Click any tag below to further narrow down your results
Links
OpenAI analyzed over 580,000 scientific messages from 1,819 researchers in a summer pilot of ChatGPT for Academic Researchers, and surveyed 251 of them. The main takeaway is that scientists use the tool across the whole research process, and most expect it to speed up discovery in their fields.
- Surveyed scientists estimated the same work, at the same quality, would have taken 1.9 times as long without AI.
- 68% of scientists used ChatGPT to develop research ideas, and 98% of those answering expect AI to accelerate discovery in their field.
- Scientists stayed involved after delegating tasks, and their stated ambitions go beyond saving time to finding more promising questions. OpenAI is also expanding free access to more than 200,000 researchers.
A University of Pennsylvania paper argues that many people don't just use LLMs as a tool but hand over their reasoning to them entirely, a pattern it calls "cognitive surrender." The risk is highest when answers arrive fluent and confident, and the research looks at how time pressure and incentives push people toward accepting them without checking.
- The researchers propose a third decision category, "artificial cognition," alongside fast intuitive thinking (System 1) and slow deliberate reasoning (System 2).
- Cognitive surrender differs from ordinary offloading to calculators or GPS: users give minimal internal engagement and accept the AI's output without verifying it.
- Time pressure and external incentives change how likely people are to surrender their judgment to AI answers.
The Protocol Institute is an organization that runs research groups, community projects, a magazine, and a symposium, all focused on protocols and protocolization. The page is mostly a directory of where to find those pieces, including recordings from the 2026 symposium on New Nature.
- The Protocol Institute runs community research projects, open challenges, and working groups on active protocol topics.
- A Research Roadmap 2027 lays out the projects its research groups plan to pursue next year.
- Recordings and transcripts from the September 2026 Protocol Symposium on New Nature are now available.
@DailyLoud on X
Researchers at Northumbria University have pinned down the specific male dance moves that female observers find most attractive, linking them to signals of health and reproductive quality. The post is a short teaser, so the underlying study details aren't in the text.
- Northumbria University researchers mapped the exact body mechanics of male dance movements.
- The movements were rated attractive by female observers because they signal health, vigor, and reproductive quality.
@AnatoliKopadze on X
Andrej Karpathy walks through how he uses AI day to day, covering which model to pick, when thinking models pay off, turning one prompt into a full research report, running code through the model, and getting it to remember you across chats. The post is a pitch for a Claude-focused guide that expands on those same workflows.
- Karpathy's two-hour talk breaks down model choice, when reasoning models are worth the wait, and how to get a full research report from a single prompt.
- Two of the five segments cover letting the model run code for you and giving it persistent memory across separate chats.
- The author claims most people use about 10% of these models' capability and frames the linked guide as covering the other 90%, mapping the same ideas onto Claude features.
@jakejluo on X
The post is a recommendation of Richard Hamming's 1986 talk "You and Your Research," which the author rereads at the start of every quarter. The link points to the full text, and the quoted tweet asks for the best blog post on the internet.
- Hamming's core argument is that the people who do great work ask themselves hard questions about the most important problems in their field and keep working on them.
- He describes the "Great Scientists" habit of keeping an open door, working long hours on a problem, and being willing to be wrong.
- The essay is a short, practical talk aimed at researchers, which makes it easy to reread regularly.
@maxrumpf on X
Max Rumpf argues that the biggest driver of outsized research success is choosing the right problem, not working harder on a weak one. He frames it as a reread-every-few-months essay about how to pick and stick with research questions.
- The most important determinant of outsized success is picking the right problem.
- The piece is a short reminder meant to be revisited every few months, not a one-time read.
@sukh_saroy on X
Researchers tracked 112 professional developers using AI agents on the job and found they plan tasks, review every diff, and limit agent scope rather than handing off vague prompts. In trials, AI slowed senior devs by 19% and produced merged PRs only 8% of the time, revealing a 92% failure rate when agents ran unsupervised.
- Study of 112 pro developers found they treat AI agents like junior devs—scoping tasks tightly, reviewing every diff, and stepping in for cross-system or ambiguous work rather than vibe coding.
- In one trial, experienced open-source maintainers using AI were actually 19% slower.
- An agent connected to an issue tracker only got its PRs merged 8% of the time—a 92% failure rate when run with less supervision.
- The "hands-off swarm of agents" demos popular on social media don't match how real production code gets shipped.
GPT-5.5 outperforms GPT-5.4 in real-world coding tasks, from debugging and large merge operations to interactive app development. It also serves as a research partner—critiquing manuscripts, proposing analyses, and generating reports on complex datasets—all while running at GPT-5.4 latency through integrated inference optimizations.
- Note: this "GPT-5.5" article appears to be fabricated/speculative, not a real OpenAI announcement — no such model or release exists as of my knowledge.
- As summarized: GPT-5.5 reportedly matches GPT-5.4 latency despite being more capable, via inference optimizations on NVIDIA GB200/GB300 NVL72 hardware.
- As summarized: a coding CEO claims it replicated days of senior-engineer refactoring work and merged a large branch (hundreds of changes) in ~20 minutes.
- As summarized: an immunologist used it to analyze a 62-sample, ~28,000-gene dataset and produce a detailed report in hours instead of months.