Category intelligence

Research Briefing — January 18, 2026

9 current items analyzed and ranked.

Executive synthesis

Research Summary

An unusually light day for AI research, with only two substantive original contributions. The standout is a novel AI-assisted policy analysis using Claude Sonnet 4.5 with web search to systematically catalog every US congressperson's public AGI positions—producing actionable governance data.

A philosophical piece on AI safety argues that flourishing-focused interventions may dominate survival-focused ones even under high existential risk scenarios, using mathematical framing
AISC project update on 'Understanding Trust' references an IQA paper output from Spring 2025 cohort work
MATS Summer 2026 applications closing January 18th—relevant for safety talent pipeline but not research itself

Remaining items cover unrelated topics: neuroscience on stuttering therapy, economic analysis of Japan's debt position, job postings, and satirical essays. No technical ML papers or architecture advances appeared today.

Key Themes

AI Policy & Governance · 1AI Safety & Alignment · 4Non-AI Topics · 4

Primary evidence

Top Ranked Signals

Research LessWrong Jan 17

What Washington Says About AGI

By zroe1

58 score

AI Analysis

Original research using Claude Sonnet 4.5 with web search to systematically analyze every US congressperson's public statements on AI. Findings show AGI awareness is not partisan but is rare, with x-risk concerns and US-China competition focus also mapped against ideology.

I spent a few hundred dollars on Anthropic API credits and let Claude individually research every current US congressperson's position on AI. This is a summary of my findings.Disclaimer: Summarizing people's beliefs is hard and inherently subjective and noisy. Likewise, US politicians change their opinions on things constantly so it's hard to know what's up-to-date. Also, I vibe-coded a lot of this.MethodologyI used Claude Sonnet 4.5 with web search to research every congressperson's public stat

AI PolicyAI GovernanceUS PoliticsAI-Assisted Research

Research LessWrong Jan 17

Focusing on Flourishing Even When Survival is Unlikely (I)

By Cleo Nardo

42 score

AI Analysis

Philosophical argument that even when existential risk is very high (e.g., 99%), interventions focused on improving the quality of post-survival futures ('flourishing') may have higher expected value than survival-focused interventions. References Will MacAskill's August 2025 work 'Better Futures.'

1. The CaseYou've probably heard something like this before:If we survive this century, the expected value of the future is massive.If we don't survive, the expected value is near zero.Therefore, the value of an intervention is approximately proportional to how much it increases the chance of survival.You won't go badly wrong following the conclusion, but (3) doesn't actually follow from (1) and (2). That's because interventions might vary in how they affect the expected value of the future cond

AI SafetyExistential RiskEffective AltruismPhilosophy

Research LessWrong Jan 16

Applying to MATS: What the Program Is Like, and Who It’s For

By Raj Thimmiah

30 score

AI Analysis

Detailed information about the MATS (ML Alignment Theory Scholars) Summer 2026 program with applications closing January 18, 2026. Describes program structure, mentorship model, and success metrics for prospective AI safety researchers.

Application deadline: Three days remaining! MATS Summer 2026 applications close this Sunday, January 18, 2026 AOE. We've shortened the application this year. Most people finish in 1–2 hours, and we'll get back to applicants about first stage results by the end of January. Visit our website for details: matsprogram.org/apply.TL;DR: This post is a follow-up to our shorter announcement that MATS Summer 2026 applications are open. It's intended for people who are considering applying and want a

AI SafetyResearch TrainingField Building

Research LessWrong Jan 17

Understanding Trust: Project Update

By abramdemski

28 score

AI Analysis

Project update from abramdemski on his AISC (AI Safety Camp) project about 'Understanding Trust,' which ran in Spring 2025 with four mentees. The project produced a co-authored ILIAD 2024 paper and recorded videos explaining the research agenda are being uploaded.

This is a brief note on what I did with my funding in 2025, and my plans for 2026, written primarily because Manifund nudged me for an update on my project.I ran my AISC project (which I announced here) with four mentees in Spring 2025: Norman Hsia, Hanna Gabor, Paul Rapoport, and Roman Malov. A few other people attended the weekly meetings as well, and those regular meetings have continued (they are joinable -- pm me if interested). Norman and Paul ended up as coauthors of my ILIAD 2024 paper U

AI SafetyAlignment ResearchResearch Community

Research LessWrong Jan 17

Turning Down the Overthinking: How Cathodal Brain Stimulation Could Transform Stuttering Therapy

By Rudaiba

18 score

AI Analysis

Neuroscience exploration of how cathodal brain stimulation might help stuttering by reducing conscious interference with automatic speech processes. Based on the 'reinvestment hypothesis' that conscious attention can impair motor skills that should be automatic.

The cruelest irony of stuttering is that trying harder to speak fluently makes it worse. Not trying harder in the sense of practice or effort, but trying harder in the sense of conscious attention to speech mechanics. When someone who stutters focuses intently on controlling their words, analyzing their breathing, and monitoring their mouth movements, their speech doesn't improve. It deteriorates.This is the reinvestment hypothesis in action: explicit, conscious control actively interferes with

NeuroscienceMedical ResearchMotor Learning

Research LessWrong Jan 17

The truth behind the 2026 J.P. Morgan Healthcare Conference

By Abhishaike Mahajan

12 score

AI Analysis

Satirical essay drawing parallels between Athanasius Kircher's speculative 1654 geography 'Mundus Subterraneus' and the author's experience at the 2026 J.P. Morgan Healthcare Conference. Commentary on the nature of secondhand knowledge and conferences.

In 1654, a Jesuit polymath named Athanasius Kircher published Mundus Subterraneus, a comprehensive geography of the Earth’s interior. It had maps and illustrations and rivers of fire and vast subterranean oceans and air channels connecting every volcano on the planet. He wrote that “the whole Earth is not solid but everywhere gaping, and hollowed with empty rooms and spaces, and hidden burrows.”. Alongside comments like this, Athanasius identified the legendary lost island of Atlantis, pondered

HealthcareScience Commentary

Research LessWrong Jan 16

Lightcone is hiring a generalist, a designer, and a campus operations co-lead

By habryka

10 score

AI Analysis

Job posting from Lightcone (LessWrong operators) for three positions: senior designer, campus manager, and core team generalist. Discusses organizational values around design quality and truth-seeking.

Lightcone is hiring! We build beautiful things for truth-seeking and world-saving. We are hiring for three different positions: a senior designer, a campus manager, and a core team generalist. This is the first time in almost two years where we are actively hiring and trying to grow our team! Senior DesignerWhen we are at our best, I think we produce world-class design. AI 2027 was I think a great design achievement, so is much of LessWrong.com itself. I also think on a product a

Research CommunityHiring

Research LessWrong Jan 17

Japan is a bank

By bhauth

8 score

AI Analysis

Economic analysis comparing Japan's debt/GDP ratio to the US, arguing that when accounting for government-held assets, Japan's net debt position isn't as alarming as headline figures suggest, and the US situation is more comparable than typically assumed.

Among developed countries, Japan has long had the highest debt/GDP ratio, currently ~232%. That seems pretty bad, and conversely has made some people say that the US debt is fine because it's still much lower than Japan's. But here are some points that might clarify the situation: First, that ratio has declined recently, from 258% in 2020. Second, the Japanese government holds a lot of stocks and foreign bonds. Its net debt/GDP is "only" 140%, and has declined since 2020. The US government doesn

EconomicsFinance

Research LessWrong Jan 16

Forfeiting Ill-Gotten Gains

By jefftk

8 score

AI Analysis

Parenting principle about enforcement: when punishing rule-breaking, the combination of penalty and forfeiture of gains must leave the violator worse off than if they hadn't broken the rule. Extended briefly to crime deterrence principles.

It's a holiday. The cousins are over, and the kids are having a great time. Unfortunately, that includes rampaging through the kitchen. We're trying to cook, so there's a "no cutting through the kitchen" rule. Imagine enforcement looks like: Kid: [dashes into kitchen, pursued by cousin] Adult: Out of the kitchen! Kid: Sorry! [Continues their path, leaving through the other door; escapes pursuit from more rule-abiding cousin] This doesn't work! The kid got what they wanted out of this interaction

ParentingIncentive Design