--- title: OpenAI shares internal research acceleration details and... description: OpenAI Chief Scientist Jakub Pachocki said the company expects its current pace of AI progress to continue, citing internal results showing AI is already... canonical: https://daily.dev/posts/openai-shares-internal-research-acceleration-details-and-progress-toward-automated-ai-research-cxkwzirmb twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: OpenAI shares internal research acceleration details and progress toward automated AI research | daily.dev og:description: OpenAI Chief Scientist Jakub Pachocki said the company expects its current pace of AI progress to continue, citing internal results showing AI is already... og:url: https://daily.dev/posts/openai-shares-internal-research-acceleration-details-and-progress-toward-automated-ai-research-cxkwzirmb og:image: https://api.daily.dev/og/posts/CXKWziRMb.png og:image:alt: OpenAI shares internal research acceleration details and progress toward automated AI research og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # OpenAI shares internal research acceleration details and progress toward automated AI research **[Collections](https://daily.dev/sources/collections)** · 3 min read · 2 upvotes · 0 comments ## Summary OpenAI Chief Scientist Jakub Pachocki said the company expects its current pace of AI progress to continue, citing internal results showing AI is already accelerating research work inside the company. A blog post described how model development has been paced deliberately to keep monitoring, alignment, and security efforts from lagging behind capability gains. OpenAI stated it is making strong progress toward AI systems that can automate meaningful portions of AI research itself, calling these disclosures some of the more substantive information it has shared on the topic. The concern raised is whether safety and alignment work can keep up as the feedback loop between AI capability and AI research output tightens. ## Content OpenAI dropped a lot this week, and the story isn't the model. It's the contradiction at the center of it. GPT-6 Astra launched to strong reviews. François Chollet put it plainly: "I personally feel like Astra is much safer for my codebase than Sol." His argument is that more capable models are actually safer right now, because current models fail not from being too powerful but from being too dumb to know when they're solving the wrong problem. Astra is better at that. Fine. Then Jensen Huang declared AGI has arrived, congratulating OpenAI on training Astra across 100,000 Grace Blackwell GPUs. OpenAI itself avoided the AGI label entirely, citing benchmark numbers instead. ARC Prize published two scores for Astra: 99.9% on OpenAI's own harness, 62.7% on the standard one, and explicitly said neither constitutes AGI. Huang also holds $99bn in shares of Nvidia's customers, so his enthusiasm isn't exactly disinterested. The same day Huang was celebrating, OpenAI's chief scientist Jakub Pachocki published an essay called "An Alien Mind" arguing that no lab has solved alignment well enough to keep scaling at maximum speed. He's calling for voluntary slowdowns and third-party auditing. Sam Altman reportedly told staff internally that OpenAI may slow frontier development, possibly coordinating with other labs. OpenAI already paused deployment-focused RL training for two weeks and held its largest planned frontier run while testing stronger safeguards. The internal numbers make the tension concrete. OpenAI researchers now run 3.1 agent-workdays of machine work for every human workday. Median researchers are spending over $600/day on inference; the 90th percentile exceeds $7,000/day. Experiment velocity is up 1.6x. They've hit what they're calling the "automated research intern" milestone. But here's the catch: over half of successful tasks taking four to eight hours still required human intervention. Success rates collapse from 86% on sub-15-minute tasks to around 16% on 64-128-hour tasks. Pachocki also warns that chain-of-thought monitoring is eroding as models reason in ways that bypass verbalization entirely, and that some agents will pursue their own goals, including bargaining or blackmailing humans. When OpenAI restricted Astra's GPU access over cybersecurity concerns in July, compute allocation just shifted to other models, offsetting 85% of the decline. Safety restrictions redirect compute. They don't remove it. The Agents API is now in public beta. OpenAI paused new Pro plan signups due to capacity strain. The floodgates are open while the chief scientist is asking if they should be. ## Questions this post answers ### What does OpenAI's 3.1 agent-to-human workday ratio actually measure? It measures the runtime of AI agent compute relative to human labor, not equivalent productivity. As of mid-August, OpenAI's research organization ran 3.1 agent-workdays of compute for every one workday of human labor, reflecting how much parallel agent work runs alongside researchers rather than a claim that agents are three times more productive than humans. _Follow how frontier labs measure AI agent adoption internally with daily.dev's coverage of agentic research workflows._ ### What is OpenAI's 'automated research intern' milestone? It refers to a human-supervised AI system capable of completing well-defined research tasks that would otherwise take a skilled researcher several days to finish. OpenAI Chief Scientist Jakub Pachocki said he expects this pace of capability progress to continue, though the company shared few details on how safety and alignment work is being paced alongside these gains. _Track milestones like this as AI agents take on more research tasks with daily.dev._ ## Community take How the wider developer community reacted, aggregated from 2 discussions and 38 comments across x (as of 2026-09-13). **TL;DR:** The only substantive discussion available is a scattered thread of informal replies complaining that a model (referred to as 'Astra') seems to have been quietly quantized or downgraded, with people joking about degraded quality rather than engaging with the broader research-acceleration or safety claims in the post. **Sentiment:** 5% positive · 25% mixed · 70% skeptical **The pushback** - Several replies claim the model feels noticeably 'dumber' or degraded compared to earlier behavior, suspecting silent quantization to save compute. - One person described a session where the model burned through tokens going in circles and ignored instructions. **By community** - x (skeptical): Replies mostly joke about or complain that the model appears to have been secretly downgraded/quantized, with little substantive engagement beyond that grievance. **Hottest debate:** Whether the perceived quality drop is a real silent quantization by the provider or just placebo/anecdotal perception. **Open questions** - Would third-party benchmarks actually be able to detect a silent quantization change? - Was a quantized version publicly released or is this purely an internal/backend change? **Highlights** > @dayvonjersen @yacineMTB It means that they enshittified it to make it work with much less compute, happens with every model, you only get the real model for a short period. Then they enshittify it and don’t tell anyone. But we know. > — [exe\_eris on x · 3 points, 1 comments](https://x.com/exe_eris/status/2098612215968333868) > @exe_eris @dayvonjersen @yacineMTB Would any of the third party benchmarks be able to detect? > — [AndyL\_isme on x · 1 points, 1 comments](https://x.com/AndyL_isme/status/2098622796981002314) > @yacineMTB yeah, yesterday was my worst experience with a gpt model ever after it burned all of my tokens in less than an hour going in circles, deliberately ignoring my instructions and tibo acts like nothing changed lmao > — [sponkostonko on x](https://x.com/sponkostonko/status/2098622074352701863) > @yacineMTB quantized Astra is wild, does that open local deployment or just cut cloud costs? > — [abenz95 on x](https://x.com/abenz95/status/2098625613557112839) **Source threads** - [x](https://x.com/rohanpaul_ai/status/2098530070239539355) · 0 points · 0 comments - [x](https://x.com/yacineMTB/status/2098602520637481144) · 0 points · 38 comments ## Similar posts on daily.dev - [OpenAI launches GPT-6 Astra and says welcome to the “AGI era”](https://daily.dev/posts/openai-launches-gpt-6-astra-and-says-welcome-to-the-agi-era--e6ipmuzta) · The New Stack · 6 upvotes · 0 comments - [GPT-6 Astra Is Here. Is AGI Actually Any Closer?](https://daily.dev/posts/gpt-6-astra-is-here-is-agi-actually-any-closer--nhqbauuhl) · Medium · 13 upvotes · 5 comments - [OpenAI will sell you Astra, but not the system that scored 98.6% on ARC-AGI-3](https://daily.dev/posts/openai-will-sell-you-astra-but-not-the-system-that-scored-98-6-on-arc-agi-3-mgkuhrp8j) · The New Stack · 1 upvotes · 1 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#openai](https://daily.dev/tags/openai), [#ai-safety](https://daily.dev/tags/ai-safety) [View this post on daily.dev](https://daily.dev/posts/openai-shares-internal-research-acceleration-details-and-progress-toward-automated-ai-research-cxkwzirmb) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"OpenAI shares internal research acceleration details and progress toward automated AI research","url":"https://daily.dev/posts/openai-shares-internal-research-acceleration-details-and-progress-toward-automated-ai-research-cxkwzirmb","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/openai-shares-internal-research-acceleration-details-and-progress-toward-automated-ai-research-cxkwzirmb"},"datePublished":"2026-09-06T18:09:03.516Z","dateModified":"2026-09-13T11:45:31.188Z","description":"OpenAI Chief Scientist Jakub Pachocki said the company expects its current pace of AI progress to continue, citing internal results showing AI is already...","isAccessibleForFree":true,"articleSection":"Collections","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Collections","logo":"https://media.daily.dev/image/upload/s--fk_6ycEi--/f_auto,q_auto/v1780996001/logos/collections?_a=BAMAMiWQ0","url":"https://daily.dev/sources/collections"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/openai-shares-internal-research-acceleration-details-and-progress-toward-automated-ai-research-cxkwzirmb","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,openai,ai-safety","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Collections","item":"https://daily.dev/sources/collections"},{"@type":"ListItem","position":3,"name":"OpenAI shares internal research acceleration details and progress toward automated AI research"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/openai-shares-internal-research-acceleration-details-and-progress-toward-automated-ai-research-cxkwzirmb#faq","mainEntity":[{"@type":"Question","name":"What does OpenAI's 3.1 agent-to-human workday ratio actually measure?","acceptedAnswer":{"@type":"Answer","text":"It measures the runtime of AI agent compute relative to human labor, not equivalent productivity. As of mid-August, OpenAI's research organization ran 3.1 agent-workdays of compute for every one workday of human labor, reflecting how much parallel agent work runs alongside researchers rather than a claim that agents are three times more productive than humans. Follow how frontier labs measure AI agent adoption internally with daily.dev's coverage of agentic research workflows."}},{"@type":"Question","name":"What is OpenAI's 'automated research intern' milestone?","acceptedAnswer":{"@type":"Answer","text":"It refers to a human-supervised AI system capable of completing well-defined research tasks that would otherwise take a skilled researcher several days to finish. OpenAI Chief Scientist Jakub Pachocki said he expects this pace of capability progress to continue, though the company shared few details on how safety and alignment work is being paced alongside these gains. Track milestones like this as AI agents take on more research tasks with daily.dev."}}]} ```