1.3M+ experiments exposing bias in critical AI decision-making
Have we been measuring AI political bias wrong? A better approach is possible.
Why ideological preferences and epistemic failure in LLMs are not the same thing — and why the difference matters
The AI Chemist: To be trustworthy, LLMs need to show their work
Good scientists reveal how they do their experiments and report their results; so should any machine-driven research
Underspecified Human Decision Experiments Considered Harmful
Decision-making with information displays is a key focus of research in areas like human-AI collaboration and data visualization. However, what constitutes a decision problem, and what is required for an experiment to conclude that decisions are flawed, remain imprecise. We present a widely applicable definition of a decision problem synthesized from statistical decision theory and information economics. We claim that to attribute loss in human performance to bias, an experiment must provide the information that a rational agent would need to identify the normative decision. We evaluate whether recent empirical research on AI-assisted decisions achieves this standard. We find that only 10 (26%) of 39 studies that claim to identify biased behavior presented participants with sufficient information to make this claim in at least one treatment condition. We motivate the value of studying well-defined decision problems by describing a characterization of performance losses they allow to be conceived.
Import AI 446: Nuclear LLMs; China's big AI benchmark; measurement and AI policy
Will AIs be jealous of one another?
Solving a Million-Step LLM Task with Zero Errors
LLMs have achieved remarkable breakthroughs in reasoning, insights, and tool use, but chaining these abilities into extended processes at the scale of those routinely executed by humans,...
AI Sycophancy and Decisions
We examine whether sycophantic AI advice distorts decisions. Our experiment involves 1,500 participants in 30 decision environments spanning core domains in eco
If You’re Going To Defend AI And Whine About Its Critics, You Should Probably Be Honest About Its Actual Harms
I think this recent post by AI industry CEO Matt Shumer is worth a read. In it, he basically explains how quickly LLMs (large language models) are evolving to supplant many developers and prog…
Can AI Agents Synthesize Scientific Conclusions?
Scientific AI agents increasingly retrieve evidence, reason across sources, and synthesize conclusions used in consequential decisions. Yet, their ability to do so in high-stakes domains such as health remains unclear. We introduce SciConBench, a large-scale live benchmark of 9.11K questions and expert-written conclusions from systematic reviews to evaluate open-domain scientific conclusion synthesis. The benchmark draws on an expert-validated automated evaluation pipeline that decomposes conclusions into atomic facts and measures correctness and comprehensiveness via factual precision and recall. To mitigate data leakage, we further introduce SciConHarness, a clean-room evaluation harness that equips agents with controlled web interaction to ensure valid measurement. Evaluating 8 frontier models and deep research agents, we find that factual quality remains low: under clean-room settings, the best agent achieves only a factual F1 of 0.337. Our clean-room setting consistently reduces performance relative to unconstrained evaluation, suggesting that leakage inflates estimates of models' true synthesis capabilities. Finally, we audit consumer-facing agents (e.g., Google AI Overview, OpenEvidence) and find they frequently generate incomplete and sometimes contradictory conclusions, even when the ground-truth answer is available. Overall, our results show that reliable synthesis of scientific conclusions remains an open challenge, and that clean-room evaluation is essential for assessing open-domain AI agents.
LLM Agents are simply Graph — Tutorial For Dummies
Ever wondered how AI agents actually work behind the scenes?
LLM Agents are simply Graph — Tutorial For Dummies
Ever wondered how AI agents actually work behind the scenes?
It’s remarkably easy to inject new medical misinformation into LLMs
Changing just 0.001% of inputs to misinformation makes the AI less accurate.
One year as an AI Engineer: The 5 biggest misconceptions about LLM reliability I've encountered
535 votes, 59 comments. After spending a year building evaluation frameworks and debugging production LLM systems, I've noticed the same…
wharton-generative-ai-labs/AIBO
An open-source tool for running controlled behavioral experiments on AI systems at scale.
