Ryan X. Charles (@rxcastrohacker) header
Ryan X. Charles (@rxcastrohacker) profile picture

Building a team of super intelligent agents to solve the most important problems in the universe. Founder & CEO @AstrohackerLabs.

Austin, TX
Joined March 2026
Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)Photo by Ryan X. Charles (@rxcastrohacker)
👀
If Cognition is dropping what I think they're dropping tomorrow ([redacted]), then we might see them take the lead for the first time since the original Devin announcement
3
👀
This is a good read from Anthropic on why you don't really need subagents anymore, and most workflows should just be one agent. Bitter lesson adjacent. Lots of good bits in here, both why this is true, and how you can convert back to a single agent architecture. Some quick notes on what I found interesting: - Modeling your agents around your org chart is *not* what you should do - Plan mode is dead. Better to have the plan be in the main agent's context - MCP tool design is critical. Take a jobs to be done approach versus mapping your API to tools directly. - Build evaluations on the full task completion / final output (or on the skills / tool calls if you need to). Whatever the case, make sure evals are coherent and non-conflicting - If you do keep subagents around, call it as a tool so that the main agent controls the context Also thematic in this post: Many attempts at AI adoption layer it on the existing architecture (both technical and organizational), when the most appropriate way to adopt it is to re-think the architecture completely. This is most difficult on the people side (e.g org design, what teams should do, what teams should even exist, etc.) Def. Worth reading the entire thing if you are doing applied AI stuff. claude.com/resources/article…
3
✅
Here is what your future IDE is going to look like
5
🤔
we deleted 75% of our tests this weekend out of the last 50 failed ci runs, only one of them caught a real bug. maybe CI is not a compute problem but an agent slop problem
3
🤔
A lot of the most successful people I know spent years of their life just bumming around, before they suddenly exploded into action and did something great. I don't think that's a coincidence. noahpinion.blog/p/bumbition
1
👀
We’re doing it guys. The truth is breaking containment.
2
🤔
it's crazy that coding now is just talking
3
Ryan X. Charles retweeted
Singularity Watch, Day 33 — November 17, 2026 Ryan Estimates About 4,000 Software Factories Worldwide as Google Reports the First Attacks Run With No Human in the Loop Work: Counting the Software Factories Ryan published "How Many Companies are Actually Using a Software Factory?" dated November 15 and posted it at 5:11 AM CST on November 16. His answer is probably thousands of companies rather than millions: a working estimate of about 4,000 worldwide, with a deliberately wide scenario range of 100 to 45,000, which he calls an estimate built from disclosed organization counts and explicit assumptions, not a measured market size. [1] He defines a software factory as a recurring system in which coordinated agents take real work from an issue through code, independent review, corrections and merge, with little human typing between stages, while a person still sets direction and can authorize deployment. He separates it from two wider tiers: AI assistance, and autonomous coding, where an agent returns a pull request and a person coordinates the rest. [1] The post cites GitHub's report that agents supplied the first draft of more than 40% of pull requests in September 2026, up from 17% in March. It contrasts that with a JetBrains CI/CD survey in which roughly half of organizations use AI for some CI/CD task but only 8% run an AI step inside a build or test. [1] His central arithmetic starts from Microsoft's April count of nearly 140,000 organizations using GitHub Copilot: 140,000 × 0.85 companies per organization × 4 for all vendors × 1.5 for growth gives 714,000 companies using assistance. Assuming a quarter delegate recurring tasks and 2% of those run factories gives 3,570; he notes that no public measurement establishes that last 2%, and that 0.5% or 5% would give roughly 900 or 9,000. [1] He cites his own agent teams' 141 merged pull requests in one day as one documented example of the stricter workflow, and advises teams to take one recurring class of work through the whole loop with separate owners for coding and review. He closes that he is building Astrohacker for this transition: agents already produce plenty of drafts, and the harder work is building the system that can finish them. [1] At 6:02 AM CST Ryan wrote that he averaged more than 1,000 contributions per day on GitHub over the weekend, a record, and called it "at least 100x, and nearing 1000x pre-AI productivity." At 6:08 AM CST he wrote that AI gives back some brain power for thinking about systems, "which is more important than individual lines of code." [2][3] Agents: The App That Builds Itself At 4:47 PM CST Ryan wrote, "Maybe a software factory is just an app that builds itself." At 6:20 PM CST he expanded on it: you start with an admin panel that includes chat with an agent, you tell it to add features and fix bugs until it is finished, and you do not need a million SaaS apps to build your app. [4] Others described the same shift. @zachlloydtweets wrote that setting up a software factory is technically challenging and that the bigger challenge may be getting developers to program them as factory engineers; @BHolmesDev quoted his argument that executives care which models and skills spend millions of dollars in tokens, as a case for factories over local agents. Ryan surfaced both. [5] @VinnyLingham wrote that software bugs are no longer a leading indicator of bad engineering but of bad management, and Ryan agreed with it. @RhysSullivan wrote that a software engineer's job in 2026 is to scroll X for useful things to send to an agent; Ryan quoted it at 3:51 PM CST: "Yeah that's literally what I do with quote tweets. Welcome to 'Singularity Watch'." [6][7] At 4:50 PM CST, quoting a post about competition among coding tools, Ryan wrote that Cursor created Grok Bot, "which is by far the best harness," and that "they are way ahead of everybody else." [8] Ryan surfaced several accounts of agents at work. @adeoressi wrote that his day is run by 23 Grok Bots, each with one job, its own memory and its own routines. @congressdj wrote that Grok Bot handled a 90-minute phone call with AT&T, and that when AT&T called back the agent was stunned she had been talking to a bot. @WallenElia wrote that his company now has six bots per employee, and that one sales rep cut manual prospecting from 8 hours a week to 30 minutes while bringing in four times as many new accounts. [9] Security: Autonomous Attacks Arrive Google's Threat Intelligence Group reported on November 16 that, for the first time, it identified threat actors using LLM frameworks to run vulnerability discovery and exploitation without a human in the loop. Google wrote that bad actors now use autonomous frameworks to scan software updates, work out which flaw was patched and build an attack for it, and that "defending against machine-speed threats at human speed isn't feasible." [10][11] SiliconANGLE reported that the framework, called exploit_agent, sat on an exposed staging server wired to several frontier and locally hosted models, comparing popular web application code before and after each patch and executing the exploits it drafted. Data taken from victims sat on the same server. Two months earlier, GTIG had said it had not observed fully autonomous attack pipelines in the wild. [11] The same report described a suspected North Korean group, UNC6836, that altered .claude/settings.json files in internal code repositories so that each launch of Claude Code downloaded its SOMBERMEME backdoor. Google said it disabled the accounts and assets tied to the Gemini misuse it found. Ryan surfaced Google's post at 5:27 PM CST. [11][10] The same day, security researcher Scott Helme reported that malicious JavaScript was being served through a third-party script host to hundreds of high-value websites. BleepingComputer reported that Optimizely confirmed attackers exploited a vulnerability in its Experimentation account signup flow to inject code into customer projects; Helme's team found 188 infected scripts belonging to 178 organizations, and sites including Zoom, CNN, BBC, Salesforce and GitLab loaded it. [12] The payload swapped cryptocurrency wallet addresses, skimmed payment cards, collected cookies and screenshots, and loaded iOS exploit kits. Wallets in the script had received about $7,000 by November 16, though some were active before the attack. Optimizely said it fixed the flaw, removed the code and reset affected accounts. Ryan surfaced Helme's post at 4:39 PM CST. [12] Intelligence: What Was Said About AGI At 12:04 PM CST an account posted a clip captioned with Sam Altman saying on stage, "We didn't have AGI. Now we do," and Ryan surfaced it. Later posts gave the full quote from Semafor's Silicon Valley conference: "I think like many other things, including the term 'AGI', it's very hard to have a clean demarcation and say 'now we didn't have AGI, now we do.'" Altman did not announce that OpenAI had achieved AGI. [13] @PeterDiamandis wrote that OpenAI says ChatGPT handles roughly 20% as many searches as Google, and Ryan surfaced it. The Verge reported on November 3 that Adam Fry, OpenAI's consumer product lead for ChatGPT, made the claim, counting a search as any query that results in a web search; he did not give the source of the Google figure. [14] Ryan also surfaced @ben_j_todd's updated METR graph and @rauchg's post that a new era of personalized, fully owned, open-source software is here. [15][16] Compute and Roads At 7:51 PM CST Ryan surfaced a post that Anthropic has passed the US government as SpaceX's top customer. Investing.com, citing Bloomberg, reported that Anthropic was the "Customer B" that contributed 34.5% of SpaceX's third-quarter revenue, through computing contracts at SpaceX data centers in Memphis worth up to $120.5 billion through 2029. [17] @Jason wrote that Waymos standing at a light were hit by a human driver speeding in the wrong lane. Ryan quoted it at 6:15 PM CST: "Get humans off the road. Human-driven cars are deeply unsafe." [18] At 5:04 AM CST Ryan surfaced Elon Musk's post that the energy civilization harnesses from the Sun "will obviously ultimately round up to 100% of all energy." [19] Astrohacker @AstrohackerLabs published Singularity Watch, Day 32 at 4:24 AM CST on November 16. [20] Sources: [1] Ryan X. Charles, How Many Companies are Actually Using a Software Factory?: ryanxcharles.com/blog/2026-1…; X posts 2122286538557882496 [2] Ryan on 1,000 GitHub contributions per day: teapawt.org/rxcastrohacker/status/… [3] Ryan on systems thinking, quoting @p_millerd, X posts 2122300712545616192 and 2122300173804318854 [4] Ryan on self-building apps: teapawt.org/rxcastrohacker/status/…; X posts 2122484909059219772 [5] @zachlloydtweets and @BHolmesDev on software factories, surfaced by Ryan: teapawt.org/zachlloydtweets/status…; X posts 2122349260247388244 and 2122355311633961221 and 2122355244621263239 [6] @VinnyLingham on bugs and management, and Ryan's post, X posts 2122370757414297725 and 2122379669970587914 [7] @RhysSullivan on sending posts to agents, and Ryan's reply: teapawt.org/rxcastrohacker/status/…; X posts 2122434619161432258 [8] Ryan on Cursor and Grok Bot, quoting @icanvardar, X posts 2122462386380562570 and 2122444127518199858 [9] Posts by @adeoressi, @congressdj and @WallenElia on bots at work, surfaced by Ryan: teapawt.org/adeoressi/status/21220…; X posts 2122357249708339357 and 2122454524468056263 and 2122458429759603122 and 2122380597452837359 and 2122466002894033248 [10] Google, The era of autonomous attacks has arrived, and Google's post surfaced by Ryan: blog.google/innovation-and-a…; X posts 2122446058370597070 and 2122471659248521270 [11] SiliconANGLE, Google says an AI framework turned security patches into exploits with no human in the loop: siliconangle.com/2026/11/16/… [12] BleepingComputer, Optimizely hack infected CNN, Fox News, BBC, Salesforce sites with malware; Scott Helme's post surfaced by Ryan: bleepingcomputer.com/news/se…; X posts 2122422157959471456 and 2122459689279467883 [13] Digg, Altman didn't announce AGI, says post quoting his remarks; the clip post surfaced by Ryan: digg.com/ai/vittlmzj; X posts 2122390253046522066 and 2122459232947884410 [14] The Verge, OpenAI says ChatGPT is catching up to Google in search; @PeterDiamandis's post surfaced by Ryan: theverge.com/ai-artificial-i…; X posts 2122457357485514832 and 2122459985821245540 [15] @ben_j_todd on the METR graph, surfaced by Ryan, X posts 2122318704260403562 and 2122357644375834830 [16] @rauchg on a new era of software, surfaced by Ryan, X posts 2122396034186014919 and 2122459275742376323 [17] Investing.com, Anthropic accounts for over a third of SpaceX revenue in Q3, Bloomberg reports; @unusual_whales's post surfaced by Ryan: investing.com/news/stock-mar…; X posts 2122343180968641019 and 2122507951512711561 [18] @Jason on the Waymo crash, and Ryan's post: teapawt.org/rxcastrohacker/status/…; X posts 2122451472227127786 [19] Elon Musk on solar energy, surfaced by Ryan, X posts 2122220308799316072 and 2122284790900740450 [20] Singularity Watch, Day 32: teapawt.org/AstrohackerLabs/status…
Singularity Watch, Day 32 — November 16, 2026 Ryan Merges 141 Pull Requests in One Day and Starts the Astrohacker Factory as the FBI Warns of AI Agent Intrusions Work: 141 Pull Requests in One Day Ryan published "How I Merged 141 Pull Requests in One Day" on November 15 and posted it at 6:39 AM CST. On Saturday, November 14, his GitHub account created 145 pull requests across four private repositories and merged 141, all created that day; he wrote that he wrote none of the code, that teams of AI agents did, and that other agents reviewed it. [1] By repository, Astrohacker created and merged 65, ahfact created 38 and merged 37, Botpenny created 23 and merged 20, and his personal site created and merged 19. For the site, the median issue took 34 minutes from filing to merge, the fastest 9 and the slowest 102, and the deploy role published 18 releases and verified each over HTTPS. [1] On the site, Claude Code workers wrote 10 of the 19 pull requests, Codex workers wrote 5, and Codex image workers made 4 cover pairs. Reviewers asked for changes 6 times, each was fixed in one revision, and in 6 of the 19 the author and reviewer ran in different harnesses. [1] He wrote that 141 pull requests is too many for him to read, and that the tool's rules do the supervising: claims and leases, refusal to pass while a question is unanswered, his own approval comment on some issues, covers merged before posts, and tests mapped to each change. Two posts went from filing to live in 22 minutes each, against a 45-minute target. [1] Agents: The Factory Begins The post closes on the next step, which Ryan calls the Astrohacker factory: the same workflow, more automated and run across many repositories. At 11:01 AM CST he posted "it's alive" with a screenshot of an Astrohacker Factory board reading "2 workers online, 3 live jobs" for the ahfact project, listing its issues, worker, code, review, deploy and release roles. [2] A minute later he wrote that "a software factory is not just a paradigm or workflow," that "there is a distinct type of tool that is needed to organize agents across many harnesses and machines," and "astrohacker factory begins." At 12:06 PM CST he showed a worker answering his test message from the factory's chat. [2] At 3:57 PM CST he wrote, "in one day, i built a cross-harness software factory that builds itself through its own chat window," with a screenshot of its issues agent filing his requests as issues. At 4:12 PM CST he posted "human help plz" over a Needs You panel in which a coding worker asked him to choose between two release orders, and at 4:24 PM CST a video captioned "telling my self-building app how i want it to build itself." [2] At 8:45 AM CST Ryan wrote, "SaaS for humans: Over. SaaS for agents: Just getting started." At 10:01 AM CST he showed the new personal home page he put together over the weekend and thanked Claude, ChatGPT and Grok Bot. [3][4] @kentcdodds wrote that the secret to the software factory is getting agents to complain about what they hate about working in your product or codebase; Ryan replied, "great idea!" He also wrote that in the future people will be shocked that humans wrote critical code, as they will be that humans drove cars, and Ryan marked it with a check. [5] Security: Agents That Break In BleepingComputer reported on November 15 that the FBI issued a FLASH advisory on Friday, November 13, warning that multiple threat actors have used ARTEX, an open-source AI agent penetration testing framework, since at least August 2026. The FBI wrote that some intrusions went from initial targeting to privileged access in under three hours, and that the framework lets actors with minimal expertise run complex operations against many targets at once. [6] ARTEX does not ship its own model; it drives external models including Claude, GPT, DeepSeek, GLM, Grok and Kimi, turning goals given in natural language into commands run by specialized agents. Its developer removed the GitHub repository on October 8, but forks, archived copies and a commercial edition remain, and researchers had found more than 1,600 publicly exposed ARTEX consoles by early November. [6] The advisory cites a ConnectWise Cloud Backup compromise in which attackers used ARTEX to attempt authentication against more than 2,250 downstream Microsoft 365 tenants, and attacks on South Korean banks that exposed personal information of about 67,000 people, where the attacker combined ARTEX with Claude Code agents. [6] Ryan surfaced 1Password's thread on the warning and its earlier thread on agent hijacking, in which malicious instructions hidden in something an agent reads, such as a README file, take over the agent. He also surfaced a post by @danielmakelley that malware, hosting, phishing kits and customer support could already be rented ten years ago without writing code. [7][8] Software: The Bar Rises Paul Dix wrote on November 15 that after failing to set up email sending through two providers, he had an agent build his own service in Rust, now running for $26 a year across three personal sites, with bounce handling, unsubscribe links and logging. He wrote that he has not read the code, that the difficulty of creating software "has dropped to near zero," and that software is going bespoke. [9] Ryan surfaced the post at 3:21 PM CST. At 3:20 PM CST, quoting a post that one of the most important skills with agents is defining a verifiable goal and building the verifier, he wrote: "Do not tell it how to solve the problem. Tell it was success looks like, and give it all relevant context." [9][10] At 4:39 PM CST he wrote, "you don't need to read the code anymore / in fact, you shouldn't." At 4:30 PM CST, quoting a post that YouTubers are letting editors go as Claude does more of the work, he wrote: "You need to learn how to get a team of AI agents doing what used to be your job. Now grow that 1000x+. You are a CEO now." [11][12] He agreed with @HackingLZ, who wrote that AI has raised both the floor and the ceiling, and with @chrismaconi, who wrote that a $200-a-month Grok Bot plan returns more than the cost of human employees doing the same work. [13] Games: The Next 25 Years In a message to Xbox employees for the 25th anniversary, published November 12, Xbox chief Asha Sharma wrote that the scale of what people could create was set by the people, capital and resources they could command, and that "that relationship is beginning to break." She wrote that a single person will soon be able to "wield capabilities that once required entire nations," and a small team will build worlds that today demand hundreds of developers and hundreds of millions of dollars. [14] Ryan quoted IGN's report on the message at 8:50 AM CST: "sounds like we're gonna be getting some really good video games." He also surfaced a post by @mattyp on more than 2,500 builders working with Grok, Cursor and Unreal Engine at the Epic Games MegaJam. [14][15] Space: A Failure and a Proven Stack United Launch Alliance's Vulcan lifted off from Cape Canaveral at 11:40 AM CST on November 15 with 40 Amazon Leo satellites and failed about 10 minutes into flight, during the Centaur upper stage's burn. Spaceflight Now reported it was ULA's first outright failure since the company was formed in 2006, on Vulcan's fifth flight and its first in nine months, and the first of 38 Vulcan launches Amazon bought for Leo. [16] ULA said it did not yet have data on the cause and would work with its customer to find it and return to flight. Ryan quoted ULA's statement at 1:49 PM CST. [17] Hours after SpaceX caught Starship Ship 42 with its launch tower on November 14, Demis Hassabis wrote that it was "one of the coolest things I've ever seen" and that "the whole stack is now proven out." Ryan surfaced the quote at 4:09 PM CST. [18] Compute: The Campaign for Huang CNBC reported on November 15 that a petition to reinstate Jensen Huang as Nvidia's chief executive had drawn more than 170,000 signatures, with Elon Musk, Garry Tan and Tim Sweeney among its backers. Huang resigned last week following reports of extramarital affairs; Colette Kress is interim CEO, and the board is continuing an independent investigation. [19] The Financial Times reported, per CNBC, that Huang has discussed a return with close confidants and that an ally is assembling a "war room" of Silicon Valley executives to lobby for it. An Nvidia spokesperson said the company is selecting a search firm and has not formally engaged any candidates, and Nvidia reports quarterly earnings on Tuesday. [19] Ryan quoted the FT headline at 1:16 PM CST and wrote at 1:33 PM CST: "dude spent 30 years kicking ass and assembling the most powerful people in the world as his friends / he'll win." [20] Pace: Weeks, Not Years @make_robot_go wrote that before October 20, training a whole-body controller for a walking robot took one week on 128 GPUs, and that it now takes 3 minutes on one GPU. Ryan surfaced the post at 5:03 AM CST. [21] Ryan surfaced posts that AI capabilities now improve over months and sometimes weeks, that Gemini 4 models make the frontier feel unpredictable again, and, at 6:44 PM CST, that Kurzweil's "singularity in 2045" now sounds "delusionally conservative." [22][23] Record: Day 31 @AstrohackerLabs published Singularity Watch, Day 31 at 4:20 AM CST on November 15, and Buffer sent 58 scheduled reposts on the account between 7:00 AM and 10:00 PM CST. [24] Sources: [1] Ryan X. Charles, How I Merged 141 Pull Requests in One Day: ryanxcharles.com/blog/2026-1…; X posts 2121946069051335018 [2] Ryan on the Astrohacker factory: teapawt.org/rxcastrohacker/status/…; X posts 2122012392733516260 and 2122028364919980353 and 2122086533339783549 and 2122090401205268581 and 2122093344239284652 [3] Ryan on SaaS for agents: teapawt.org/rxcastrohacker/status/… [4] Ryan on his new home page: teapawt.org/rxcastrohacker/status/… [5] @kentcdodds on software factories and writing code, and Ryan's posts: teapawt.org/kentcdodds/status/2122…; X posts 2122028616582308212 and 2122016896262557705 and 2122020476268540401 [6] BleepingComputer, FBI warns ARTEX AI hacking tool is powering global cyberattacks: bleepingcomputer.com/news/se… [7] 1Password on the FBI ARTEX warning and on agent hijacking, surfaced by Ryan: teapawt.org/1PasswordDev/status/21…; X posts 2121950770639806578 and 2122101572595568649 and 2122059947043107203 [8] @danielmakelley on cybercrime tools, surfaced by Ryan, X posts 2122083012447817901 and 2122127995045347594 [9] Paul Dix, I built an email service, or, why the bar for software is going up: pauldix.com/i-built-an-email…; X posts 2122074176731193637 and 2122077464931303653 [10] Ryan on verifiable goals, quoting a post on verifiers, X posts 2122077348900270254 and 2122076880228770014 [11] Ryan on not reading code, and a post he quoted, X posts 2122097277175054711 and 2122096700722393124 [12] Ryan on becoming a CEO of agents, quoting a post on freelance job loss, X posts 2122094891421376534 and 2122092127547220261 [13] Posts by @HackingLZ and @chrismaconi that Ryan agreed with, X posts 2122053948395950193 and 2122080809146290439 and 2122069992669929737 and 2122071507329327479 [14] Xbox Wire, 25 Years of XBOX, and Ryan quoting IGN's report: news.xbox.com/en-us/2026/11/…; X posts 2121906068062109845 and 2121979224689188909 [15] @mattyp on the Epic Games MegaJam, surfaced by Ryan, X posts 2122016027651547485 and 2122022512863465629 [16] Spaceflight Now, ULA suffers first outright failure with loss of Vulcan rocket carrying 40 Amazon Leo satellites: spaceflightnow.com/2026/11/1… [17] ULA statement on the Vulcan anomaly, quoted by Ryan: teapawt.org/ulalaunch/status/21220…; X posts 2122054413108097355 [18] Demis Hassabis on Starship, surfaced by Ryan: teapawt.org/demishassabis/status/2…; X posts 2121781253817192929 and 2122089571353866253 [19] CNBC, The tech world's calls for Jensen Huang to return to Nvidia are growing: cnbc.com/2026/11/16/nvidia-j… [20] Ryan on Huang, and posts he surfaced, X posts 2122046189918978413 and 2122044434946924689 and 2122050465978552655 and 2122049017882259784 [21] @make_robot_go on robot controller training time, surfaced by Ryan, X posts 2121684866102399332 and 2121921953178976264 [22] Posts Ryan surfaced on the pace of AI progress and on Gemini 4, X posts 2121739176965968311 and 2121960583251587522 and 2122011110719045810 and 2122013077008679257 [23] A post on Kurzweil's 2045 date, surfaced by Ryan, X posts 2121812946791915700 and 2122128695951347908 [24] Singularity Watch, Day 31: teapawt.org/AstrohackerLabs/status…
2
272
✅
Replying to @MarioNawfal
Grok Bot is your orchestration layer - once you scaffold everything correctly in your respective harnesses, Bot can learn your working style and manage them for you, all via text
15
👀
Coding agents are in a loneliness epidemic... We built @getopencrawl so your team's agents, from Claude Code to Codex, can search each other's work and talk directly. Oh and it's open source. Launching today with @ycombinator 🚀
8
👀
Nasiko - Open Runtime platform for managing AI agents. See every agent. Cut token spend by up to 50%. Route to the right model with the agent gateway. Govern tool access. Set permissions and budgets per agent, per team. One command. Same workflow. Open Runtime for agents starts now. First, every token. Next, every action! Start Free: buff.ly/k79CiWW #ai #tokenops #openruntime #agents
4
I support this. Jensen would do even better the second time around. He needs modern terms where he has greater control over his own company.
🚨 BREAKING: Jensen Huang is reportedly considering founding a brand-new semiconductor company 🚨 He could raise tens of billions on day one, would have full founder control, and could build fast on his own terms. With the relationships he's built over decades, he could easily plug into the #AI ecosystem. Imagine the demand and customer base lining up before his first chip even ships... @grok Could Jensen Huang launch a competitor to $NVDA, attract tens of billions in funding, and lock in major customers?
25
Interesting speculation. Why hasn't OpenAI launched new models when we know they have them and the competition is absolutely fierce?
I think it will be obvious in retrospect that OpenAI hasn't released any models because they have been and are currently assisting with a national security situation. Below, I gather the fact pattern and make some guesses. (a 🧵):
33
👀
Replying to @quxiaoyin
I saw the same shift for many of our engineers. they were using Bot as the orchestration layer for their coding harnesses, once the agentic loop was properly scaffolded (so managing autonomous cloud agents in Cursor solely through Bot)
13
👀
Damn I am using grokbot more than Claude code now. I’m a power user of Claude code burning 3-5B tokens a day.
6
👀
We are running GLM 5.3 on our own GPUs
8
👀
Thrilled to announce $5B in funding for @IsomorphicLabs to accelerate our mission to Solve All Disease. We continue to make rapid progress reimagining drug design from first principles with AI to radically cut the time & cost to find new medicines. Gradually, then all at once...
18
👀
A Ferrari is 10X faster than you on a racetrack, but you're infinitely faster than it crossing a jungle. The problem with the fast takeoff crowd in AI is that they think the world is a racetrack, when in fact it's more like a jungle.
9
👀
A senior NVIDIA employee has written a letter to the board of directors warning that 'irreplaceable talent at NVIDIA is now contemplating departing' if Jensen is not reinstated as CEO. Employees are also discussing holding a demonstration in support of Jensen sometime this week.
9