Your curated collection of saved posts and media
we are about to hit 1 9 of availability while coding is largely solved https://t.co/4NJB1YNsPk
we are about to hit 1 9 of availability while coding is largely solved https://t.co/4NJB1YNsPk
It's very common for people to claim that open LLMs will be used to commit cyber attacks at massive scale. What public evidence is there for this claim? The best (and one of the only) accounts I've seen of a cyber LLM attack was done using Claude https://t.co/v63Lolv5iH
Looking for user feedback about the upcoming ggml official Debian and Ubuntu packages https://t.co/8lcGZzSgLK
ๅ ้ฑใSakana AIใฎใชใผใๆ่ณๅฎถใงใใKhosla VenturesใฎVinod Khoslaๆฐใๆฅๆฅใใพใใใ๐ฏ๐ต ๅฝ็คพๅ ฑๅๅตๆฅญ่ ใฎไผ่คใจใจใใซ็ๅฑฑใใคใ่ฒกๅๅคง่ฃใๅ ฑใซ่กจๆฌใใๆฅๆฌใฎ็ฃๆฅญ็ซถไบๅใ้ซใใAIๆฆ็ฅใใๅ ฌๅ ฑ้จ้ใซใใใๆๆฌ็ใชAIๆดป็จใซใคใใฆใใฐใญใผใใซใช่ฆ็นใใๆ่ฆไบคๆใ่กใใพใใใ ็ถใใฆใVinodๆฐ ใซใฏSakana AIใฎๆฐใชใใฃในใซใใ่ถใใใใ ใใพใใใCEO David Ha (@hardmaru) ใ CTO Llion Jones (@YesThisIsLion) ใไบคใใๆฅๆฌใฎๅฎๅ จไฟ้ใป้ฒ่กๅ้ใซใใใAIๅฉ็จใๅซใใๅตๆฅญๆใใๆๅพ ใใใใฆใใใ ใใฆใใๅฝ็คพใฎใฆใใผใฏใชๆ่กใๆดป็จใใๅฝๅ ๅคใฎใใพใใพใช็ฃๆฅญใปใฏใฟใผใซใใใAIๅฑ้ใฎๅฏ่ฝๆงใซใคใใฆๅฏพ่ฉฑใ่กใใพใใใ โโโโโโโโโโโโ Last week, Vinod Khosla (@vkhosla) of Khosla Ventures, Sakana AIโs lead investor, visited Japan. ๐ฏ๐ต Together with our Co-founder COO Ren Ito, they paid a courtesy visit to Finance Minister Satsuki Katayama to exchange views from a global perspective on AI strategies to boost Japanโs industrial competitiveness and the fundamental integration of AI within the public sector. Following that, Vinod visited Sakana AIโs new office. Joined by CEO David Ha (@hardmaru) and CTO Llion Jones (@YesThisIsLion), we discussed the potential of deploying AI across various domestic and global industrial sectors using our unique technologyโsomething he has supported since our founding. This included conversations on utilizing AI in Japanโs security and defense fields.
Can AI companies restrict government use of their technology? They do it all the time. Whether and how depends on the acquisition pathway, contract type, and terms. My explainer: https://t.co/QHSZrooFoH #Anthropic #openai #pentagon #DoD #govcon
@CharlieBul58993 @JTillipman @bridgewriter (former NSC counsel) - https://t.co/K8WEStCDhc
A deep dive in @lawfare on the many legal problems with the Pentagon's designation of Anthropic as a supply chain risk. https://t.co/6mlWhgwMge
New research just exposed the biggest lie in AI coding benchmarks. LLMs score 84-89% on standard coding tests. On real production code? 25-34%. That's not a gap. That's a different reality. Here's what happened: Researchers built a benchmark from actual open-source repositories real classes with real dependencies, real type systems, real integration complexity. Then they tested the same models that dominate HumanEval leaderboards. The results were brutal. The models weren't failing because the code was "harder." They were failing because it was *real*. Synthetic benchmarks test whether a model can write a self-contained function with a clean docstring. Production code requires understanding inheritance hierarchies, framework integrations, and project-specific utilities. Different universe. Same leaderboard score. But it gets worse. A separate study ran 600,000 debugging experiments across 9 LLMs. They found a bug in a program. The LLM found it too. Then they renamed a variable. Added a comment. Shuffled function order. Changed nothing about the bug itself. The LLM couldn't find the same bug anymore. 78% of the time, cosmetic changes that don't affect program behavior completely broke the model's ability to debug. Function shuffling alone reduced debugging accuracy by 83%. The models aren't reading code. They're pattern-matching against what code *looks like* in their training data. A third study confirmed this from another angle: when researchers obfuscated real-world code changing symbols, structure, and semantics while keeping functionality identical LLM pass rates dropped by up to 62.5%. The researchers call this the "Specialist in Familiarity" problem. LLMs perform well on code they've memorized. The moment you show them something unfamiliar with the same logic, they collapse. Three papers. Three different methodologies. Same conclusion: The benchmarks we use to evaluate AI coding tools are measuring memorization, not understanding. If you're shipping code generated by LLMs into production without review, these numbers should concern you. If you're building developer tools, the question isn't "what's your HumanEval score." It's "what happens when the code doesn't look like the training data."
Gift link: https://t.co/S1D5ZMpE3l
๐ซ @bradlightcap has stopped following @GaryMarcus (๐ค๐: any thoughts on this?) https://t.co/kI0mBNCoxY
Gift link: https://t.co/S1D5ZMpE3l
In the last few days, OpenAI and its executives have claimed that its DoW deal prevents its models being used for mass domestic surveillance. As I write in a lengthy explainer for @ReadTransformer today, that appears to be misleading at best. https://t.co/IdlpVUSY0p
Be like Sam Altman > runs YC > starts a open-sourced non profit to regulate ai & protect humanity > raise money for non profit > use that money to build a closed source AI > create a new for profit company > raise money & kick out existing investors > use our data for ads in ChatGPT > go on the news and stand up for Anthropic against US gov > 24hrs later sign a deal with US to do exactly the opposite
Folks, this is not normal. Four American soldiers have died, but let me tell you about the curtains. โI always liked gold.โ https://t.co/1Kt9tvNi8g
Shocked! https://t.co/8EX9ADZibS
Satya Nadella just said what the entire industry is too invested to admit. Every CEO signing $100 billion data center contracts right now is making a bet that history may not honor. Nadella: โWe are one sort of innovation away from the entire regime changing.โ Right now, every major player is running the same play. More data. More GPUs. Bigger clusters. Same architecture. Theyโve convinced themselves scale is destiny. Theyโve convinced themselves the biggest balance sheet wins. Theyโve convinced themselves this is a resource war. Itโs not. Nadella: โIf you look at where weโve gone, it was all about pre-training scale, then it was about post-training, then we came up with reasoning, then we said, โoh, thereโs RL.โโ The architecture isnโt stable. It never was. Itโs been mutating the entire time. Each shift rewriting the rules. Each breakthrough making the previous moat irrelevant. And the companies that didnโt see it coming didnโt get a warning. They just woke up behind. Nadella: โA new model architecture that could even be more efficient in its performance.โ When that lands, the $100 billion clusters donโt matter. The hoarded GPUs donโt matter. The multi-decade infrastructure advantage doesnโt matter. Every castle built for the current paradigm becomes a monument to the wrong bet. This is what makes the AI race unlike anything in history. In nuclear competition, more warheads meant more power. The advantage was permanent. Cumulative. Compounding. In this race, one person with the right insight at 2am in an apartment somewhere erases a trillion dollars of infrastructure before the market opens. No warning. No negotiation. No second place. The most dangerous competitor in this race doesnโt have a data center. They just have the equation.
Anyone else having those weird dreams where future generations hate you? From @TheOnion https://t.co/y2OI4SqdpC
Markets have a history of overreacting to narratives long before underlying economics change. AI is no exception. Capital tends to swing between euphoria and panic when it struggles to price uncertainty. The work, as always, happens between those extremes. What's your take on this development? @pchamard @Khulood_Almani @antgrasso @GlenGilmore @Shi4Tech @CurieuxExplorer @FrRonconi @chidambara09 @theomitsa @Analytics_699 @Nicochan33 @nafisalam @pierrepinna @smaksked @Corix_JC @amalmerzouk @AdityaRPatro @quepasachico @IngridVasiliu @EstelaMandela @sonu_monika @RLDI_Lamy @SpirosMargaris @IanLJones98 @Timothy_Hughes @avrohomg @bimedotcom @HaroldSinnott @c4trends @mvollmer1 @DG_Collective @bamitav @rwang0 @ipfconline1 @sijlalhussain https://t.co/ZxZhQSG0py
โI probably spend a third, maybe 40%, of my time making sure the culture of Anthropic is good,โ Anthropic CEO Dario Amodei said. https://t.co/wATWuk7bZO https://t.co/Pbo7qTjhEm
Larry Ellison $ORCL highlighted something critical: models like ChatGPT, Gemini, Grok, and Llama are all trained on largely the same public internet data. When everyone trains on the same information, models inevitably converge. Thatโs why AI is moving toward commoditization. The real moat isnโt the model itself. Itโs the proprietary data behind it. Companies that can train on exclusive datasets gain an advantage competitors canโt replicate. Having data that no one else has will allow you to dominate your market.
Warren Buffett: "I like the way I lived 30 years ago โ and I live that way now. The only difference is I have a plane to travel around privately. But in terms of what I eat, the clothes I wear, the books I read, the television I watch ... it's what I want to do in life." https://t.co/0XiZu1RgMJ
The CEO of a $380 billion AI company said something that should concern every developer, every startup, and every government on earth. He called open-source AI a "red herring." This is Dario Amodei, the man who runs Claude. His argument sounds technical but it's not and it's about money. Here's what he said: "I don't care whether a model is open source or not. The only thing I care about is, is it a good model?" Sounds reasonable, until you look at his books. 75% of Anthropic's $14 billion in revenue comes from one thing, charging companies per token to use Claude through an API. If enterprises could run their own models for free that revenue disappears. So when Amodei says open source "doesn't matter," what he means is, please don't look at open source. His technical argument, AI "open source" isn't real open source. You get the weights, just numbers not the actual source code and you can't see inside the model. Fair point but it misses the bigger picture. Companies running open source models don't need to see inside. They need three things, lower costs, data privacy, and freedom from vendor lock in. Open source somewhat delivers all three. A Berkeley study found open source AI models cost up to 90% less than closed APIs. Hospitals can keep patient data in house. Banks can meet compliance rules. Defense contractors don't send classified data to someone else's servers. Amodei brought up DeepSeek to prove his point. "I don't think it mattered that DeepSeek was open source," he said. But DeepSeek's release crashed Nvidia's stock by the biggest single day loss in market history and it mattered. Here's the pattern that keeps repeating in tech: Linux was a toy until it ran the internet. Android was "fragmented" until it owned 72% of mobile. Open source starts cheap, then it gets good, then it wins. Anthropic is posting record revenue. 500+ customers spending over $1M a year and 8 of the Fortune 10 on board. But so were BlackBerry and Sun Microsystems. So was every incumbent that dismissed the disruptor. The real question isn't whether open source is a red herring. But the real question is whether the man running a $380 billion closed source empire is the right person to ask.
Anthropic just launched a weapon against OpenAI. Not a new model, benchmark, or price cut. But a copy-paste button. Here is what is happening and why it should terrify every AI company on earth. For two years, Silicon Valley assumed one thing would keep you locked into ChatGP
The price of intelligence is falling fast. As AI becomes cheaper and more capable, agentic systems are starting to take over tasks once done by human workers. The shift isnโt gradual. Itโs economic. https://t.co/qpha8z0s7S @abcnews @AlanKohler
Inside Amazonโs layoffs, AI and โleanerโ operations are reshaping the culture. Survivorโs guilt and rising workloads are becoming part of the transition as automation accelerates. Whatโs happening there may become a blueprint other companies quietly follow. https://t.co/cRMN54FsNi @ft @rafeuddin_
AI is starting to shape scientific discovery itself. In particle physics, machine learning systems inside detectors now decide which signals are worth keeping and which are discarded. When algorithms filter reality, they quietly influence what scientists get to study. https://t.co/xjGX5V2QEZ @IEEESpectrum
Investors are piling into AI-resistant โhaloโ stocks. Heavy-asset, low-obsolescence companies are driving UK and EU markets to record highs. In the AI era, scarcity and stability are suddenly back in favor. https://t.co/aQwauFrnR9
McKinsey says agentic AI could fundamentally reshape global banking. But it warns banks not to get trapped in endless pilots and proofs of concept. The competitive edge wonโt go to those experimenting the longest, but to those scaling fastest. https://t.co/rrt7ROCCBl @DigWatchWorld @mckinsey
AI is beginning to decode the electrical noise inside our brains. Signals once thought too complex to interpret are now being translated into patterns and meaning. When machines start reading inner thoughts, neuroscience enters an entirely new era. https://t.co/4ZB8TUqxiG @LauraCReporter @bbcnews
AI literacy isnโt optional anymore. As automation spreads across industries, understanding how AI works becomes a core economic skill. If policymakers underestimate that urgency, the competitiveness gap will widen fast. https://t.co/ncHCj37goe @epc_eu https://t.co/XtYStA1f1V
WE WON THE @MistralAI LONDON HACKATHON ๐ฌ๐ง๐ซ๐ท We made Mistralverse, here's our demo vid. @HarryStebbings who says the UK isn't shipping?? https://t.co/lVWr43XkNj
We won the @MistralAI London hackathon๐ฌ๐ง We built Mistralverse, a fully autonomous AI powered civilization, where every citizen is an agent with its own land, money and opinions. The agents even democratically elected a mayor to run the economy! Thatโs 2 AI lab hackathons won
NEW: When OpenAI announced its Pentagon deal Friday night, people immediately challenged Sam Altman's claims. Why, they asked, would the DoD suddenly agree to red lines when it had said it would never do so? The answer, sources told me, is that it didn't. https://t.co/DkF9uWVHa4
