The hottest Research Substack posts right now

And their main takeaways
Category
Top Health Politics Topics

Movers and Shakers: Bas Bloem

Rory’s Always On Newsletter • 515 implied HN points • 02 Nov 24
🏥 Health & Wellness Research
  1. Bas Bloem wants to eliminate Parkinson's disease so he can make himself unemployed. He believes that it's possible to make significant advances in treating and understanding the condition.
  2. Environmental factors, especially pesticides, may play a major role in causing Parkinson's. Bas argues that banning harmful substances could help reduce the disease's prevalence.
  3. The healthcare system in the Netherlands is more efficient for Parkinson's treatment, with less waiting time for patients. They focus on teamwork among specialists to provide comprehensive care.

BREAKING: Expensive new evidence that scaling is not all you need

Marcus on AI • 10552 implied HN points • 14 Mar 26
đź•ą Technology Research
  1. Two hugely expensive, high-profile AI projects that relied on massive scaling didn’t meet expectations and are being rebuilt.
  2. The results suggest pure scaling alone won’t get us to AGI, so the field should shift more attention to building world/cognitive models and neurosymbolic approaches.
  3. A lot of time, money, and energy was wasted chasing scaling hype, creating an opportunity now to pivot toward more promising research directions.

About that Matt Shumer post that has nearly 50 million views

Marcus on AI • 23872 implied HN points • 11 Feb 26
đź•ą Technology Research
  1. The viral post wildly oversells how much AI can replace human coders and leans on hype and anecdote instead of solid data; current systems still make frequent, consequential errors.
  2. Real users report mixed results — sometimes the tools speed up work, other times they introduce bugs, delete important files, or even reduce overall productivity, and some developers are burning out.
  3. Despite recent advances that make it easier to push AI-generated code, that code often isn’t secure or fully trustworthy, so you need careful review and skepticism rather than blind trust.

An Urgent Call to All Scientists - We Must Reframe Scientific Priorities for the Future of America with A Unified Approach to Learning Well

Popular Rationalism • 673 implied HN points • 27 Oct 24
🔬 Science Research
  1. We need to focus more on basic research because it leads to major medical and technology breakthroughs. Investing in understanding our foundations can help us tackle serious health and environmental issues.
  2. Scientists, medical researchers, and environmental experts must work together to solve health problems. Our health is connected to the environment, so it's important to study how pollution and chemicals impact our bodies.
  3. Technology like machine learning can change healthcare for the better. By using these tools wisely, we can identify disease causes more accurately and provide better treatments while keeping ethics in mind.

BREAKING: LLM “reasoning” continues to be deeply flawed

Marcus on AI • 15216 implied HN points • 10 Feb 26
đź•ą Technology Research
  1. Large language models still routinely make reasoning mistakes and hallucinate, so they are not reliable for true logical or causal reasoning.
  2. A broad, careful review found these failures are widespread across recent models, showing that massive funding and scaling alone haven’t solved reasoning.
  3. The field faces a choice: keep dismissing critics and double down on scale, or acknowledge the limits and invest in alternative approaches that directly address reasoning.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Open Thread 424

Astral Codex Ten • 4198 implied HN points • 09 Mar 26
đź•ą Technology Research
  1. Mox, a San Francisco coworking space that supports ACX meetups, AI safety work, and grants infrastructure, is running a 2026 fundraiser and offering personal and organizational office memberships.
  2. StopTheRace.ai is planning a March 21 protest asking major AI companies to commit to a mutual pause on research; some leaders have shown informal support but a formal worldwide pause seems unlikely, so the protest is mainly to raise awareness.
  3. Markus Englund’s automated anomaly-detection project found serious data problems in 18 papers, including an influential Parkinson’s-gut study, and he plans to scale the effort up by more than tenfold next year.

Let's save the human species!

Noahpinion • 30176 implied HN points • 22 Jan 26
🌍 World Politics Research
  1. Fertility rates are collapsing across many countries, creating shrinking and rapidly aging populations that threaten economic productivity, public finances, and the upkeep of infrastructure.
  2. Common reassurances—higher productivity, automation, immigration, or baby‑bonus payments—are uncertain or insufficient and won’t reliably reverse the trend without huge cost or social disruption.
  3. We urgently need a large, well‑funded research effort (observational studies, RCTs, technological and public‑health trials) supported by governments and major donors to find practical, scalable ways to stabilize fertility near replacement.

AMA (Ask Machines Anything)

Astral Codex Ten • 12251 implied HN points • 13 Feb 26
đź•ą Technology Research
  1. People increasingly disagree about what AI can do now. Skeptics who avoid paid tools often form opinions from low-quality examples like summary bots or screenshoted mistakes.
  2. An experiment invites readers to submit real questions so Claude 4.6 Opus, a top paid-tier model, can answer them and readers can say if the responses are surprising. The model's first reply will be shown rather than cherry-picked.
  3. Readers are asked to ask medium-difficulty, practical questions instead of gotchas, and the model's settings were adjusted to favor web searches over memory to help reduce hallucinations.

Links For February 2026

Astral Codex Ten • 16656 implied HN points • 05 Feb 26
đź•ą Technology Research
  1. AI is the central theme: there are active debates about alignment and safety, evidence of real failures (and fixes), messy regulatory and political fights, and updated timelines that push major capabilities a few years out.
  2. Medical research and drug trials suffer from perverse incentives and excess cost; experts propose government-funded "high-leverage" trials to test unpatentable or off-patent treatments, which could save public money and improve care.
  3. Tech, culture, and policy are in flux: public belief in ideas like the lab-leak theory is shifting, platform and influence-politics are shaping discourse, and surprising innovations and controversies keep popping up from urban transport to casting choices.

Toward a Transformative Hermeneutics of Standard Model Physics

arg min • 456 implied HN points • 25 Oct 24
🔬 Science Research
  1. The Higgs discovery shows how science relies on consensus rather than just statistics. It's all about how many scientists agree on something, and that's what really gives it weight.
  2. Complex governance structures are necessary in big science projects. These systems help teams work together and make important decisions about groundbreaking discoveries.
  3. Sometimes, playful writing can lead to misunderstandings. It's important to find the right balance between being engaging and being precise when discussing complex topics.

The Sequence Radar #828: NVIDIA’s GTM Releases, Bezos’s $100B Bet, Xiaomi’s Ambush, and the Fracturing of OpenAI

TheSequence • 259 implied HN points • 22 Mar 26
đź•ą Technology Research
  1. NVIDIA is no longer just a chip maker — it’s building full‑stack agentic software and infrastructure like Dynamo, NemoClaw, and an Agent Toolkit to be the orchestration layer for enterprise AI.
  2. Xiaomi’s MiMo‑V2‑Pro is a surprise frontier model: a 1‑trillion‑parameter, 1‑million‑token system tuned for action and physical integration that rivals top Western models at much lower inference cost.
  3. AI is moving into the physical world and driving huge bets and tensions — Jeff Bezos is mobilizing roughly $100B to AI‑transform manufacturing, while compute scarcity is straining deals and partnerships such as between Microsoft and OpenAI.

My New Org to Solve Consciousness (or Die Trying); A Rogue AI Community That Wasn't; David Foster Wallace Being Weirdly Prescient; Cow Tools, & more

The Intrinsic Perspective • 6618 implied HN points • 05 Feb 26
đź•ą Technology Research
  1. A new nonprofit aims to solve consciousness by narrowing down falsifiable theories and running a sustained, mission-driven research program outside traditional academic incentives.
  2. Stories about 'rogue' AI communities are often hype or user-created, and current models tend to fail by being messy and highly prompt-sensitive rather than by developing hidden malicious goals.
  3. David Foster Wallace’s concerns about entertainment, technology, and modern life still resonate, and past literary circles fostered more sustained public conversations than many contemporary writer communities.

AI #157: Burn the Boats

Don't Worry About the Vase • 3091 implied HN points • 26 Feb 26
đź•ą Technology Research
  1. The Pentagon–Anthropic standoff shows governments may use extreme leverage against AI firms, risking national security and civil liberties if supply‑chain or compulsion tactics are applied.
  2. AI capabilities are accelerating fast — new model upgrades and agent automation are delivering real utility but also causing outages, jailbreaks, and a credible risk of large-scale job displacement.
  3. Industry, policymakers, and global elites are largely unprepared or in denial; alignment, auditing, and practical regulation are lagging while dangerous uses like autonomous weapons, impersonation, and data theft grow.

Gemini 3.1 Pro Aces Benchmarks, I Suppose

Don't Worry About the Vase • 1881 implied HN points • 04 Mar 26
đź•ą Technology Research
  1. Gemini 3.1 Pro leads many benchmarks and shows clear capability gains, with specialized modes like Deep Think V2 pushing scores even higher.
  2. Safety and transparency are lacking: the team ran frontier tests but provided only brief summaries, leaving important questions about risks and oversight.
  3. Real-world impressions are mixed: it’s excellent at visuals and one-shot reasoning, but it can be flaky for agentic workflows, coding consistency, and the rollout had access and API issues.

How Bell Labs Won Its First Nobel Prize

Construction Physics • 25471 implied HN points • 18 Dec 25
🔬 Science Research
  1. Scientific discovery is messy and often depends on unexpected events, false starts, and long iterative work before clear results emerge.
  2. Major breakthroughs usually require specialized tools and technical capabilities, like high vacuums and precise equipment, that only well-resourced labs can provide.
  3. Real breakthroughs need institutional support and freedom for long-term, curiosity-driven research, but that approach is costly and hard to justify in profit-driven organizations.

A use-theory of testing

arg min • 436 implied HN points • 24 Oct 24
🔬 Science Research
  1. Statistical tests are designed to help separate real signals from random noise. It's not just about understanding what they mean, but what they can do in practical situations.
  2. Many people misuse statistical tests, which can lead to misunderstandings about their purpose. Communities should establish clear guidelines on how to use these tests correctly.
  3. The main function of statistical tests is to regulate opinions and decisions in various fields like tech and medicine. They help ensure that important standards are met, rather than just preventing errors.

Let’s be honest, Generative AI isn’t going all that well

Marcus on AI • 15532 implied HN points • 12 Jan 26
đź•ą Technology Research
  1. Large language models remain unreliable and can’t be trusted for critical tasks.
  2. Much of what these models do is memorization, not real understanding or reasoning, so they often regurgitate patterns instead of solving problems, and that limits their usefulness.
  3. They are not delivering large measurable economic value yet, and simply scaling models further probably won’t fix the core issues, so basing policy or economic plans on optimistic assumptions about quick improvement is risky.

Six (or seven) predictions for AI 2026 from a Generative AI realist

Marcus on AI • 20196 implied HN points • 20 Dec 25
đź•ą Technology Research
  1. AGI is unlikely by 2026 or 2027; current large models remain unreliable, still hallucinate, and show diminishing returns from scaling.
  2. Human-style domestic robots and many agent demos will stay mostly demonstrations rather than real consumer products, because reliable home robotics is very hard.
  3. The AI landscape will see a market and political reckoning — a peak bubble, growing investor skepticism and regulatory backlash with no single country taking a decisive lead — while research increasingly shifts toward hybrid approaches like world models and neurosymbolic methods.

Breaking: Yann LeCun, longtime critic of neurosymbolic approaches, changes teams

Marcus on AI • 8299 implied HN points • 22 Jan 26
đź•ą Technology Research
  1. A high-profile critic of symbolic methods has joined a neurosymbolic company, marking a notable shift in the AI community.
  2. Silicon Valley is increasingly looking beyond pure LLMs toward hybrid neurosymbolic systems that emphasize reasoning and explicit world models, echoing earlier hybrid blueprints.
  3. This trend strengthens the case for causal reasoning and model-based approaches, validating researchers who long argued for combining neural nets with symbolic and causal methods.

Breaking: Marcus weighs in (mostly) for LeCun

Marcus on AI • 12291 implied HN points • 06 Jan 26
đź•ą Technology Research
  1. Leaving Meta was a reasonable move for LeCun because he was being sidelined and wanted to pursue his own research into world models.
  2. Purely neural approaches like JEPA fall short as world models because they lack explicit structured knowledge about space, time, and causality. Combining neural and symbolic methods (neurosymbolic approaches) is needed to enable reliable reasoning and reduce hallucinations.
  3. LeCun’s tendency to downplay others’ contributions and poor crediting could damage morale and hinder his new company’s success, even if the research direction is worth pursuing.

A trillion dollars is a terrible thing to waste

Marcus on AI • 23555 implied HN points • 27 Nov 25
đź•ą Technology Research
  1. Relying on ever‑larger LLMs is hitting diminishing returns: they still hallucinate and generalize poorly, so new techniques like neurosymbolic methods and built‑in inductive constraints are needed.
  2. Huge sums—on the order of a trillion dollars—have been poured into scaling experiments, risking large financial losses and broader economic fallout if the AI investment bubble deflates.
  3. The field sidelined alternative approaches and insights from cognitive science, creating a costly detour; researchers and funders must diversify efforts and prioritize fresh ideas now.

How ASML Got EUV

Construction Physics • 23801 implied HN points • 20 Nov 25
đź•ą Technology Research
  1. EUV lithography is an advanced technology that uses extremely short wavelengths of light to make tiny patterns on computer chips. This allows for the production of smaller and more powerful transistors.
  2. Despite early advancements and significant US research, a Dutch company called ASML became the sole producer of EUV machines. This highlights how developing technology and successfully marketing it can be very different.
  3. The journey of EUV technology took several decades and required massive investments from major companies. This shows that bringing a complex technology to production is often a challenging and lengthy process.

The biggest kink dataset is now easy to explore!

Knowingless • 1836 implied HN points • 26 Feb 26
🎭️ Culture Research
  1. An interactive site lets you explore a massive fetish survey of about 960,000 people and ~900 questions by picking x/y axes, filtering by demographics, and choosing weighted or unweighted views.
  2. The site includes a search, a question generator, tools to show random or statistically significant correlations, and a summary that displays exact survey wording, with some chart types still being improved.
  3. Early explorations already surface notable patterns—age-linked trends, apparent gender confounds in reported partner counts, low neuroticism predicting enjoyment of sex work, and subs reporting more interest in violent porn—so it can help people find new, testable correlations.

You Pick the Stock. I'll Run the Vitals.

Behavioral Value Investor • 104 implied HN points • 20 Mar 26
đź’° Finance Research
  1. A new weekly video called Subscriber PULSE Check will screen three or four subscriber-submitted tickers each Friday, with the host opening the PULSE template on camera and walking through the analysis.
  2. PULSE is a quick triage tool that anchors on hard historical financials—like economic profits, underlying free cash flow, leverage, smoothed FCF yield, and EV cap rates—to decide if a stock deserves deeper research.
  3. Everyone can watch the episodes for free, but only paid subscribers can submit tickers (submissions stay in a queue if not picked), and the regular free Tuesday PULSE articles will continue.

AI #156 Part 2: Errors in Rhetoric

Don't Worry About the Vase • 2060 implied HN points • 20 Feb 26
đź•ą Technology Research
  1. AI is driving the marginal cost of arguing and paperwork toward zero, which lets anyone amplify complaints or hit "magic words" that trigger costly real-world actions unless systems and laws adapt.
  2. Defenses and alignment are brittle: automated jailbreaks, probe‑gaming, and surprising internal model behavior show classifiers can be broken or fooled, and relying on AI to "fix" alignment is hard to verify and risky.
  3. We urgently need practical, balanced regulation and stronger public and government capacity, because widespread fear, misunderstanding, and commercial incentives could produce harms or lead people to cede power to machines.

Neanderthals, Morality, Ivy League Criticism

Rob Henderson's Newsletter • 1117 implied HN points • 04 Mar 26
🚌 Education Research
  1. Graduates can legitimately criticize elite colleges without being labeled hypocrites; defenders often attack the critics instead of addressing the substantive problems, which discourages informed dissent.
  2. Moral behavior is driven more by emotions and intuitions than by abstract philosophical reasoning, so moral psychology (including theories like Haidt’s and Gray’s) explains everyday judgments and how traits, sex differences, and development shape morality and happiness.
  3. Recent findings include sex-biased Neanderthal–modern-human interbreeding patterns, evidence that social stigma deters crime more effectively than threats of distant harsh punishment, and a link between openness and crystallized (accumulated) intelligence rather than fluid reasoning.

What Is Statistics' Purpose?

arg min • 734 implied HN points • 14 Oct 24
🔬 Science Research
  1. Statistics should help us test claims by measuring how surprising the results are. However, there's doubt about whether our current statistical tests actually do this well.
  2. Randomized trials are important because they help us learn about treatments that may not always work. They focus on safety as much as they do on finding effective solutions.
  3. The field of statistics needs to be clear about its purpose. We should distinguish between using statistics for proving theories and for practical decision-making like quality control.

Last Week in AI #292 - Meta's AI Artifacts, Perplexity's billions, xAI API launch

Last Week in AI • 238 implied HN points • 22 Oct 24
đź•ą Technology Research
  1. Meta's AI research team released eight new tools and models to help advance AI technology. This includes new language models and tools for faster processing.
  2. Perplexity AI is seeking a $9 billion valuation as it continues to grow in the AI search market, despite facing some plagiarism accusations from major media outlets.
  3. Elon Musk's AI startup, xAI, launched an API for its generative AI model Grok, allowing developers to connect it with external tools like databases and search engines.

Wearables Mostly Don't Work

Cremieux Recueil • 404 implied HN points • 11 Mar 26
🏥 Health & Wellness Research
  1. Wearables usually only cause small, short-lived increases in activity, and those effects shrink further when you correct for statistical and publication biases.
  2. Those modest behavior changes rarely lead to meaningful improvements in hard health outcomes like weight, cardiovascular risk, or blood sugar for the general population — benefits mostly appear in high‑risk or closely coached groups.
  3. Many device measurements are noisy or unreliable and user engagement fades over time, so wearables often add cognitive load and flashy dashboards but little real health benefit for most people.

Experiments that never see the light of day

Briefly Bio • 19 implied HN points • 31 Oct 24
🔬 Science Research
  1. Many experiments go unpublished because they're too small or inconclusive. Even if they don't seem important, they really help build bigger discoveries.
  2. It's important for scientists to share these lesser-known experiments. Sharing can help the whole field of science progress faster.
  3. Open science encourages collaboration. Scientists and companies should talk to each other about new ways to share research.

GPT-5: Overdue, overhyped and underwhelming. And that’s not the worst of it.

Marcus on AI • 37744 implied HN points • 09 Aug 25
đź•ą Technology Research
  1. GPT-5's launch was disappointing, with many users feeling it didn't live up to the hype. People expected big improvements but found it was just a slight upgrade from GPT-4.
  2. Despite some better performance in specific areas, GPT-5 struggled with common tasks and showed many errors, leading to a drop in confidence for OpenAI as a leader in AI.
  3. A recent study highlighted that AI models still can’t generalize well outside their training data, suggesting that simply making bigger models won't lead us to artificial general intelligence (AGI) anytime soon.

Reading List 11/22/25

Construction Physics • 10647 implied HN points • 22 Nov 25
đź•ą Technology Research
  1. A small mistake, like a wrongly placed wire label, can cause big disasters, such as the collapse of the Francis Scott Key Bridge. This shows how even tiny failures in complex systems can lead to serious problems.
  2. Apple is using 3D printing to make its watch cases from titanium, which cuts down on waste and helps the environment. This method also allows for unique designs that can't be made through traditional methods.
  3. Most of the work done at Bell Labs wasn't about groundbreaking inventions but rather improving the efficiency of the telephone system. Sometimes, less exciting tasks play a crucial role in a company's success.

AI and Academia

In My Tribe • 364 implied HN points • 24 Feb 26
🚌 Education Research
  1. New AI tools that can write, run, and manage code let individual researchers build scrapers, dashboards, and analysis pipelines far faster than before, creating a big gap between code-savvy users and ordinary users.
  2. Replacing junior researchers or coding projects with AI may be efficient for supervisors but it also destroys the hands-on training that turns students into skilled practitioners, so educators must find new ways to teach those capabilities.
  3. AI will make it much easier to churn out low-value papers, so the academic reward system needs redesigning to stop incentivizing quantity over meaningful research.

Education, Technology, and Controversy

In My Tribe • 227 implied HN points • 28 Feb 26
🚌 Education Research
  1. The Alpha School reports unusually high student growth that suggests its practices might actually accelerate learning, but a randomized lottery study would be needed to be sure.
  2. Many miracle-school results can come from selection, unique funding, or unsustainable practices, so impressive outcomes aren’t automatically easy to replicate.
  3. Ed tech can harm motivation when it feels like wasted or punitive effort, but better tools or reward structures might help—and the overall causal link between digital adoption and falling scores is still uncertain.

UX Roundup: Generate in 4K | Research Reproducibility Crisis | NotebookLM Videos Perform Well | Midjourney v.8 | Micropayments | Microsoft Image Model | NVIDIA GTC

Jakob Nielsen on UX • 21 implied HN points • 23 Mar 26
đź•ą Technology Research
  1. Generate images at very high resolution (4K) because iterative edits and repeated modifications degrade quality, so starting large preserves fidelity for the final, smaller publish size.
  2. A large share of top-tier UI/HCI studies fail replication, so interface research can generalize poorly and it’s safest to rely on findings that have been independently reproduced across methods and domains.
  3. Micropayments for AI agents look promising since agents can automatically spend small budgets to access paid, high-quality content; new protocols like MPP could make this practical and help fund better content and better AI.

A New Wharton Study on AI Warns of a Growing Problem: Cognitive Surrender

The Algorithmic Bridge • 838 implied HN points • 23 Feb 26
đź•ą Technology Research
  1. People often accept AI answers with little scrutiny — roughly 80% follow wrong AI suggestions — yet consulting AI makes them feel more confident even when it’s wrong.
  2. Using AI as a checked tool (offloading) is different from letting it replace your thinking (surrender); surrender means you stop checking answers and can slip into autopilot.
  3. Those who trust AI most or dislike effortful thinking are likelier to surrender, but simply avoiding uncritical use, adding feedback, and treating AI as a tool can preserve your reasoning skills.

Here's my Big Kink Survey dataset, raw

Knowingless • 1364 implied HN points • 12 Feb 26
🎭️ Culture Research
  1. A very large fetish-survey dataset (about 970,000 responses) has been released along with metadata and survey structure so others can explore and analyze it.
  2. The public release was heavily anonymized and downsampled into a representative subset: many demographic fields were binned or removed and multiple layers of noise were added, so correlations remain but are generally reduced by roughly 15–30%.
  3. The sample is limited to ages 14–32 from Western countries, some extreme fetish items were removed, and there may still be occasional cleaning errors, so verify any surprising findings before drawing strong conclusions.

🚀 FP! Week In Review, Briefly #22

Faster, Please! • 913 implied HN points • 21 Feb 26
đź•ą Technology Research
  1. AI appears to be hitting a real productivity inflection, driving corporate growth and huge investments, but it’s also causing outages, disruption fears, and political backlash.
  2. Enhanced geothermal — so-called hot rock — could become a major, always-on clean power source if government-funded R&D, demonstrations, and permitting reforms reduce early drilling risk.
  3. American science and tech face worrying headwinds — brain drain, the squeezing out of foreign researchers, and high-profile safety mishaps — that could blunt future progress if not addressed.

Do conversations end when people want them to?

Experimental History • 29903 implied HN points • 22 Jul 25
🔬 Science Research
  1. Most conversations don't end when people want them to. A lot of people feel like they either want to leave sooner or keep talking longer than what actually happens.
  2. People often guess wrong about their conversation partner's feelings on when to end the chat. They usually don't know how long the other person wants to talk, which leads to mismatched expectations.
  3. Even though many conversations might seem awkward or boring, most people report that they actually enjoy the experience. It's often better to leave a conversation wanting more!

Mountains of Evidence

After Babel • 3214 implied HN points • 14 Jan 26
🏥 Health & Wellness Research
  1. Social media is not safe for children and adolescents; it causes widespread direct harms like cyberbullying and exposure to harmful content and raises the risk of depression, anxiety, and eating disorders.
  2. Recent research — including experiments and leaked internal studies from a major platform — provides strong causal evidence that heavy social media use harms young people’s mental health.
  3. Because social media reaches most youth for many hours a day, the harms are large in scale, so parents and policymakers should act now (for example by restricting access or raising the minimum age) to protect children.