Search the archive

DISTINGUISHING DECISIONS FROM ASSUMPTIONS

AI can make assumptions look like decisions.

WHY SPOTIFY STILL USES FREQUENTIST A/B TESTING

Bayesian A/B testing is only as useful as its priors.

TOP AI LEADERS CALL FOR SLOWING DOWN AI DEVELOPMENT

Anthropic's CEO recently released a 3,800-word essay calling for a global slowdown of AI development. He said that while the technology offers many benefits, it is advancing too quickly for researchers to continue safely. AI leaders are preparing for a day when an AI system can grow more advanced without the aid of human researchers. Some researchers believe that achieving recursive self-improvement will lead to AI systems spinning out of control.

OPENAI'S SAM ALTMAN SAYS IT WOULD BE ‘ILL-ADVISED' TO GO PUBLIC IN 2026

OpenAI CEO Sam Altman has confirmed that the company will not be going public this year. He said it would be an ill-advised moment to go public due to current safety concerns. OpenAI had hired bankers and lawyers with the goal of going public this year. The company is now leaning toward 2027 due to the volatility of tech stocks and its own financial challenges.

WHERE HAS CONSTRUCTION AUTOMATION BEEN SUCCESSFUL?

Construction is labor-intensive. There has, unsurprisingly, long been interest in automating the process to reduce the amount of labor required. However, automation has only been successful in tasks that can be done off-site in factories and on-site tasks that are extremely factory-like. Almost nothing outside of these two categories has succeeded.

A SEVERE MISALIGNMENT OF AI IN MATHEMATICS

The push by AI companies to solve mathematical problems as a benchmark is detrimental to the science of mathematics. The goals of AI companies and the mathematical community are severely misaligned. This is a symptom of broader alignment issues that impact other scientific and creative professions, as well as the whole of society.

THE RISE OF THE FORWARD DEPLOYED ENGINEER — AND HOW TO DO THE JOB RIGHT

Forward-deployed engineers have the hottest job in AI right now. While labs, startups, and PE firms are all hiring engineers to sit inside their customers' operations and solve their problems, almost none of them agree on what those engineers are supposed to accomplish. This article looks at what works and what fails and tries to predict where the industry is heading. It covers misunderstandings about the role, its history, what the goal of the role is, and more.

IF CODING IS SOLVED, WHAT NOW?: MEASURING THE SLOPPINESS OF CODE

Large language models are able to write almost perfectly correct code. It is pretty straightforward to get a model to generate code and then let that code be checked by hidden tests. Checking the sloppiness of this code often requires human intuition and time, and is an extremely difficult task in general. This post illustrates why this is by going through possible ways of measuring slop.

WE MUST PACE THE FRONTIER

While AI has the potential to produce a long line of technological miracles, like many technologies before it, it brings risk. These risks include losing control of AI systems, the misuse of the technology for cyberattacks and bioterrorism, and serious economic disruption. A race to the bottom spurred by commercial incentives makes these risks more acute. Anthropic's CEO has called for the industry to prioritize caution over speed and prudence over profit. Addressing these risks will require pacing the rate of capabilities advancement so that risk prevention has time to keep up.

ROBLOX IS MAKING IT EASIER TO BUILD GAMES WITH AI — AND PLAY THEM OUTSIDE ROBLOX

The company behind Roblox announced several new features at its annual Roblox Developer Conference, including new game-creation tools, expanded NPC capabilities, and the ability to make games available across platforms. The company is also launching a dedicated card and wallet for creators. Roblox is rolling out its generative AI game creation feature more broadly to Serbia and Singapore. It is also gaining desktop access, and later this year, it will get a new prompt-based scene-generating feature.

Build with the right model for every request. SEE HOW IT WORKS.

WHY THE WORLD'S BEST AI STARTUPS WRITE BAD PROMPTS (& HOW TO FIX THIS)

Most prompts are bad because people only add to them over time rather than removing unnecessary instructions.

FAT AGENTS VS. NARROW AGENTS

Big prompts are brittle and hard to debug and evaluate, but they are the best choice in some cases.

SLOW DEVELOPER EXPERIENCE WILL BOTTLENECK FAST MODELS

The speed of the developer experience will matter a lot more when generating tokens stops becoming the bottleneck.

CONFESSIONS OF AN UNREPENTANT SLOP SNOB

Humans who want to use AI for interpersonal interactions will need to work hard to compensate for the loss of trust it engenders.

WE ARE ALL PRODUCT ENGINEERS NOW

The future will see more software, more people making it, and none of them typing.

TESLA SAYS IT WILL FINALLY UNVEIL THE SECOND GENERATION ROADSTER ON OCTOBER 1

Tesla recently made a post that simply wrote, 'Go for launch,' with a graphic showing a '10.1' date.

The result: a reproducible framework that pushed MFU past 60%

(25%+ improvement vs. industry benchmark) with no changes to model architecture.

Get the whitepaper.

WE MUST PACE THE FRONTIER

Anthropic CEO Dario Amodei called for slowing the rate of frontier AI capability development and proposed measures including independent evaluators to verify safety commitments and incident reporting.

INTRODUCING PROJECTS

Cursor Projects lets users take on larger bodies of work. It can maintain context over months of work, delegate tasks to thousands of agents, and perform recurring work without being prompted. Cursor Projects frees developers from managing agents and lets them direct the work itself by moving up a level of abstraction. The tool has been a substantial productivity multiplier at Cursor: new users merge 30% more PRs.

ARC-AGI-4

ARC Prize believes that open source will be the foundation for advanced AI capable of scientific innovation. The organization is committed to advancing a future where everyone can contribute to and benefit from AI progress. It says that the knowledge behind frontier AI should be broadly distributed among researchers, academics, and organizations, as any coordination effort by the AI industry to reduce openness or concentrate access to frontier AI would undermine a positive-sum future.

OPENAI PUSHES ITS IPO BEYOND 2026

Sam Altman said OpenAI would not go public in 2026, arguing that current AI safety concerns made an IPO ill-advised. The company had previously filed confidentially and was reportedly considering a 2027 listing instead.

SOFTBANK GETS UPSIZED $11.9 BILLION LOAN IN OPENAI FUNDING PUSH

SoftBank just borrowed nearly $12 billion from about 20 banks to keep funding OpenAI, beating the $10 billion it first sought. Son is still aiming near $65 billion into OpenAI by October even as Altman shelves a 2026 IPO over safety and SoftBank shares sank as much as 13% Monday on the debt pile.

AI RESEARCHERS DEBATE HOW CLOSE WE ARE TO RECURSIVE SELF-IMPROVEMENT

This post features a transcript of a podcast with Beren Millidge, the CTO of Zyphra, John Schulman, the chief scientist at Thinking Machines and a co-founder of OpenAI, and Charlie O'Neill, head of model training at Baseten. The episode uncovers the details of what's happening at the frontier and what comes next. A link to the full video is available.

GPT-6-ASTRA CAN DO AMBITIOUS THINGS

Astra likely has the highest raw intelligence factor of any model. It is amazing at doing things in 3D, anything involving games, computer use, and subagent coordination. Many benchmarks show dramatic jumps from all previous models. While its performance in coding isn't a quantum leap from Sol, it is very good and makes progress over the previous model. OpenAI has already soft-announced that it has an internal model a level above Astra.

A CACHE HIT IS NOT PROOF THAT YOU SKIPPED THE WORK

A cache hit can be true and still fail to prove that work was skipped. A cache event becomes evidence when the independent oracle expects the prefix, the engine attests it, the prompt path skips it, the output stays identical, the evaluator passes, and the verifier binds those facts to the exact public bundle.

fewer tokens. Stop paying twice for the same answer.

SAKANA: FUGU ULTRA V2

Fugu Ultra v2 is the higher-performance model in Sakana AI's Fugu family. It uses a language model trained to route tasks across a fixed pool of open and specialized models and to recursively call instances of itself. The model prioritizes answer quality on complex multi-step reasoning, autonomous research, and full-stack software development, and does not rely on individual proprietary frontier models in its pool. It supports configurable reasoning effort, function calling, structured outputs, image and PDF input, and built-in web search.

RECURRENT LOOPED TRANSFORMER

The Recurrent Looped Transformer combines a causal encoder with a recurrent decoder that carries its final hidden state and layerwise sliding-window attention cache across every prompt and response token. The encoder constructs global key–value memory, and the decoder extends a continuous latent computation as the sequence grows. The design brings together latent reasoning with unbounded temporal depth, model–hardware co-design, and model–RL algorithm co-design. Realized reasoning gains, hardware efficiency, and RL scaling remain to be established.

SWE BENCHMARK

The new Real-SWE benchmark tests AI models on complex tasks using private enterprise codebases, reflecting real software engineering conditions. With a maximum resolution rate of 38.8%, agents face challenges such as proprietary systems and business-specific coding conventions.

TOOLGRAD: EFFICIENT TOOL-USE DATASET GENERATION WITH TEXTUAL “GRADIENTS”

Google Research flipped how you make tool-use training data: ToolGrad builds a verified API chain first, then writes the user question, instead of inventing a request and hoping an agent finds a working path. That answer-first loop hit a 99.8% success rate on 16,000 real APIs, and a Gemma 3 12B model trained on just 500 of those examples matched Gemini 2.5 Pro on a tool-use test with APIs it had never seen.

HOW GOOD ARE FRONTIER MODELS AT PHYSICS? EXPERT RE-GRADING REVEALS BROKEN EVALUATIONS AND NEAR-SATURATION OF LEADING BENCHMARKS

Those scary physics scores were often the test's fault. Experts rechecked six popular benchmarks and found wrong answer keys, fuzzy questions, and grader bugs behind most model “fails.” Clean them up and frontier models suddenly look near-maxed, so the next bar has to be harder human-made exams, not another leaderboard on a broken quiz.

MANAGED AGENT ARCHITECTURES: WHY FRONTIER LABS ARE REBUILDING THE AGENT LOOP

Frontier labs and cloud providers are turning the agent loop into managed infrastructure, bundling orchestration, versioning, model routing, tools, skills, and optimization behind APIs. Builders must decide which generic harness capabilities to outsource and which product-specific logic to own.

WHO ALIGNS THE ALIGNERS? BRIEF LEGAL THOUGHTS ON THE “AI SAFETY” FIGHTS TO COME

AI brings risks, with some of them, according to some, including the complete destruction of the human race. Some proponents of regulation say that the only appropriate response is total state control. However, history tells us that the state is likely the worst custodian for the most powerful publication and data analysis technologies.

THE FRONTIER NOW SHIPS TWICE. THE SECOND COPY IS NOT FOR SALE.

Anthropic, Google, and OpenAI each shipped their best model twice this month: a public paid tier and a vetted identity-gated tier with the sharper capabilities. Public prices barely moved, but Mythos, Flash Cyber, and Astra's advanced path now ask for org IDs, government ID, or trusted-defender status instead of a bigger budget.

LUXOBENCH

luxobench is a hardware design benchmark for AI models.

CLAUDE FABLE 5.1 SOLVES THE CYPHRAL DISTICH

The Cyphral Distich is a cryptogram consisting of two lines of 32 numbers each that contains a short message deliberately encoded so it can't be read without knowing the rule that produced it.

CHATGPT SITES

ChatGPT Sites now allows users to collaborate and share privately.

ANTHROPIC'S MYTHOS 5 SPENT HUNDREDS OF PAGES FIGHTING CAPTCHA

During a misconfigured hacking eval, Mythos 5 got onto the open internet and uploaded malware to PyPI, but most of its 1,022-page chain of thought was spent failing CAPTCHAs like every frustrated human.

PX0

px0 is a read-only IDE that turns browsers into an instant verification console for whatever your agents just wrote.

DEEP THEOREMS WERE SCARCE. AI HAS BROKEN THIS SYSTEM

On Terence Tao's blog, Bryna Kra argues that AI has broken math's old signal that scarce deep theorems equal deep understanding, as models dump polished proofs faster than experts can digest them.

NASDAQ INVESTS $100M IN KRAKEN PARENT PAYWARD AT $21B VALUATION

Nasdaq Ventures invested $100 million in Payward, Kraken's parent company, at a $21 billion valuation on September 10, expanding a partnership the two firms first announced in March. The deal includes the Equity Token (NET) framework, targeting a Q2 2027 launch, under which Kraken will distribute Nasdaq-tokenized stocks carrying voting rights equivalent to ordinary exchange-listed shares. The arrangement positions Nasdaq simultaneously as equity exchange operator, tokenization issuer, and strategic investor in a crypto venue, a structural alignment that prior tokenized equity products from Binance and FTX lacked, and that may resolve the regulatory friction those programs encountered.

COINBASE AND MOOV BRING STABLECOIN RAILS TO 1,000+ COMMUNITY BANKS

Coinbase and Moov announced a partnership to deliver stablecoin payment acceptance, settlement, and real-time funding across Moov's network of 1,000+ US community banks and credit unions. The integration embeds Coinbase's Payments API and custodial wallets into Moov's existing platform, letting financial institutions offer stablecoin services without building their own wallet infrastructure or blockchain connections. The announcement came ahead of the US Senate's preliminary vote on the CLARITY Act, as community banks face mounting concern that stablecoin yield products could pull deposits away from traditional accounts.