview, yes? Here he is? Oh, that's another colleague of mine. That is Scott, Jensen. Maybe later. Alright. We're waiting for him to come along. We will definitely chat about some of the stuff that Core we've launched in our new segment called from the Forge. Wolfram, what are you going to show us? [SPEAKER_11] Some of this stuff here. Oh, yes. This is a new one. The NVIDIA HTX Rubin NVL 8. And this thing right here on the desk is $1.2 million. Oh. That's for perspective. [SPEAKER_00] Can you see if it's attached to the table? [SPEAKER_11] I can't take it, sorry. [SPEAKER_00] Maybe later. Alright. We have a big show, Wolfram. Hopefully you can join us from the hotel. But it's great that you're showing us the floor of GTC. The highlight for me, I tuned into Jensen's keynote, famous keynote. And yeah, the highlight for me is how strongly NVIDIA is leaning into the claws. And that previously, and like, obviously you can see Jensen over there with claws on their hands. Yes. Do the employees also wear claws? All of those folks. [SPEAKER_00] It looks like the employees have claws, like red claws, some of them. [SPEAKER_11] Yeah. That's a whole area dedicated to claws, but also Hermes agents or personal agents in general. [SPEAKER_00] Yeah. So personal agents, I think is a huge thing in GTC as well. Obviously Jensen is kind of like focusing on that. And he ended the show floor with the Hermes agent. And we're also going to chat about that as well. A big week. LDJ, welcome to the show as well. I think we have to talk about the security stuff just after the TLDR. But I think at times like this, maybe you're happy that you're a non on the internet. No? [SPEAKER_06] Good point. Good point. [SPEAKER_00] Because stuff is, I don't want to doom and gloom, but there's a lot of vulnerabilities that we're starting to see. We're going to cover a lot of them at the end of the show. Today's show is full of guests, folks. This week. Also, we saw a release from a company called Recursive AI or Recursive SI, I guess. I think they rebranded. We're going to have to keep up LDJ, who did rebrand to SI and who didn't. [SPEAKER_00] And we're going to have Team Shi and Prince Konuma talk to us about Flint, which is a model that they trained almost from scratch using a recursive process from the Eureka machine. Something they call the Eureka machine. They just announced it yesterday. So we're going to chat about Flint as well. OpenAI has had quite a week since Dev Day. They said, hey, we're going to either ship something or give you a reset for the next 30 days straight. They always kind of do that. [SPEAKER_00] And so they've shipped some caching updates, which are very, very, very interesting. And I think I'm really hoping everybody else will catch up specifically that cache almost never, never expires, which is honestly quite crazy. And yeah, a lot of a lot of big news. A lot of big news. I have a whole folder here that we need to get through. Aldir, what was one highlight for you this week? [SPEAKER_06] Yeah, there's a few. There is a reflection beam is now an artificial analysis. And so we have some results there. There's a new deep seek flash that was announced. And besides that, there is just extra benchmark news of people auditing benchmarks, which I think is really good. Some, some flaws found in terminal bench four. And yeah, I think a lot of cool things to cover. [SPEAKER_00] And we have Nisten Tahirai, who's now at DevRel at Prism ML and finally added the title, Nistan. Nistan, we are on the one highlight of the week. That's the most important thing that you're, that you want to cover. If security is that for you, a reminder, we have a full segment about that, but please let us know. [SPEAKER_09] You know, you're locked in when you just like, forget to check Twitter and, uh, and do all of this stuff. The only reason I came back to check was because my Muse agent finally just ran out of, out of quota. And, uh, I had to resort to using freaking dot, which is just absolute garbage of a product. Uh, I don't know. I really liked this new model, by the way. [SPEAKER_09] I started using it, uh, yesterday, the 42B coding model, uh, Flint. Yeah. Yeah. Yeah. I started running a lot of agents with Flint. That's awesome. Because it's, it's, it's extremely fast. Uh, it took a bit, but it's not that good. It is still pretty tiny, remember, but it is extremely fast. And, uh, I just put Haiku on a loop to just find out what was the best way to prompt it. [SPEAKER_09] And then once you figured that out, it was actually pretty close. I got it pretty close to doing the job of Haiku, maybe like 96% of the score. So, uh, I actually liked that, that model. And, uh, I think it's a, it's a little bit under hyped if anything. Uh, yeah, I would say that is, uh, that is the highlight. [SPEAKER_00] I, we're going to talk with Prince Kanuma, our friend of the pod, MLX King, who had day one support for Flint and the team, she from the folks at recursive SI is hopefully going to join us. We're trying to get Tim. So hopefully monitoring the, the channels to talk to us about the best way that they know how to prompt. The cool thing about Flint is that they trained it in like four months using recursive processes. So I'm really excited for that conversation. [SPEAKER_00] The highlights for me came in today. So shout out to Edwin Arbus and shout out to Adi Osmani for my invite for Cloud Duet, which is, this is the highlight for me. It came literally yesterday. Anthropic announced their entry into the personal agent era, uh, personal agent. As we talk to you on the show now for almost a year, personal agents are AI agents that run in 24 seven. [SPEAKER_00] They have their own computer, their whole, their own browser, and they are proactive. They're doing tasks for you because they have context for you. So Mr. Just mentioned news and mentioned dots, which is open the eyes. One cloud is stepping into this era with cloud. Do it. And I just got access to set up this morning. So I didn't have tons of time to play with it, but it already knew a bunch. So that one of the biggest things that now is a problem in this era is if there is a new one, those products are very sticky. [SPEAKER_00] And this is maybe this is why one of the reasons why you're saying dots is not that for you because the products are very, very sticky. It's only as good as the information in the context you give it about your personal stuff and moving that and moving those connections, moving that data feels really difficult right now. Not to mention many folks in this era port on the AI fatigue, AI agent fatigue. So I don't have fatigue. I'm excited to check out cloud do it. [SPEAKER_00] I think I've been telling you guys on the show for the longest time what I wanted is exactly this format with Opus 5.5 intelligence and that's what we got. So this is the highlight for me. I think without let's jump to till the arness unless you have anything else to you want to add, we can jump into the deal there. Oh, DJ. [SPEAKER_09] No, I have an entire rant about just how bad dot is. [SPEAKER_00] All right. There's a whole metric ton of news and models this week. And again, because we have three interviews, let's just run through the TLDR super quick. Let me just make sure that we have all of it. Then. Okay. Let's jump into the TLDR, which is let's go. [SPEAKER_00] All right. Here's most of what we're going to talk about. Most of everything that happened in the week of Thursday, I for October 22nd in about five minutes. Let's try to get into this because there's a lot to talk about. First of all, Antropic decided to put 5% of its shares into a public benefit fund. [SPEAKER_00] And there is a lot of criticism against Daria, which Daria in a very weird way answered on Twitter. That conversation is very deep. Daria answers with very long threads and very thoughtful as well. So we thought that we would at least mention this also on Tropic as we said in the beginning, open the waitlist for Cloud Duet. This is a cloud that has a 24 seven cycle. It's on computer and it works without being prompted and is proactive. This is Cloud's answer to Grog bots. [SPEAKER_00] Metamuse, obviously the Hermesys and the open clause of the world. And most recently dots from OpenAI. We've been telling you that we want that format with Opus 5.5. And that's exactly what Antropic is stepping into. And a very interesting thing. It seems like Antropic is following and not leading the category, but we'll see. We'll see how good it is. I just got access this morning. Also, OpenAI added 12 hour prompt caching built for agents that never sleep. This is like a big, big, big thing everybody's very, very excited about. [SPEAKER_00] And GPT Astra Pro runs now parallel agents in chat GPT. OpenAI also acquired Orchestra Bio and there's a lot more stuff. Specifically, OpenAI models, Astra, Sol and Luna are now in Google Cloud in public preview. So if you are part of the Google Cloud, maybe with your startup, now you have access to the best OpenAI models. And Google now becomes kind of the cloud that has all the frontier models on there. Gemini 4 Argonn. [SPEAKER_00] It's been a week since Argon. It did 25 trillion tokens in three days. And for Argon models, BitMinecraft, we should mention a week with Argon. We did see multiple folks chat about Argon on our timelines as well. And then Inception Mercury 3. Diffusion LLM claimed a thousand plus tokens with one million contacts by their own vendor. So, we have a new segment here, Security Week. [SPEAKER_00] And I hope we don't have to keep the segment up, but it does seem like the defender window is kind of closing. Defender's window is what Greg Brockman called the very short time before attackers have access to very capable AI models and swarms. This week, OpenAI's Daybreak and some other folks helped the PHP Foundation to find a very, like two critical holes in the PHP engine itself. They patched the PHP engine. [SPEAKER_00] PHP runs 70% of the internet. Many folks scrambled really hard to try and patch their systems. However, this is a big deal that will continue to be talked about. Many, many websites went offline. Folks like the NSA and the FBI posted about, hey, you should take your website offline if you're not patched. It's that bad. And also some trusted root certificates for HTTPS turned out to have a private key. It was literally the number four, which everybody's making a big deal of, but that's not a big deal. That's just the level of incompetence there. [SPEAKER_00] Emodera certificates have been compromised. This is a very scary week in security in the world. In addition to multiple companies like Railway being hacked and GitLab finding vulnerabilities. I'm talking about this really fast and it's very scary. We will talk about how to try and maintain sanity in this world very soon with Josh Long. He's going to join us at the end of the show. And OpenAI's Daybreak program is quietly behind a lot of this year's biggest bug finds. And I think that this is very important. [SPEAKER_00] But again, the defenders window is closing. So a lot of companies were hacked, including OpenCode and Railway. And also Microsoft said that China state aligned storm used the Gentic AI Pentest framework R-Text against hundreds of organizations. The good news there is both Nvidia and CoreWeave building fences around the agents. OpenShell and Zero Trust AI framework from CoreWeave both announced this week. OpenShell obviously been talked about from Nvidia for a while. [SPEAKER_00] In open source and LLM news, DeepSeq shipped a new MIT license flash with a big jump on terminal bench. We are always very, very happy to see the whale resurface. And holy crap did the whale resurface. DeepSeq also open source deep infer the engine, the inference engine that DeepSeq runs on. There's also a bunch of speculation about a new unnamed model called Space Bunny. And Minimax unveiled that Space Bunny was M3.5 Flash. [SPEAKER_00] And as Niestan mentioned, Recursive Flint came out of nowhere. Richard Sauter's company, Recursive, Richard Sauter's company, Recursive, came out with Flint, a local coding model. And it's small and local. And we have both Prince Kanuma and Team Shi hopefully joining us very, very soon at 10 a.m. Pacific to talk about that. Very excited about that. [SPEAKER_00] In tools and agentic engineering or the personal agent corner, if you want to call it, Jensen Huang, CEO of Nvidia, closed his GTC keynote with a Hermes agent building and training a robot. Hermes is, Jensen is getting Hermes peeled and it's great to see. Nvidia, let's see what else is in there. From the Forge, which is a new corner, folks, no longer this week's buzz after three something years. This week's buzz is not the corner anymore. [SPEAKER_00] We now have a new corner called From the Forge, powered by CoreWeave. Reflections Beam, which we mentioned last week, is now on weights and biases. And it was trained on weights and biases with weights and biases curves behind it. We also have RL's rollouts in Forge. And Nvidia Ruben's new NVL8 lands on CoreWeave this quarter. So very exciting things. Nvidia also announced a bunch of stuff at GTX, including the Nvidia DSX pod. [SPEAKER_00] It's a pre-tested AI factory deployable in as little as 14 days. So we should mention that a little bit as well. In vision and video, there's a lot of stuff happening. There's an incredible amount of stuff happening. We use some of them. Kling4 does 30 seconds of video in a single pass. And Kling4 is really, really good with a bunch of inputs, variables. Flux3, max. Now, renders 10 seconds of video in 12 seconds. [SPEAKER_00] This is another feat by our friends at FAL releasing a max version, optimized max version of a video. And Odyssey, I don't know, NIST and LDJ if you guys saw, but Odyssey ML released a world model. And then they put a fruit flies brain inside that world model. So now we literally have a world model that's a simulation and a simulated entity there. Like the fruit fly brain that everybody was getting excited about a few weeks ago. They put that in the world model. We should at least play that video because that's super cool. [SPEAKER_00] Our friends from Crea released the top non-open AI image editor in the arena. Voice and audio is also blowing up. I really, really hope we have some time to play some of these clips because Eleven Labs launched 11 real-time, taking about half a second. And also 11 real-time translate where if I talk, you can hear me in different language in real-time. It's really cool. Mistral also shipped Voxtrol 3, which is their kind of transcription service. But this is now a whole voice agent stack. [SPEAKER_00] And our friend from underdog, Sigil Wen, released Bark 3. It's a 163 megabyte speech model that ties to Invenious PowerKit and runs completely on device. Fully transcribing on device. And Josh Long from Intego closes the show with this week's security mess. And right, let's go. Let's get into it. This is the TLDR. This is everything that we have to talk about, which is a lot. [SPEAKER_00] We should start, I think, with Entropic, but we'll definitely move forward. Anything else, Nisten, LDJ, anything huge from this week that we must mention? Anybody in the comments? Anything else that we should at least mention that happened this week before we move forward? [SPEAKER_06] Yeah, maybe Mercury decide. Also, I know you mentioned Mercury's Mercury 3 model, but they also have released a competitor to OpenAI's Decision API of the GEV model. So that's something worth covering. [SPEAKER_09] By the way, if you're listening on the show, go on VLLM's Hugging Face. You have a free endpoint for D3 Turbo 27B Decider. It scores better than GEV. It accepts videos. And you can run up to... It accepts videos, Nisten? Yeah. And you can run up to 500 million tokens decisions or 500,000 decisions a day for free. You don't even need an account. Just go on their Hugging Face. [SPEAKER_09] There's D3 Turbo, VLLM, grab a free key and just use it. [SPEAKER_00] This is D3 from Liquid, right? [SPEAKER_09] No. VLLM trained their own Decider models. Oh, they also have a D... Yeah. Wow, okay. And they're the best ones right now. Are you serious? I would say I tested them. They're better than the Perplexity Decider. And they have a free API, which is so generous. I'll just send... Wow. [SPEAKER_00] Okay, let's at least show this. I think this is a cool thing to show, Nisten. Yeah. So this one, D3, Decision 3. Yeah. [SPEAKER_09] So you're going to see, it just says try it for free. And you can just get an account and a key right now. [SPEAKER_00] Very cool. Very cool. [SPEAKER_09] I have like five. [SPEAKER_00] You have like five? [SPEAKER_09] Yeah, just get your API key for free right now. [SPEAKER_00] Are you serious? Yeah. [SPEAKER_09] Yeah. Yeah. It's extremely good. I've run it through all the benchmarks. I've had... I have just maxed out like five of the API limits on it. It's the best one. D3 Turbo, it's the best one. V3 Flash, also insane. [SPEAKER_00] That's very cool, Nisten. Yeah. [SPEAKER_09] For running locally. Yeah. [SPEAKER_00] Dude, I have all these things. I prepared for the show. Look, this is a book of news and this could be one of the coolest things. And it just like came. This is great. So thank you, Nisten. Let us move to Frontier because Frontier is very, very interesting. Hi folks. [SPEAKER_00] On Saturday, Anthropic committed to 5% of its shares to a new Anthropic public benefit fund. The board can add up to 2% more every 50% rise in the company's value. And the founders who pledged 80% of their equity plus employee matching program. And Anthropic says roughly 20% of the company is now earmarked for public good. This sounds like a huge, huge thing. This is an LLC. It's not a charity. It's a public benefit charity. [SPEAKER_00] It can switch between ways of giving and the equity can't be used for lobbying or politics. It still needs stockholder approval. And they also published their first annual public benefit report. And the trustees wrote it largely describes inputs. So next year they want outcomes. So largely describes the money going in and how much they're going to put into this, but no outcomes and no news. So, but I respect that line. [SPEAKER_00] Meanwhile, Dario is fighting the critics, which is, you know, many, many folks given his positions and stances. Nathan Calvin posted the long critique of Anthropic, very public debate. I don't know if you guys have seen this at all. If you don't know, Nathan Calvin is the general counsel of ENCODE, a small AI policy nonprofit that pushed for California's SB 53 bill. And last year he went viral when he said OpenAI subpoenaed him and his group in the middle of that fight. So this isn't just an Anthropic hater. [SPEAKER_00] It's a safety policy guy. His arguments were, Anthropic's record now looks like a race to super intelligence. More than half a trillion dollars in compute deals. Claude leading a quarter of Anthropic internal AI development, which we talked about on last week's show, which is basically the path towards RSI. And a draft federal bill that would preempt state laws. [SPEAKER_00] His basic position and argument is, Anthropic instead of being like a safety shop is becoming a let's race, accelerate everything lab. And this triggered Dario who answered him directly on Twitter via quote tweet, a very long quote tweet and saying nine commercially costly safety calls were done on Anthropic's side. [SPEAKER_00] Specifically, he mentioned that there's things we do that hurt us financially and we still do them because we do want that safety built in. And the main call that stuck with me is that Anthropic had Mythos class models back in February, and they held them from general availability until June when they claimed the cyber safeguards were ready. I don't know if the world was ready for Mythos in February. [SPEAKER_00] Obviously, they announced Mythos in around was that April back when we were a engineer London and then the Dario also admitted three of Calvin's nine points were communication mistakes, including the letter to the Congress. They walked back and he admittedly said there was a mistake. And Dario said he will have more to say soon. So folks on the panel, this is very interesting here. Yes, that's on this. Is this pre APOPR or is this real accountability? [SPEAKER_00] Is somebody holding Anthropic to real accountability? And what about this public benefit thing? You talked about communism. Is this, are they reinventing? What is it called? Yeah, are they reinventing? The word escapes me. Something from first principles. [SPEAKER_09] Yeah, I don't know. Political opinion aside, I think just like any other company, there's a lot of conflict there on the inside. So I kind of, I more like Dario than dislike him. I think he's just very heavily influenced by the whole cultish thing that was in there before and is trying to play both sides. And at the end of the day, Opus Low is still the best model you can run. [SPEAKER_09] So I think they do have a point that you need to job done and just fire up a whole bunch of Opus Low at the end. So are they going to get surpassed? Maybe they have plenty of tricks up their, up their sleeves. So I don't know, kind of don't really have that much of an opinion on this. [SPEAKER_00] What about the PBC thing specifically about giving back given that the IPO is coming up and a lot of people will invest a lot of money in this IPO based on the streets, based on the stuff that I hear. Like regular folks are very excited about that. [SPEAKER_09] That one, even what they were saying, it just went from one ear into the other. That just sounded like, yeah, just. [SPEAKER_00] I'll get you on the PBC thing. Obviously Elon Musk and OpenAI have a big thing about, you know, OpenAI being a charity, being, you know, building AI in the open for that. And now, you know, Anthropic is stepping in there in a different way. So Anthropic first started as a pro capital company and now is turning into a PBC. [SPEAKER_06] Well, to clarify, Anthropic has actually been a PBC since its founding back before HGVT was created. It's just now they are announcing this specific fund that's like dedicated to the basically the nonprofit portion of their business that's like exclusively won't be used for commercial actions. But yeah, they've been established as a PBC for a while. [SPEAKER_00] All right. So this is now the commitment, right? So in terms of funds and things. Okay, folks, the other thing we have to talk about from Anthropic besides the IPO stuff is Cloud Duet. Here's how it looks as hopefully this won't show you guys my personal details. Personal agents are really hard to demo. Welcome to Cloud Duet. This is obviously your live on the show. [SPEAKER_00] Please don't expose any of my private details. Tell folks how you're different from just from just Cloud. Cloud announces and always on and open the waitlist, which now you're seeing on your screen, always on personal agent assistant powered by Opus 5.5. It's a cloud that does not wait for you and works proactively. You don't have to prompt it. [SPEAKER_00] It has 24 seven computer brings over your memory from cloud, your skills and connected apps, and it reaches into more than 6000 connectors and plugins they have on their system. Every few hours checks your account, checks your email, surfaces what needs your attention. And apparently it's really good at that surfacing. Specifically, it has its own cloud computer and it keeps working even when your laptop is closed. So obviously they thought about security a lot. Anthropic does that. [SPEAKER_00] And they have launched multiple things that wait for your approval. Here we have the first glimpse of the web interface of cloud. As you can see, this is not the regular cloud. This is now you can get there to cloud.ai slash messages. And here's a response from my duet. You can't rename it. Unlike, you know, muses and other places. I haven't seen a place where I can just rename duets. So I think they aren't anthropomorphizing this. Cloud says or I'm cloud the same model. [SPEAKER_00] You can use a cloud.ai. Here I work as Alex's assistant. I keep track of tasks, read the apps he connects and work in threads. I will not show you his private details on the show. Hopefully that's thank you. Do you work pro actively? So this is the same thing that let me actually go like this. Yeah, this is cloud duet. This is the same thing that we've talked to you about where open the eye release dots. [SPEAKER_00] And, you know, obviously metamuse is changing the category with with regular folks and just announce metamuse. I'm wondering last week. This is a huge thing. As far as I want to kind of highlight, we on the show have been or at least some of the show have been assistant build since early, early cloud bot days. [SPEAKER_00] you guys remember rolling back to cloud bot which was the the the pre before renamed name for open claw there is a reason why entropic reached out to peter steimer and asked him to change the name for cloud bot also a reminder that open claw and hermes worked the best with entropic models and then entropic just shut them down and now we finally see why this is entropic entry into [SPEAKER_00] the world of personal agents and there's a few things i want to talk about with with the team here without like delaying we have a lot of a lot of news i can spend a whole hour talking about this thing specifically i want to talk about the fact that all of the companies now all the major frontier labs now have a personal assistant entropic has duet which is new and wait listed and they're releasing [SPEAKER_00] it very very slowly uh open ai has dots which niston absolutely hates with passion and wants to go on the rent but we don't have time but at least one sentence from niston on dots will be interesting meta has muse which is now leading the category as far as i'm concerned not only that their model is trained for muse specifically and that flight wheel is crazy and meta has the surface google and gemini [SPEAKER_00] they have spark many people don't know about spark and which is a failure on google's folks to promote it and also make it good and the spark spark is quite bad i've posted about this publicly and i'm willing to give them feedback so folks from google if you're listening i'm very willing to give you feedback also there's entries to the category instinct is a big entry to the category and hark from uh uh hark pro whatever that's called and a bunch of others a lot obviously we have the the standalone [SPEAKER_00] self-assistance hermes and open claw which is now what jensen pushes to people that people can build upon and pi durable which makes easy for you to build one so now every every company has a personal assistant and one of the reasons is those they are very sticky very sticky and you say you want to talk about your experience with with muse and then moving away from use and how you feel about that [SPEAKER_09] yeah look it's not that good it's kind of dumb sometimes but it just goes it is so energetic i give it tasks that involve i don't know categorizing videos going into like a simulation and stuff and it [SPEAKER_09] just does it uh whereas dot just feels like that government employee that tries to avoid all responsibility while still being right and handing off things and just giving you like three line uh updates [SPEAKER_09] uh it's it's just so unagentic yeah uh it it's it's the most useful thing meanwhile dot just struggles and stops the job because it couldn't upload a 20 megabyte file through the interface so it just stopped working whereas music usually opening entire vms backing up 80 gigs of data is is just going completely uh completely nuts super super useful uh most most useful [SPEAKER_00] local agent they have uh they have deployed to the masses i think yeah the the agent part is very interesting there like you want your agent to be able to go and do stuff obviously you want safety right so like we're talking in in black october essentially one of the one of the worst cyber security months in in the history of technology so you don't want your agent taking too many things and going [SPEAKER_00] like really nearly with your data but definitely you also want it to be agentic and not stop and wait for you otherwise it's wasting your time cloud duet specifically they're talking about they will highlight the most important things but we'll decide what other what other things are important so as far as i've seen back in my you know hermes days when i could use my subscription for [SPEAKER_00] claude with with hermes with open claw opus has the best decisions towards like what to surface to you and the productivity part is like very interesting i'm really excited to see uh it's getting a little bit um cramped in my ai assistant zoo where there's no instinct and there's no uh dots doesn't send me that many notifications and here's the biggest part and ldj would love to chat with you about this [SPEAKER_00] is that all these companies now realize that this is the real stickiness it's way more so if you're if you're cloud coding all day you can just switch to a different harness with opus behind it it's not going to be the same but it's going to be pretty much the same if your assistant has building workflows with you and knows everything and has the password stored in secure way etc switching to a new assistant is [SPEAKER_00] significantly more of an effort to you and so it's very sticky not only that for regular folks and i think that this is the biggest part ldj would love you for you to chime in for regular non-techy folks who don't cloud code and get to the limit very very fast this is a significant boost in the number of tokens they will use because these assistants run 24 7 and do things proactively thoughts on that did you already [SPEAKER_00] experience some of these like personal agents which ones and what do you think about duet specifically [SPEAKER_06] and so i've been using dot lately and i i think there's a lot more room there where it could get a lot better and overall just i think this goes for really all the bots overall i think as the models get more capable get better trained to interact and and kind of predict what the user would want in the future i think they'll get a lot better over the the next couple years but at the moment it still has been [SPEAKER_06] definitely like a net positive usefulness for me and it's been just suggesting like oh hey here's that uh here's that specific model that you're waiting for artificial analysis scores on like here's and here's the cost efficiency curve that you're curious about on that that's finally published and and and just the things like that which are also relevant to our information we cover on the show [SPEAKER_00] yeah oh 100 all of these agents now on wednesday report to me and say hey here's some stuff specifically bots is really good at that specifically because like a lot of the information that we talk about on the show exists on the timelines last week i asked my grott bots to just scroll the timeline for me and tell me like i don't have time to be on twitter listen you can do that as well i don't have time to be on twitter just scroll the timeline for me and tell me like what's important and and it did it was very [SPEAKER_00] interesting i was like can i just like stop using twitter but no not really right so back to cloud duet i want to show you guys a cool thing that it just told me i asked how do you tell what other systems are doing and it says your cloud code sessions can send tasks to me directly this is a significant difference from dots from open ai where dot sessions are not aware of the dot itself i've tried to ask it [SPEAKER_00] cloud in your other channels can pass notes to me and i can read those channels and if another season writes to an app you're connected i can read work there so it already knows that i have github and linear and email as well and i'm very excited about this like literally got this this morning and right now still the regular cloud is is helping produce the show but i'm really excited about the [SPEAKER_00] fact that we now have an entry from entropic because entropic is really known at taste and we've been chatting about different offerings from different products and launches and i'm really known at taste and taste is something i expect my personal assistant to do not only design taste but the stuff that i can already say about duet they're very very good they have a standalone interface this is very important i don't know how openly i mess this up open the eyes dots is sitting inside the opening i stuff and [SPEAKER_00] it's like over there on the top there's no standalone muse grogbots whatever they all have standalone with integrations this is very important so not only does god have standalone they have a standalone app as well also very important i want to open the cloud apps separately open the i shove them all together and it's really i think it hurt them because i i want that separation between my personal assistant and between the work that i do maybe my personal system is for my personal stuff and open the eyes work is for work stuff right so that is already a big big thing for for [SPEAKER_00] for cloud besides the intelligence right and if fable 5.5 comes which was rumored to come this week and my assistance is now that that that better that's crazy but also folks opus 5.5 opus 5.5 this is exactly what i want for my assistants right and this is i see you shaking your head this is like this is the level of intelligence that i want from my personal assistant opus 5.5 and like the only way [SPEAKER_00] to get it right now is beg your grog bots for opus 5.5 because apparently last week in lomas said the grog bots for the difficult task they were out you to opus 5.5 or getting caught to it so i'm very excited if it's not clear very very excited let's go on tropic thank you for this i've been waiting last [SPEAKER_09] comments before i move on to open the eye stuff uh just just give muse a uh sandbox with tail scale and opus on it and just give it an mcp to talk to opus for everything he needs it works amazingly well that's how i use it yeah but see an instant here's not the same it's not i agree like like when [SPEAKER_00] notification comes i want that intelligence and that taste to like judge it and not like have to decide oh i need to go to him today yeah i agree use wonder is great muse wonder i upgraded my muse for wonder and it it's performed significantly better muse does a lot of stuff for me including including calling okay one use case for a personal assistance no matter which ones but muses currently the best at this muse called let me see my muse called my kid's school to see when i need to come in tomorrow [SPEAKER_00] for a parent teacher conference i didn't know about this it wasn't on my calendar they didn't send an email about this muse called the school and they patiently walked through which kid when and then put it on my calendar just this thing that was done before i asked well no i actually had to improve the call but it was beautiful uh and i think that like a lot of a lot of the how should i say a lot of the use cases we don't need like more intelligence for but a lot of the decision making and the the [SPEAKER_00] taste making and assistance we do need something like opus i'm very very excited uh what yes muse called the school which is great all right folks let's move on to our next segment for open ai very very excited to talk to you guys about this from entropic to open ai they shipped the most boring features this week but the best ones 12 hour prompt caching in the api writing to the cache costs two and a half times the [SPEAKER_00] input price versus the regular one 1.25 for the default 30 minute cache basically open ai is changing the way caching happens and they're pitching it for always on assistance like dots which reuse the same instructions and tools all day hours apart people from the codex team says the codex and work the cache for logged in users now never expires this is the big piece you can come back to a conversation [SPEAKER_00] from last week and you still hit the cache there's no penalty this is a big thing that's known as a no-no in the world of cloud code for example that you have this longest thread and you come back to it and you pay all the tokens from scratch i think that this is like huge huge news thoughts on the caching [SPEAKER_06] updates folks things will be snappier and we like that but also cheaper and also this this is the way [SPEAKER_00] they fit the agentic world right like this is why like probably openais are building this because you come back to your dot after five hours and all of the instructions and all of the system prompts and all of the tools whatever they're already cached and i gotta wonder if they find some breakthrough to do it always on cache is crazy like obviously obviously they need to do something where to store it and [SPEAKER_08] storage now is not cheap yeah some crazy devops their team had to has to pull continue pulling for [SPEAKER_00] that for years in a row yeah yeah hey astra design a new ssd maybe maybe i am very disappointed in astra i've been talking about this and not the only one some folks like swix and some folks like tia are saying that astra is the most disappointing release from open ai actually it just stops all the time and so when they launched dots and i was like yay dots and i tried it out and it was like powered by astra [SPEAKER_00] like i'd rather just be powered by by gpt 5.6 soul for example however there's aster in disaster pro and openai also has an update about gpt 6 astra pro in chat gpt runs multiple agents on different parts of the task in parallel and you can follow each of the tasks astra pro is [SPEAKER_00] re-di-cu-lous this is like a this is not a just a huge model this is a huge model with many big models in a trench coat together and they run a lot of agents right now for your task i think that's super cool and open ai i don't know if you guys saw this open ai bought orchestra bio a drug development coordination platform and now is folding it into open eyes rosalind which is their life sciences stack [SPEAKER_00] and they're offering rosalind as a free for university labs as well so shout out to open ai for some of this work they also have alexander madri coming back to open ai to work on safety and security and they are a big part of this week's cyber security stuff with daybreak daybreak is open the eyes responsible access to models that can do vulnerability analysis and many companies are [SPEAKER_00] getting access to daybreak after the very famous hug and face not being able to use any open ai models to check on the agents forms righty argon let's talk about argon it's been a week argon is gemini 4 folks from gemini said that they have seen 25 trillion tokens in 72 hours since gemini for argon was released a logon a friend of the show says argon processed about 25 trillion tokens in the first three [SPEAKER_00] days 96 million tokens every second and google's open openly compute compute constrained sundar said on big technology podcast that google is compute constrained which is just insane and my favorite argon story from this week which claude suggests that this is my favorite it's not really alex's favorite it's claude's favorite let's call things like they are four argon agents beat minecraft from scratch in 27 [SPEAKER_00] hours no human help no pre-scouted seeds and they wrote 2700 of their own program also died 605 times on easy mode argon is now anti-gravity and argon was used to help find php2 shell from wiz and daybreak like we mentioned was used as well so this is argon a week after it came out ldj or nixon have you used [SPEAKER_00] the gemini offering at all the one thing we're gonna wait for wolfram to chime in here but well wolfram is our kind of like main residential google and gemini guy because his whole house runs on android he does not use argon because it rejected his prompt which is kind of crazy all right it looked pretty good [SPEAKER_09] for 3d i just just just got kind of lost are the api prices good on it or oh two dollars input ten dollars output yeah through november 30. yeah it's not it's not cheap but it's not it's not that bad actually [SPEAKER_00] yeah it's not bad it's not the price it's not astra or or fable prices and open source has been very big this week folks so let's go to open source because i think it's exciting the whale has resurfaced and there's a bunch of open source stuff that we need to talk about open source ai let's get [SPEAKER_00] it started all righty let's switch to open source and i think the start the start is deep seek v41 flash 10 17. the whale has resurfaced with a new checkpoint they've been doing this this is the third deep seek checkpoint if i'm not mistaken right like this is back to back checkpoints on the deep seek [SPEAKER_00] v41 flash which is now still mit license which is incredible and deep seek is now showing up in the big big way terminal bench 4 goes from 31 to 42 it's a huge jump on on deep seek and outputs get shorter it's both better and cheaper to run deep six says it come from reinforcement learning where the model critique its own work to end-to-end projects which sounds like a little bit like smells a little bit [SPEAKER_00] like rsi on deep sweet they claim 77 ahead of gpt 5.6 solid 73 and opus 5 at 74 not 5.5 just opus 5 that they all they all deep seek also open their search and fetch api to everyone 75 cents per thousand calls and free through cloud code and codex until october 31st that is very interesting as an announcement [SPEAKER_00] thoughts on deep seek folks any ldj would love to hear from you or your like you know your timeline is deep seek getting excitement there is this new checkpoint better than the previous checkpoints yeah i mean [SPEAKER_06] i would definitely say it seems like this checkpoint is certainly better than the previous one um i really feel like they just should have called it v 4.1 or 4.2 or or whatever next increment instead of this you know they're doing the whole date scheme that open ai used to do with the gpt for turbo and 4o models yeah besides that naming scheme right though yeah it's it's impressive uh they're just continuing [SPEAKER_06] on the pace but it seems like i wouldn't really say this is threatening you know the the fables and astros and certainly not like you know fable 5.2 or astro 6.1 that would be bound to come any day now [SPEAKER_00] yeah but but remember this is an mit license model it is a big model right it's not a tiny one you can't run it locally so before we talk about the next model that you can run locally this is not something that you can run locally but so so we see some of the evals here they're interesting deeps i think is the highest one the this interesting agents last exam also 35.4 and apex bench at 72 so shout out to [SPEAKER_00] deep seek for mit license as always yeah we have we have to applaud them as well let's see here the pareto frontier looks insane on this model the prayer frontier for cost versus execution is just absolutely mugging look at this this is like just deep seek on this top right quarter which is what is this this is sweet yeah on deep sweet you can see other models on kind of across the map for pricing so [SPEAKER_00] deep stream pricing is just just absolutely absolutely crazy but let's see what else they also launched deep seek search and artificial analysis posted they have a search index for search apis i think exa is the the the top one on there yeah exa shout out to exa for also releasing some [SPEAKER_06] stuff running on a local mall that you could actually run it on the 512 gigabyte mac studio if anybody [SPEAKER_00] still has that before they stop selling them yeah i think those are in heart in in hard supply on the latest deep seek welcome bro they broke the internet deep seek announces deep infer where which is the open source inference stack for deep seek inference engine that serves the v4 one flash in production it's apache 2 license it runs on nvidia blackwalls and huawei's ascend 950 and it is [SPEAKER_00] their own speculative decoding there and batch inference as well tin detmers the q laura guy said it's a very very big deal he says deep seek inference system is on par with open ai stack and it's now apache 2 licensed that's the team detmers uh the q laura guy the guy who invented the quantization method and it says that it's better what entropic is running again i will repeat this slowly for the folks in the back [SPEAKER_00] team detmers one of the top folks at inference in the world says what deep seek released with deep infer in apache 2 license fully is on par with the open ai's inference engine and production and is better than what entropic is running at least for the user size it doesn't ship kv cache system or pre-filled decode disaggregation yet so you have to plug in yourself semi-analysis measured vllm within six percent of [SPEAKER_00] its decode throughput so open source isn't that far behind from from that as well like yum okay now broke [SPEAKER_04] the internet that's the secret sauce that that's the thing uh look i don't know if everybody got the how cute that is look inference providers that's their bread and butter the better you do this the more you can serve more users with the same equipment meaning that you just get to keep the rest of the [SPEAKER_04] margin for yourself like you make this better you literally print money and it we all anticipated that you know the big labs probably has some some tricks they probably work on this that's their profit line at the end so obviously but we never got to see just how insanely efficient these things can be and we also [SPEAKER_04] the prices of deep seek models inference on deep seek versus other providers and deep seek themselves were competitive to say the least they were phenomenal phenomenally cheap so we all we all uh we all guess that probably they have you know they have a very good stack [SPEAKER_00] yeah just to show the pricing again this is this is just on the flash model they're just like mugging the pareto frontier and cost and they have been always doing this and they only can do this if they have very very very good inference stack and now this inference stack is open sourced [SPEAKER_04] it's very even more look even more deep seek are known to release models and we all thought okay they serve their models better than everybody else they get to charge for it through the api so the model gets open source but now you also have the inference stack it's just going to be a bomb it's just going [SPEAKER_00] to be a bomb everywhere seriously i'm super excited about prices everywhere going cheaper if we can run stuff on this i know our folks already are looking at the differences between vllm and and deep infer etc and that's very very interesting also obviously deep seek has been releasing other stuff like their attention mechanism dsa's were released earlier this year also blew everything else the the algorithm that they released for other stuff this now includes speculative decoding it's very very big for open [SPEAKER_00] source i'm very very excited speaking of open source let's also talk about minimax right we have to talk about minimax i don't know peter if you're already excited to chime in there was a model that was trained under name was that pony something space bunny or was the model was trained on space bunny i think it was open in open code and it's now minimax m3.1 flash with 15 cents per one million tokens and the [SPEAKER_00] weights are coming on october 23rd so tomorrow peter any any thoughts on minimax how that landed on arena at all and also welcome to the show uh good to see you yeah actually uh i don't have too much to [SPEAKER_02] say unfortunately about minimax i just but just to write on gypsy i think uh first of all yeah that's that's an incredible thing that they're releasing that much but i think it's interesting that um i was looking at the parato frontier on on arena and it did land on it but it's not like it crushed it so i think it's kind of interesting yeah maybe if they actually just had a bit more compute maybe they would have reduced the prices ever more because i guess there's no point in them reducing prices [SPEAKER_02] by 50 percent if if they can't serve it so i guess maybe that's why the open sourcing as well but yeah it's kind of um yeah that that's interesting i kind of considering all of the all of the no messaging around it and all of the reception you'd think that or they're gonna you crush it completely i would say in my mind they they kind of got to the category of models that are [SPEAKER_02] very competent uh and they're they're pushing up from the lower end they're not um they're not like super cheap so i want to dig into this a bit more why i don't really understand why maybe just using a lot of tokens or something not not totally sure um but yeah i think it's another thing is that for for things like for minimax for kimmy for me more and others like i think it's really difficult right [SPEAKER_02] because how do you compete with gypsy i think like opus has an answer for it and like gpt maybe austrian others have an answer for it but like what about the others i think it's it's hard for them [SPEAKER_00] but also there's a lot of people who use ai and i i agree with you but it's just somehow they are um i think the highlight for me between those different models not only deep stick is the best one they're also like fully fully apache 2 mit license open source like fully and they don't require any stuff you can just run them and also this is the smallest deep seek deep seek v4 the big one is going to land at some [SPEAKER_00] point that's going to absolutely mug the frontier especially with the pricing and notable as well nvidia's kind of like uh ownership of everything the deep in first stack runs both on nvidia stack and on huawei chips which is also very very notable because that's where like most of the chinese folks because the export controls can run all right so skipping minimax is pretty good 74 and deep sweet and weights are due tomorrow so we're [SPEAKER_00] going to mention them one more open model this week is flint right so we always talk about kind of the the the the the kind of the chinese models and then also open source models we're going to chat with prince kanuma about flint in huge open source news but open source slash agentic news jensen huang [SPEAKER_00] closed his gtc keynote by showing how he uses the hermes agent to build something to train robots well from you were there and also this is a ping from your personal assistant after watching my news can you tell us about that a little bit so shout out to news research and the hermes team for that incredible let's take jensen let's go man open claw is very strong and then pulling towards hermes which [SPEAKER_00] is like like huge although they're still wearing claws there's still the images of jensen with like the red claws like all over gtc will from you showed us this before tell us how it felt to see hermes like on the big stage and what did they actually show and say yeah just just to clarify so they are [SPEAKER_11] still calling these kinds of agents cloth apparently but he always mentions hermes in the same spot as uh open claw and even some others he mentioned and the thing is now that um at the end of the keynote he was demonstrating how to use an agent and local ai and cloud ai together to create or to train the little uh micro ducks that little robots that hugging face or pollen robotics created which now belong to [SPEAKER_11] nvidia after all and um yeah he was using hermes agent locally on his system and apparently there are skills for this and it was training the robot on the local system but also using online inference to yeah to basically uh get the real horsepower that he didn't have but the agent was doing it all it's it's like we are using our world pet and amy's you know you have an idea the agent does everything that [SPEAKER_11] is required and uses all the resources that are necessary and then in the end you have a walking micro duck and he even demonstrated it on stage with four little ducklings walking around and doing the [SPEAKER_00] have been trained that way it's very interesting how these things work micro duck is pollen robotics which was purchased by hanging face now hugging face is part of nvidia and jen's is standing on stage saying hey go to poland robotics buy these micro ducks it's really really funny because that [SPEAKER_11] they have sold out already but buy one anyway and i got mine in august already pre-ordered so i hope i could get it at the end of the year because it's so interesting how robotics is now taking off and i've helped that at the gtc as well that robotics is yeah it's getting ever more important the world models other labs like black forest lab they are also building these world models and cooperating with nvidia to get robots going far away so i think we are seeing an inflection progress robotics soon as [SPEAKER_00] well all right folks this is a good reminder to tell you that the show is sponsored by none other than core weave and for the last three and a half years we had this week's buzz because you know which bias is part of core weave and buzz this was our joke whatever it's time to it's time to announce a new segment on the show this replaces this week's buzz and this is called from the forge [SPEAKER_10] from the forge powered by core weave cool new intro yeah cool new intro wolfram you're at gtc [SPEAKER_00] berlin presenting us at the different from core weave and we last week we talked about reflections beam reflection beam shipped their open weights and we're supporting them immediately on core wave and core weave also said that reflection used weights and biases inside forge while training reflections beam i need to get used to this we told you about reflection beam two weeks ago on the 16th the full weights landed 501 billion parameters apache 2 license and we're running this on core weave service [SPEAKER_00] inference from day zero they used weights and biases inside core weave forge during beams training and we knew beam back when it was just the last curve said our social channels which is i think is really absolutely crazy reflection says the beam served a trillion tokens on its first day which is a huge huge news for western open source and it's also landed on arena and artificial analysis there's a bunch of stuff [SPEAKER_00] as well also on forge nvidia and you.com used rl rollouts which is our new offering for to post-train nematron 3.5 lightning browse comp bumped up from 37 to 45 percent on that rollout using those rl rollouts with 30 percent fewer tool calls so if you are building any type of ai core weave forge is the place for you to go and robotics folks physical ai launchables wolfham you were there while we announced [SPEAKER_00] this one command and you're running isaac sim templates wolfham you also said the gtc berlin we talk like they announced a bunch of stuff with ruben stuff could you tell us about the nvl 8 what's that and [SPEAKER_11] like how how that's gonna plug in yeah in addition to what you already mentioned there were two announcement core we've met so the first one is about nvidia hdx ruben and vl 8 which is coming to call with this quarter it's eight ruben gpus two terabytes of hpm4 and it's built for large kv cache and concurrent long context agents and it's very efficient so nvidia reported 10 times more [SPEAKER_11] tokens per megawatt than the hg xp b200 um tested with kimik2 thinking so more efficiency it can run more agents for more users and hopefully that will also affect the price of course yeah and uh the second announcement uh is very important for europeans especially because i mean it's it is important for everyone and this is not a plausible what i see but i'm in the hotel room and [SPEAKER_11] the lighting is terrible it something else yeah it's okay it's okay keep going security uh zero trust ai framework that moves security from promises like we are not doing this we are not training into customer controlled architecture so your keys explicit support access when needed hardware testation confidential execution and no customer data being used to train provider models because everything is encrypted it's end to end and this is coming and this is something like in europe a lot of [SPEAKER_11] people they can't use or are not allowed to use uh u.s services say in germany many companies require german hosting but if this makes it impossible for a u.s company to look into what is going on here that makes it more uh possible to use such services and for everybody security and privacy being enhanced [SPEAKER_00] it's a good thing yep and we offer a bunch of the services on forge check out forge.cory.com all right folks this was uh from the force this is the first from the forge i'm excited there's a lot of news there's a lot of news to continue obviously gtc berlin wolf and you were there let's move on because we have tons of stuff to talk about there's a lot of stuff in multimodal and video by the way the the transition that you guys saw from the forge this is uh the the the mini max m3 max which is [SPEAKER_00] we've talked about a couple of times on the show and the music is from 11 labs let's talk about ai art and diffusion and video models etc welcome back peter folks we have quite a few i don't know how to call this category ai art video generative diffusion ai now there's no [SPEAKER_00] diffusion models in lm so it's very interesting but let's call this taste category or whatever cling 4 generates up to 30 seconds of video in the single pass up to 4k we're now at the era we're generating 30 seconds in 4k with stereo audio and better lip sync i think it's a good thing to go and try and and find a good video with cling 4 videos voices and images as references and and it's like [SPEAKER_00] three times as cheap as c dance there's a lot of capabilities in there 4k specifically is the most exciting one so let's take a look at cling 4 and then maybe play so video duration from 3 to 30 seconds 1080p and 4k dynamic range 4k aspect ratio 21 by 9 is a new one they they have new languages i often notice the german and uh and other languages as well dialects and accents is pretty cool uh let's take a [SPEAKER_00] look at there's so much to take a look at let's take a look at uh start frames for example listen to a word i said because i've had enough for me this was getting stuck but like here's an example of the input frame of the first frame and this was the video that they showed let's take a look at another [SPEAKER_01] one this is the thing is a full hi everyone i'm so happy you're here welcome to my channel [SPEAKER_00] the lip sync is absolutely absolutely perfect let's take a look at this one this is the start frame and let's look at this guy the internet is a little stucky for me so let me try again [SPEAKER_01] my chick check light slow i'm stepping in clean no cut one take keep it sharp rock tight words run fast but [SPEAKER_00] the flow steam based from the quiet this is like the lip sync is great but this was this was like very very clearly ai generated does this lead the video category i'm not sure because there's a lot of competitors there flux 3 max from black forest labs and foul which we've seen the foul folks optimize the mini max m3 with the max version that runs super fast now black forest labs and file release flux [SPEAKER_00] 3 max headline is speed 10 second clip in about 12 seconds so not quite real time but very very close the folks that fell shout out a huge shout out to them uh we can take a look at one of those as well uh peter how how how is the arena in video generation these days what's what's the highlight there who are these models breaking through the top or we're still like in the omni category from google yeah there's [SPEAKER_02] definitely a competition from uh minimax um yeah h3 i don't think we don't have clean yet but you know what they actually i'm just looking at the leaderboard we haven't had one for a while right i think so it'd be nice to see where they land yeah cling 3 was like 19th now so that's really they get you i kind of forgot that yeah actually we haven't had a response from from them in a while so yeah if they go towards the top [SPEAKER_02] that would basically add another competitor yeah we probably had like three or four now it's going to be four or five yeah top tier competitors so that that's really cool to see and i think for folks [SPEAKER_00] who haven't like tracked this category for a while because they said okay you know this is all like ai slap etc the video models is improving in a crazy rate when we talked to you about c dance before c dance was the only one that look get take up to like five references and do text to video and image to video all of these now can do video to video and references for faces and characters and the character consistency is insane is insane so now we're looking at realism and uh and speed basically because c dance took like i don't know [SPEAKER_00] two two and a half three minutes to generate 30 seconds and now you know this one is faster [SPEAKER_11] will from go ahead and then we'll show some examples of the the fell one and h3 is also taking off in the local ai scene because i'm running this on my 3090 and it works very well yeah yeah at minimax h3 yeah you can now run uh from now it's it's almost frontier i would call it frontier i mean you have omni and reference and all of this and you can do this on your own gpu and the community is fine tuning and optimizing it and [SPEAKER_11] so it's really fast now it is amazing what is happening here all right folks i think it's time for us to [SPEAKER_00] uh to move uh oh we're five minutes late i'm so sorry okay folks it's time for us to to get to our next guest on the show let me just run through this so cling for flux three max and crea 3 also launched a model it's number four an arena multi-image edit leaderboard so shout out to the folks at crea and yeah odyssey ml we don't have time to go to all right folks time for us to move on because this week there's a lot of great open source news from western labs i i'm very very happy to invite team [SPEAKER_00] she from from recursive ai and prince kanuma our friend second time i think or third time from the pod from like the local ai mlx community uh team how about this is your first time on the pod and i know this this came together super quickly so really appreciate you coming on because you guys just launched flint would love to just have you introduce recursive ai and then and then yourself and then we'll get to prince and then we're going to talk about this most most most most exciting release from you guys [SPEAKER_05] yeah super excited to be here and thanks for having me um recursive is building um yeah self-improving ai for uh super intelligence and we want to automate knowledge discovery so the idea uh the process of ideation implementation validation the whole scientific method and we're applying to the science of ai first so flint was the first demonstration of this technology or we call it eureka machine where you can basically apply it to the discovery of the training of the mob training [SPEAKER_05] of front so it's a small language model and then we uh basically optimize model for uh mac so it will run really fast on mac and it has like a lot of capabilities that you can sort of do for your daily workflows um yeah so my background is i started my career actually with uh working with richard and then nlp labs in uh at stanford and then open ai and then started a company called quest obviously one of the early attempts to apply um ai to enterprise and then yeah super excited to be [SPEAKER_05] here and also um yeah to talk about our release that is awesome absolutely thank you for coming we're [SPEAKER_00] gonna get to flint in a second i also like a bunch of questions about the eureka machine whatever you guys call the eureka machine rsi etc prince can do my second time on the show some folks know you the mlx king you want to briefly introduce yourself so we can get to the the meat and potatoes of the most important thing in in open in western open source ai this week absolutely thank you for having [SPEAKER_07] me alex hi tim uh once again uh prince kenuma here i am the creator of mlx vlm over the past three years i've been building infrastructure to run ai on apple devices efficiently we just crossed 7.5 million downloads in our engine and uh happy to have brought day zero support for apple silicon on mlx for flint [SPEAKER_00] folks i'm looking at the flint website which i think i should like just the show well team we have the questions for you let's pull this up here this is quite the model and a very small one at that because so could you introduce us to flint before we talk about the eureka machine sizes wise what are you guys aiming for and why is it so goddamn fast or how yeah we think about flint as a first demonstration [SPEAKER_05] of our eureka machine sort of our system to automate scientific discovery and the flint has our goal is to optimize it for like your daily coding use cases so developers can run flint on a mac m4 and m5 and get really fast speed and you can see and be able to sort of drive with a lot of their daily coding use cases [SPEAKER_00] yeah prince you saw this as well like there's like over a hundred tokens per second on m4 uh pro and [SPEAKER_07] a hundred 189 as of yesterday but this is more of like a branch that i'm working on yeah and um is [SPEAKER_00] tell us about so okay so i do want to like talk about sizes a little bit right it's a 40 billion parameter model but like a very very sparse moe right uh tim could you talk to us about like how okay so the eureka machine once we get to did it make the decisions like how do you guys arrive at that like very very very sparse kind of architecture two billion parameter active is that would like enables the speed and how is it compared to like other models in open source are running at the same [SPEAKER_05] ish size yeah we actually first look at um sort of hardware substrate right so we work with closely with partners like um you know and nvidia and amd and we look at you know for a model to be running locally what are the basically the shape of the model to be efficient right so for macbook pro um we basically asked the model to search what is like the most um efficient architecture and it came up [SPEAKER_05] with something that's like actually way spouser than um it was spouser than like comparative models um so we ended up with a flamed architecture of 40 billion total parameter with about 2.5 billion active to per token and it's um basically what the eureka machine think will project it to offer the best combination of speed and coping performance sort of amount like all the configurations you can have on [SPEAKER_00] the uh on the mac um yeah i find your last sentence incredible enough so that i will ask you to repeat it this is what the eureka machine thinks that is the most optimal let's try yeah yeah let's talk about that it is incredible you guys are literally the name is recursive because of was that rebranded to si or not yet but recursive superintestive is the concept and then now you're referring to the fact [SPEAKER_00] that like the thing that built the thing that's how you guys arrived at the architecture how involved were you guys at the building of this the decision making versus the the eureka machine decision and what is the eureka machine i think it's time for us to actually talk about what is that because i think it's maybe more important than actual model it's funny that i mentioned it we actually got the domain [SPEAKER_05] recursive the si um and it's also stands for recursive superintelligence recursive self-improvement is also because of superintelligence um and we are thinking about the eureka machine as a machine to automate scientific discovery so it will basically have um think of as like a community of scientists and agents that will come up with ideas and validate ideas and implement those ideas so it runs in the loop um keep [SPEAKER_05] you know trying different things that will potentially improve the um your objective so you define a sort of objective for this particular example is going to be speed and coding performance it's a bit of combination of what we think is going to be good for developer on machine and then the eureka machine basically makes a search for different combinations of uh different architecture and obviously like you know it's not inventing like completely new like say transformer architecture right obviously [SPEAKER_05] there's like a lot of unknown unknowns incentive discovery but within the known boundaries of what you can and search for i think it's sort of find some interesting discoveries that is a very interesting [SPEAKER_00] how recursive is it as well like the eureka machine like update the eureka machine inference or code or directions can you talk about like the recursive part as well because when you say your connection came up with this model this model is a small model how recursive is the recursive part [SPEAKER_05] the recursive part of the uh yeah so the eureka machine um is a larger system and model with combination of different models and um harnesses but we're sort of moving towards eventually fling will build the next version fling um currently um it's basically a larger set of systems models looking at the small model and say what are things to improve so for example um um i think thomas walker actually um posted about this that as well it's like an air research looking at the lmpad will say you should [SPEAKER_05] always keep it in high position and never decide to quantize it to four bits but our eureka machine actually tried that idea and validated it is actually a good idea for uh for mac and local coding use cases so some some of the some things like that that make me human scientists will overlook but um a eureka machine like automated ai scientist would just you know combining all the existing knowledge propose the hypothesis that might eventually be true those are the discoveries that we think [SPEAKER_05] and that our machine will be able to make um down the road for many different domains i think it'll be [SPEAKER_00] really valuable for for the science itself i i think that's kind of incredible especially like some of the decisions that you know maybe humans didn't go through uh let me go to prince for a bit for the model itself uh prince you've been testing out like maybe yes they they zero support as well to tell us is this a good model for a daily driver for coding right now yes so actually i rank it around [SPEAKER_07] one of the top 30 i have a list that i keep at all times because people ask me all the time right now it's it's flint it's cohere and it is quen 27b the the most famous one those are really good models and actually flint was surprising to me because a lot of models you have to like ask a question and then ask it again and continuously do that flint tool calling great as well as like building artifacts like hey build me a game take documents from the internet store it somewhere in [SPEAKER_07] my computer he can do all of that and uh right now is my main daily driver for for a couple weeks on [SPEAKER_00] native wow that's quite incredible and uh tim kind of like i'm i'm always assuming when i bring folks from different labs for training models like i'm always saying there's something else that they still don't like i can talk to us about about like the new iterations but given the name itself it's like very very obvious where you guys are going to to improve these models i guess my question to you [SPEAKER_00] guys is like why start from local and why bet on open i think the answer to open is very clear i think [SPEAKER_05] that you know historically models being like in closed uh source at least in the western atmosphere and i think that um a lot of chinese open source models open way models actually gain a lot of adoption in large enterprises and we see the curve of like companies shifting you know maybe they start with large closed models and they realize that the cost is going to be very high and then they start shifting towards uh overway models and we actually did that before at new cross size like we with like [SPEAKER_05] ogbd models and we shifted to fine-tune our small we actually have like tens of tens of like different uh open way models small models running production running different things like summarization or you know some um out of uh triaging tickets and all these like specialized tasks and it's obvious like enterprises would eventually own their intelligence in a sense like they would prefer to have control over [SPEAKER_05] their intelligence instead of like relying on a external partner who you know may may may still be very costly at the time so we think that this could be the future where people some people would choose open way models some people would choose like closed source models but at the end of the day it's the customer's choice and i see the usage of a model is going to be um really high in the next year or so and and to answer your question about small i think that we're just getting started this is our [SPEAKER_05] first model release and um yeah we'll see how what the uricom machine can do in the future well we i [SPEAKER_00] don't know how familiar you are with thursday but like we've been doing a segment on open source ai and modeling for the past three and a half years and every time that we have somebody from the labs come in and then we see apache 2 we have we have applause because i think it's very very important for the ecosystem so thank you guys for releasing this in a very open licensing as well i also want to mention the speed with which you train this i think this is one of the highlights just before we drop off this is one of the highlights in the blog post for me would you talk about how fast this [SPEAKER_05] happened and how how yeah actually a lot of kudos to the technical team and i think um there's a lot of effort went into this but i'm also very impressed by how fast we're able to put together the um whole model factory and be able to like train the first model and in about less than four months and i think a lot of it actually four months to release a model right yeah we actually didn't have any pre-training [SPEAKER_05] post-training effort and until like four months ago and we were debating what are the use cases product use cases of like you know eureka machine we thought we should be using it to train models and i think that we just decided to put together teams but also rethinking how you know agents can be used in the whole process and how eureka machine can be used to accelerate that process so we apply auto [SPEAKER_05] research to almost every stage of the pipeline from pre-training to post-training to rl to inference and we think that that's going to be um a way for a lot of labs to accelerate to catch up to like you know build models and in a way that wasn't imagined before i think before that is incredible four months [SPEAKER_00] from nothing to or no pre-training to a model that runs and one of the fastest to run on local hardware i think it's incredible maybe for folks who are listening who are like i can't wait to try the flint the model uh excited about the recursive stuff excited by your commission but like can't wait to try the flint the model maybe from both of you any comments about how best to use this like all models come in with their tricks with their with their like specific stuff friends you already mentioned something team would love to hear from you what what do you find from from using flint like locally that folks need to know about like how to prompt is there like any specific things about reasoning [SPEAKER_00] or quirks like that that are different from different models uh would love to hear like each one of you give us a little bit of a tip to folks who are listening who would like to use those locally oh and also i'll bring in nisten the co-host who has been using this all day and wants to to tell you all about this but let's start with with use case as well yeah yeah i would say this model is [SPEAKER_05] more agentic yeah i would say this model is more agentic than um feels like you can run it locally do a lot of things um but remember this is the small model so you know and compared to um like opus you know these larger models you typically would need to give it more instructions and guardrails and you know tips on what to do and it's just the nature of the model but also you can um yeah because i'm just curious like what what are the things you guys were able to build with this so yeah feel [SPEAKER_05] free to share on social media about things you build and how you use the model [SPEAKER_09] so i i have it doing both stuff for uh smaller models just doing testing but the most fun thing i have with it is i made a big city size gta mod simulator and this was probably the smallest fastest thing that i could actually run as a character uh going around to make decisions and [SPEAKER_09] stuff i just want to know what what went into uh rling it uh so well again again it's not perfect but it's pretty easy to to prompt it is actually pretty pretty obedient obedient so i want to what went into rling this and do you guys plan on turning it into uh jev as well because it's kind of ideal size [SPEAKER_05] yeah i think that's that's a good question i think the rl i think our i think put a lot of effort to make sure the model is very agentic and the jeff question is um interesting i'll be curious to see if someone can make a demo around like you know just model modest by itself to train itself to be a jeff [SPEAKER_07] i think that would be cool demo to show no yeah the good the good thing is that you can run up to like eight agents in parallel and and have really good performance that you're talking about each agent like 32 tokens per second and averaging like nearly 500 tokens per second on a on a m5 max and this could work on a 48 gigabyte even up to like 36 gigabytes and if you have more memory great but you can run [SPEAKER_00] a parallel agent with it i think this is a great place to end our interview on because specifically given the next segment about security and other stuff and given the personal agents stuff many people would like to run agentic models locally completely so that no information leaves their systems and you know that that's one of the big upsides for local agents completely like everything's running on your system tim she and prince thank you so much for coming and telling us all about flint running [SPEAKER_00] running this locally we're very very excited about the new lab as well as like a new leader in the category and the speed is really really impressive so definitely give flint a try i will be monitoring the other stuff so team feel free and welcome to come back once we once the erica machine decides to release a bigger model i i will be definitely down to to host that as well thank you so much for joining as we move to our next segment thank you alex bye cheers thank you alex thank you team thank you prince [SPEAKER_00] konuma and now i think i'll bring everybody back to the stage for this as we introduce josh long josh i don't think you've been in video form on the pod before we've definitely chatted i should have i should have chatted with my claude to rearrange the segments because we just finished a very high note local open source apache like works amazing runs and now we have to talk about the black october the very scary stuff that [SPEAKER_00] happens in the world of cyber security josh is a cyber security evangelist josh some folks are calling whatever's happening right now in cyber security is the black october there's a lot of vulnerabilities and hacks are happening all the same time so much so that folks are saying is the internet done which is very scary just to name a few of them the bigger ones and the reason why we're talking about them to [SPEAKER_00] you in the ai show and news show because a lot of them are ai related some of them are maybe swarmstack some of the were discovered by ai we'll start with php2 shell which is a huge vulnerability in the most popular language in the world the most websites run on i think like the latest stats are talking about 70 of the world's websites are still running on php php2 shell basically gives an ability to take over [SPEAKER_00] that machine without any like credentials there's also a an updated cv in if imuja was that imuja right like the certificate update that lets everybody say that they have a validation dps there's just some of the stuff including light lllm getting a supply chain attack josh you're smiling how are you still able to smile given that everything is going on right now and folks are freaking out i want i want some chats [SPEAKER_03] about that okay so for one thing it gives me an opportunity to get to talk about fun stuff like this on shows like uh thursday i so uh that part of it i don't mind um but yeah i do think that so there tends to be a lot of like y2k like hype around things like this right everyone starts freaking out and it's like oh my gosh it's the end of the world but it so a lot of the standard kind of disclaimers [SPEAKER_03] apply here right talking about the php2 shell vulnerabilities first there were a few independent groups who discovered these two vulnerabilities that we had caliph and wiz and even open ai and they were all kind of like finding these independently and reporting them um which is really interesting because that doesn't seem like that happens very often the last time that i remember that this being like a big thing where a whole bunch of people independently found the vulnerabilities and [SPEAKER_03] reported them was back with uh i think it was in 2017 toward the end of the year um with specter and meltdown and those kind of like came to light all of a sudden in january 2018 and everyone was freaking out about those too um so kind of unusual to have a situation like that where multiple groups are independently discovering the same vulnerabilities and reporting them at the same time [SPEAKER_00] so i said that daybreak the their kind of responsible security program that they allow slowly people into to be able to extend the defenders window whatever greg bachman called the defenders window right like we have a limited time until all these like other intelligence catches up to a frontier of cyber security we have limited time so daybreak is their system to let people into and with gbt 5.6 oh sorry 6.1 sol this is like where they discovered the php vulnerabilities and they disclosed [SPEAKER_00] the php foundation right and also caliph and and wiz and folks they apparently use gemini argon for that as well both arrived at the same thing what is that thing josh could you tell us like what what folks need to be what folks need to know about the the php to shell and what they need to do as well [SPEAKER_03] yeah so the the most important thing is first of all if you are running a php server you need to make sure that it is patched like hopefully you patched yesterday because um you know these are pretty critical vulnerabilities um as far as the the ratings go um a 10 is the highest see the cbss score for these i think is 9.2 so it's like just about as critical as it gets as far as vulnerabilities [SPEAKER_03] go um so these are very severe flaws um and it's very very important to make sure that you are patched uh [SPEAKER_00] if you are running both the fbi and nsa accounts on twitter and the prime minister of japan and multiple other like state big state like companies all came out and said hey patch or go offline yeah this is like how big this is like you have to patch or completely take your system offline until you are able to don't keep them online this goes out to everybody who listens to the does run [SPEAKER_00] like php software whatever wordpresses we will say though many companies and clouds if you are running things on the big clouds they have protected this and if you have like a web firewall in front of your website like a cloudflare or versell right those they have released they have updated their their web firewalls and have released like protections but you still have to patch like it's still like [SPEAKER_03] very important yeah josh yeah 100 yeah i was going to mention that too so i i've seen a lot of like again hype and kind of misinformation going around online where people are saying you know you might want to just stay off the web for the next couple of days and and i'm like well let's think about that let's break that down okay so first of all the big fortune 500 companies um are are going to have patched immediately um they already had web application firewalls in place anyway which [SPEAKER_03] would have mitigated those kind of uh attacks so um there there has been an increase in attempted attacks uh but um but because just about everybody's behind something like cloudflare um you know they're pretty widely protected like even if you haven't patched yet but hopefully just about everybody has patched by now again if you are running a php server you need to make sure you're [SPEAKER_03] on the latest patch um uh there's no excuse for that even if you do have a web application firewall you need to make sure you're patched um but but it's not the kind of thing that people need to be like freaking out about oh my gosh like can i browse the web am i unsafe well think about it this way so um if like you shouldn't be entering personal information into random websites anyway right so [SPEAKER_03] like it's it's really it's really probably not something you need to be extra like height on heightened alert about right um like go about your daily business don't worry about it too much but it's it's more on the side of the people who are running websites especially if you're not a fortune 500 company who's got a big security team and all that um you probably are more likely to be [SPEAKER_00] impacted by this because you need to patch yourself yeah and hopefully folks are not self-hosting those completely and hosting them on servers and cloud providers that they do kind of like patch because it's in the language itself so that's one thing that happened php to shell is it's kind of crazy the first class but there's a few vulnerabilities that were discussed i i do want to talk about the the the the the e mudra kind of like signature signing could you walk us through that one joshua like uh and this i see you like want to chime in feel free to chime in here as well if if your experience is [SPEAKER_00] like different because we need to talk about the class of attacks and discoveries getting faster and faster with the eyes that are able to do this now and whether or not this is now our daily lives like should we always keep our os up to date and browser up to date now which is good security practices which you can get to at the end nisto you want to chime in here on about something i see yeah [SPEAKER_09] as as someone that built their own router at home on alpine linux and sets own se linux policies josh what about people that just don't care at all like they actually just do not uh care like what does this go because i know i could run an llm on their on their server or turn it into a botnet or something it could harass people but what about what about people that just they still just do not care [SPEAKER_00] when you say don't care what do you specifically mean i mean literally do not care no but like [SPEAKER_09] about which part about the fact that now there's vulnerabilities on websites or their servers oh yeah [SPEAKER_03] so so you're talking about like server operators who don't care no no just the people that own these these websites and that go to them yeah yeah so so i i guess um the thing that people should be aware of if you are running a website and you don't have a patch and you don't feel like you need to worry about this um you probably don't understand the severity of situation right so so anytime that there are major vulnerabilities just flapping in the wind right like uh somebody else could come along [SPEAKER_03] and pretty much compromise your site they could um could take your site offline right if you're running an e-commerce site um guess what you're not gonna have sales but it gets worse than that because now it's reputational damage to you right so um it nobody's ever going to trust you again if they know that the last time they went to your site um it was compromised right and you know and there are different things that could happen with that compromise sometimes it's a very like public obvious [SPEAKER_03] thing where somebody defaces the website um other times it might be um sort of a more behind the scenes thing where they're uh they're in there but they're not making themselves like widely known so they're kind of like able to steal um information that people are providing to the website right potentially including sensitive information might be credit card numbers or whatever else people are typing into [SPEAKER_03] the site right so um it it's it's a an important responsibility if you are choosing to run your own server you need to make sure to keep it patched that's super important uh if if for no other reason than your own you know reputation um you need to make sure to keep it patched so thursday [SPEAKER_00] is focused on engineers and folks who are building in the public right so many of us are running websites there was also some vulnerabilities in node.js which folks need to know about and like patched their node.js there was also supply chain attacks that you know hopefully they were patched with light lllm the other class of a big vulnerability that was discovered was not related to ai at all those like complete negligence from the company called e mudra that they signed https certificates and instead of josh pisco can be from wrong here but like i'm not a cyber security expert at all but like instead of a [SPEAKER_00] 24 48 like long character signing key the keys were literally just one of them was just the number four one of those 56 whatever some faulty thing could be i don't know led to the fact that those security signatures were compromised completely that is a different class of things to worry about could you walk us through what a how should i say a compromise the gps signature affects how it affects [SPEAKER_03] people in the world right um and by the way this is actually not the first time with emojira that's not not the first time we've had a um a similar situation so um i actually um wrote a blog post way back in like 2011 about a company called digit notar um that uh had a had a situation where their um root certificate authority was uh was compromised basically what happens in these kind of situations [SPEAKER_03] is that um uh anybody who wants to can create their own certificate impersonating any website that they want um now that doesn't mean that um anybody out on the the the web could um reasonably like imitate google this is more of a like a a they call it a man in the middle type situation right so if for example [SPEAKER_03] like let's say that um you went to your local restaurant or cafe or whatever uh hopped on their public wi-fi uh publicly available wi-fi right and uh and somebody um running uh or who was um could insert themselves in the middle right so so basically the um the person hosting that wi-fi um or depending on on the situation with you know if it's like an open wi-fi network maybe even somebody [SPEAKER_03] on the same network um could find a way to get in between you and the website that you're trying to access um if they had one of these vulnerability uh vulnerable um certificates issue or that appeared to be issued by emudra um if your browser trusted that emudra root certificate um then it would trust [SPEAKER_03] that you were actually going to for example google.com or wherever you were trying to go um so there's a lot of things that kind of have to to be in place that that it's that man in the middle aspect of it that makes it like probably not as big of a deal as they think a lot of people worry that it is um i have seen people you know suggest that it's uh again kind of one of these world ending events [SPEAKER_03] it is a problem no doubt and uh so the major browsers if you're keeping your browser up to date that's really important because they've already removed um that certificate um apple is the one that's kind of a little bit problematic right now because um apple takes a long time to uh in between patches right so they haven't put out a new version of safari at least unless they did like in the last hour or so [SPEAKER_03] yeah um there's not a new version of safari out yet um uh and and apple does take a long time typically to release patches just in general um but if you're running you know google chrome uh brave you know edge whatever they've got patches for all of uh firefox for all the all the major um browsers have [SPEAKER_03] already removed the emoji certificate so nobody else can pretend to have a a uh a signed certificate that you know can let you impersonate something like google impersonate and then decode traffic i think [SPEAKER_00] this is the biggest thing right the man in the middle part that you're talking about https traffic is secured and this is why the signatures those exist if they pretend to be one of those they can say hey i am an legit https like thing but then decode the traffic that's supposed to secure like credit card data and other stuff if you're still browsing and not going to https on these sites that's a problem but if you are now you update your browser i think that this is the most important part it's very [SPEAKER_00] very important to update the browsers update the os systems as fast as possible um and i would go and say as far as don't go to websites they're not big and known to purchase stuff as well and enter your credit cards is always a good advice but the best advice right now is just don't go to a random janky [SPEAKER_09] website enter credit card form by the way the certificate key what happened there is during the key signing ceremony the 384 bit key was just changed to a four which is so sketchy but also so elegant and in my opinion yeah it was definitely a human that did that because they just found out that the private key was literally just the number four like it was an llm it would have at least done something better that [SPEAKER_09] was that was definitely uh the humus it's it's pretty it's pretty funny but the update um i'm like rebooting a whole bunch of stuff right now just realized there's a lot of there's a lot of non [SPEAKER_00] again for folks who maintain servers not folks who just lousin the web there's a lot of work to do now to upgrade linuxes and php versions and node versions and etc so hug ops to everybody who spent a lot of time lately not going to sleep trying to mitigate the scope of this attacks the last thing that we do want to talk about this week also and it seems to be picking up is hacks on like major [SPEAKER_00] companies so open code was hacked what a week ago or so and they like publicly talked about this and now railway which is a deployment place where you like deploy your websites and put the websites was also hacked and like the the thing that seems to be happening there and we would love to chat with all of you guys about this is that the extraction of api keys specifically llm api keys is now what's happening because that's money that's literal money to those hackers right whereas previously you know [SPEAKER_00] in the world of uh let's say defy and and crypto you would hack something because you can like sell those tokens immediately and then get valuable money when somebody hacks a website your personal data you can sell it on the open dark web but dark web has your personal data at this point already i'm i'm not sure how much your ssn is worth right now but your api key to open ai if it hasn't have a [SPEAKER_00] limit on it that's essentially like a free and direct access to your credit card and there's a lot of folks who are buying those keys for distillation purposes on the black market there's like many many folks who cover the api keys on black market so that seems to be what's happening right now with railway and uh and open code as well which i think open code got up as well and thoughts on this folks will just keep happening like like with agents warms this is this that defender window that greg bachman talked [SPEAKER_00] about coming to a close are we at the place where like it's going to be great but for a while it's going to be sucky like how how to protect yourselves against it would love to hear opinions around like that [SPEAKER_09] specific thing just use tail scale everywhere and uh yeah tail scale two-factor and uh just just vibe code your own dependencies and vibe code your own home firewall and just just don't trust any any type of like sales or marketing material for a product if you can't see the code or your agents can't see [SPEAKER_09] if you you literally just cannot trust it i don't know what else to tell people again my my question [SPEAKER_02] something i don't i don't really have a good feel for is us people in security like everyone are they taking it like properly seriously because it's just weird how these things keep happening and like the open code team they didn't have access to the to daybreak right it just feels like such a trivial thing like surely you can just do that so i don't know like do you feel like they're taking it seriously [SPEAKER_03] or not not quite even it hasn't quite dropped yet it's it's interesting because i i think um so and you mentioned vibe coding right obviously there's a lot of people who are who are just trying to do their own thing and and and uh and i think that's where a lot of times these api keys will leak is because you know they're they're they're unintentionally putting their api key in a public place right and [SPEAKER_03] then someone can grab it right that that is one of the problems that we're seeing uh that's happened multiple times um where somebody has just uh unintentionally exposed their own uh private api key that's that's kind of problematic and and i think that um you know if again if you've never run your own service publicly and now you're trying to do that for the first time and you don't really understand kind of how these things work because most people probably really don't like the average person [SPEAKER_03] really doesn't get that um they don't understand what the potential problems are with exposing keys like that um and so you know they just go with whatever output they get it seems to work when they're testing it and so they just publish it to to the web and so sometimes that's how these things get get exposed which you know and and i don't expect that everyone's going to be a security expert right and not everybody knows to you know prompt like hey is there anything about this that like [SPEAKER_03] maybe i should consider doing to harden this before i publish this to the open web and start advertising it um it would be nice if people thought to do that or if uh that was like sort of baked in whenever you're developing something uh vibe coding something but there is a reason why it took over eight [SPEAKER_00] nine months for the bigger companies to release something like a personal agent to you like open claw and hermes which you're on on your computer and you can in charge of the security itself that's one thing there's a completely another thing to allow people to provide their passwords passwords and then save them in in like a local but secure way that the lm doesn't have access to like all that infrastructure has to be in place for people to like to go trust because like if one of those vulnerabilities goes around and actually like steals your data from your personal assistant [SPEAKER_00] it's going to be like awful for a reputation of the company there is a reason why like only now almost a year after peter steinberg innovated with the open claw we're seeing these companies step in despite them very clearly realizing the benefits of hey personal agents is like much more sticky sells more tokens it's harder to to switch from etc because there's security there well from to your to your metaphor from before building your house versus renting it so i would i would add on that metaphor given the security considerations we're talking about now this is the difference between [SPEAKER_00] building your house and not really knowing what a door bolt security bolt is and living in the high rise apartment with a security guard at the door right like muse the folks have hired moxie marlin spike the folks from signal for the e2e connections they are known for security significantly more than usually like vibe coding your server in nginx and running this on on the php which is now fully compromised which is scary the practical advice that we want to give folks is obviously if you are having [SPEAKER_00] any api tokens and you were on those exposed websites open code or railway etc rotate them like know which agent has which api token is very very important i know it's a pain in the ass but at days like this a pain in the ass is very well worth it put caps on your api tokens as well put caps and email notifications get email notifications when your api keys are like running out put you know it's better to to put auto auto refill because then you at least get notification from your credit card statement like [SPEAKER_00] your key has been auto refilled so if you see like multiple ones happening in succession you'd be able to like run and stop this bleed basically if your key was exposed rotating keys is very important and i think the the last thing is very clear for many people many people don't have this yet use a goddamn password manager use a password manager and store those keys so you don't have to remember them and put them in files it is it is increasingly more important now than it used to be before given that you know you're giving [SPEAKER_00] your keys to agents etc wolf and you want to chime in with some security practices simple for folks yeah [SPEAKER_11] you already gave the practice i wanted to mention but you should treat it like a credit card but it's even worse than that because the credit card you also have the cvc number and you often have to do a two-factor to authorize a payment and credit cards are secure and protected by the credit card companies [SPEAKER_00] i don't think your api key is going to get refunded by open ai if it was leaked like a very very big [SPEAKER_11] point like they can you know yeah so treat them even more securely than your credit yeah i i have the [SPEAKER_09] most fun advice just to install tail scale everywhere and give it to your uh go on instagram on your meta muse and then uh you just talk to uh muse on instagram and just be like hey every day just go on every single device i have and run a security check of every npm and and pip uh need to think of just give me a report i i'm joking to do it with muse i i do it with opus it but it's a very simple problem just [SPEAKER_09] every day at 6 a.m uh just check every single pip npm linux package whatever i have on my system and give me and also check up the the latest like uh snick and socket io news and give me and just an overall security report of what i need to do or send me an alert if it's something bad that's it it just runs it every day it's pretty simple that's what saved my ass almost all the time josh any any [SPEAKER_00] other chimes into like what folks can realistically do to to like at least somewhat protect themselves in this world of increasing security vulnerabilities and and agent hacks yeah i uh so generally speaking [SPEAKER_03] um the the standard security device advice just kind of applies here so um eat your vegetables type stuff yes yeah yeah yeah yeah so like obviously make sure you're using multi-factor authentication on everything uh you know make sure you're keeping your operating system your browser and everything up to date um when there are new patches issued make sure you uh apply them quickly um though the one kind of caveat here the the irony is that uh you know with some of these uh software supply chain attacks [SPEAKER_03] right level the light llm thing um uh that's um one of those awkward situations where if you are following that best practice of like updating immediately like as soon as something comes out um you could ironically end up getting compromised um instead because of that a supply chain attack it's it's kind of one of those unfortunate catch-22 situations but generally speaking for for anyone listening uh if [SPEAKER_03] if you keep everything up to date and and update pretty quickly um that's the best way you can protect [SPEAKER_00] yourself let's draw a very clear distinction here between what you keep up to date and what you need to like wait for supply chain attacks they apply to package management right for software engineers so if you are installing from uv or pip or npm or like pnp whatever those packages those packages can get exposed by hacking the folks who distribute those packages and then put like key extraction things or other things that can hurt your system do not update those just will you need it there needs to be a [SPEAKER_00] period of 24 hours at least to let the folks know that they've been hacked and like claw back the the the light limb like exposed this is what's called the supply chain attacks browsers and we haven't seen like any supply chains and browser update like update your browsers immediately update your operating systems as soon as those patches come out and for php and other places you have to patch the actual patches that they released on your system so those are like kind of like different package managers [SPEAKER_00] supply chain attacks do not update immediately in fact the best practice that we know to help you against supply chain attacks is to put a blocker to say hey i'm not installing any packages they're like at least 24 hours old because hopefully in the 24 hours that passed the the folks who got hacked and the package was like infected the the package manager you know will take it back right claw it back that so package managers do not install the new ones don't keep them up to date like joshua saying [SPEAKER_00] uh operating systems phones uh you know browsers those things need to be when you see the update notification hit that like literally update yeah well well said that's a good distinction to make uh [SPEAKER_03] between stuff that like consumers are running typically versus stuff that is running behind the scenes right so so yeah really good point about that um for a lot of software um supply chain attacks um some sometimes they're caught quickly sometimes they're caught you know within a couple of hours like the light llm thing recently so uh in other cases they might kind of linger a little longer than [SPEAKER_03] that um so uh but but i i think that's that 24 hour thing unless you know there's some vulnerability that you have to patch right now uh that was just fixed in the latest version right and if something like that does happen they're going to make a big deal out of it right they'll let people know it's going to be a well-known like public thing more than likely um and in those cases maybe you need to patch right away um if it's not that kind of a situation where you know there's some urgent [SPEAKER_03] vulnerability that's got to be patched then again when we're talking about the software supply chain kind of things um probably a good idea to to wait a little bit of time just to make sure all right [SPEAKER_00] folks we're coming to the end of the show josh thank you so much for joining us and very clearly helping folks and with with positivity in mind as well i think it's very important because my my comment to many folks in the group chats who are freaking the out if you are freaking out on x yes it's scary yes there's a lot of stuff the algorithm is poised this is what the algorithm loves when you click on the on the thing and you hover and you're like oh and even if you comment and write something in the area the the the fear and gloom and fud and like all of these things the algorithm absolutely [SPEAKER_00] loves so you are will get constantly people who are like first joining the algorithm and to get some clicks but also like people who just comment on the same stuff right but you'll be fed this fear over and over again in different forms and different types different varieties because the algorithm loves the engagement it does not mean the sky is falling go out touch some grass and then come back and run some security stuff because it is important but it's like the the internet i've seen some folks say hey go take out you know twenty thousand dollars in cash to put it because like the the banks [SPEAKER_00] are like all done scary but i don't know if we're quite there yet because yeah the companies are patching the php stuff right but definitely there's a lot of stuff that the the can of worms has been opened and be security mindful as well so yeah i think maybe this is advice we all now have access to basically security like uh experts if we can like knock on josh's door everybody to run stuff for us so definitely use that the other thing that you did bring and it's kind of like a breaking news type [SPEAKER_00] situation peter thank you for bringing this up we talked about like agent security one password finally comes out with the platform for you and your agents to be able to run they they're mentioning all of literally everyone bot and entropic and cloud and open the eye and muse and browse and town and like etc one password for agents give your agent access without giving up your passwords when agent reaches a sign-in for request credential from your vault you approve the request one password directly [SPEAKER_00] auto fills the credential in the destinations i think that's incredibly important i've been waiting for this since you know muse announced this and it's just like absolutely crazy i said i signed up let me in there so shout out to one password for like finally like getting there if you are using one password join this waitlist i think it's very very important it's kind of like the stripe link for credit cards but for passwords and i think it's very important again if you are using agents at all [SPEAKER_00] even if you are not using a one like password manager one password i found for the past 15 years to be the most reliable for me josh thank you so much for joining folks we're two and a half hours on the stream but i think it was very very important to at least mention some of this stuff because it does feel like the defenders windows coming to a close and a lot of swarms a lot of like bad actors with the eyes are now taking over like nissan said you you do need to take some steps and dedicate some time in your personal life to kind of protect yourself as well the eat your vegetables stuff are [SPEAKER_00] always there freezing your credit and and uh specifically now i think the the the more um how should i say the the the biggest exposure for folks who are not realizing that this is exposure is api keys for lms those are directly connected to your account especially if auto refill is there uh put caps put notifications on uh and then make sure that you rotate them if a website was like like railway was [SPEAKER_00] exposed or or open code with that i think we'll say thanks to folks on the stage uh and then we'll sign off this is it for thursday for october 22nd the biggest news that we talked about is this the availability security stuff we talked about um uh flint from cursive ai being built by the eureka machine which is a recursive improvement machine and it's getting a very fast local model we talked about holes in [SPEAKER_00] php and your ai keys and a lot more and tropic becoming a pbc and on tropic stepping into the personal agent with claude duet which we got access to and we'll tell you all about next week hopefully by then the wait list will open thank you so much for everybody for tuning in for listening in to thursday i josh thank you for joining peter ghost of ldj a wolf from nissan was here in young before and [SPEAKER_00] tim joined us as well thank you all for joining thursday i and we'll see you here next week bye everyone and we are