Contemporary AI lacks the imagination to diverge or negate in science

• URL: http://arxiv.org/abs/2606.08251v2
• Date: Tue, 09 Jun 2026 03:31:39 GMT
• Title: Contemporary AI lacks the imagination to diverge or negate in science
• Abstract summary: Authors of 121,640 recent preprints across biology, medicine, chemistry, and the social sciences invited to judge ideas that large language models generated from context and puzzles of their own papers.<n> 6,749 scientists returned 25,139 sets of ratings on novelty, empirical feasibility, probability of being true, and favorability of adoption.<n>Second, scientists reward ideas that resemble their own and prize probability over novelty, though social scientists tolerate risk more readily than life scientists.<n>Third, automated evaluators on which the community currently relies agree only weakly with expert judgment, and retrieval augmentation and scientist persona prompting yield only marginal gains.
• Score: 4.024882945191284
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Bold projections that artificial intelligence will accelerate scientific discovery have raced ahead of evidence from working scientists, and the field still lacks large-scale, scientist-in-the-loop tests of these claims. Here we mount the largest such evaluation to date and map what AI cannot yet do for science. We invited authors of 121,640 recent preprints across biology, medicine, chemistry, and the social sciences to judge ideas that large language models (LLMs) generated from the context and puzzles of their own papers. 6,749 scientists returned 25,139 sets of ratings on novelty, empirical feasibility, probability of being true, and favorability of adoption. Three patterns emerge. First, non-reasoning LLMs collapse into a narrow "hivemind" of similar ideas; reasoning models roam a wider hypothesis space, yet no model class spontaneously proposes null hypotheses -- a move humans make more freely. Second, scientists reward ideas that resemble their own and prize probability over novelty, though social scientists tolerate risk more readily than life scientists. Senior social scientists are the harshest critics, and their skepticism is well-earned: LLMs falter most in pluralistic fields like the social sciences that demand context-aware interpretation and evolving theories. Third, automated evaluators on which the community currently relies -- LLM-as-a-judge, artificial metrics, and even state-of-the-art (SOTA) models -- agree only weakly with expert judgment, and retrieval augmentation and scientist persona prompting yield only marginal gains. A Qwen3-14B reward model we post-trained on human ratings captures field taste nuances, beats SOTA models by up to 27%, and closes the gap to the inter-rater consistency of independent peer reviewers. For all the hype, today's scientific AI still represents a collaborator whose imagination, outputs and judgment benefit from human grounding.

Related papers

• AgentIdeaBench: Benchmarking Scientific Ideation in the Agent Era [55.37558486468993]
  We introduce AgentIdeaBench, a benchmark that evaluates scientific ideation under two matched settings, static observation and active exploration.<n>We report matched Static-Active evaluations for 33 LLMs across 40 densely scored subfields spanning five disciplines.<n>Active exploration reveals considerably more capability headroom, and that headroom is unevenly distributed across models.
  arXiv  Detail & Related papers  (2026-09-07T15:19:13Z)
• Can LLMs Discover Scientific Laws in Real and Parallel Worlds? [58.76147897928744]
  We introduce SCILAWS-BENCH, a benchmark for scientific law discovery built from published research and real scientific data.<n>It comprises 118 problems drawn from 381 scientific papers, covering 291 candidate laws and roughly 8M real data points across six scientific disciplines.<n>We find that predictive fit can diverge from scientific validity, memorization shapes whether models reproduce or move beyond published formulas, and our best-of-N study reveals a selection bottleneck.
  arXiv  Detail & Related papers  (2026-09-01T17:17:59Z)
• AI Can Learn Scientific Taste [123.03470627335189]
  Great scientists have strong judgement and foresight, closely tied to what we call scientific taste.<n>Here, we use the term to refer to the capacity to judge and propose research ideas with high potential impact.<n>Our findings show that AI can learn scientific taste, marking a key step toward reaching human-level AI scientists.
  arXiv  Detail & Related papers  (2026-03-15T16:31:51Z)
• Autonomous Agents for Scientific Discovery: Orchestrating Scientists, Language, Code, and Physics [82.55776608452017]
  Large language models (LLMs) provide a flexible and versatile framework that orchestrates interactions with human scientists, natural language, computer language and code, and physics.<n>This paper presents our view and vision of LLM-based scientific agents and their growing role in transforming the scientific discovery lifecycle.<n>We identify open research challenges and outline promising directions for building more robust, generalizable, and adaptive scientific agents.
  arXiv  Detail & Related papers  (2025-10-10T22:26:26Z)
• AI Scientists Fail Without Strong Implementation Capability [33.232300349142285]
  The emergence of Artificial Intelligence (AI) Scientist represents a paradigm shift in scientific discovery.<n>Recent AI Scientist studies demonstrate sufficient capabilities for independent scientific discovery.<n>Despite this substantial progress, AI Scientist has yet to produce a groundbreaking achievement in the domain of computer science.
  arXiv  Detail & Related papers  (2025-06-02T06:59:10Z)
• Scaling Laws in Scientific Discovery with AI and Robot Scientists [72.3420699173245]
  An autonomous generalist scientist (AGS) concept combines agentic AI and embodied robotics to automate the entire research lifecycle.<n>AGS aims to significantly reduce the time and resources needed for scientific discovery.<n>As these autonomous systems become increasingly integrated into the research process, we hypothesize that scientific discovery might adhere to new scaling laws.
  arXiv  Detail & Related papers  (2025-03-28T14:00:27Z)
• Evaluating Sakana's AI Scientist: Bold Claims, Mixed Results, and a Promising Future? [20.188659973082643]
  Sakana recently introduced the 'AI Scientist', claiming to conduct research autonomously, i.e. they imply to have achieved what we term Artificial Research Intelligence (ARI)<n>Our evaluation of the AI Scientist reveals critical shortcomings.
  arXiv  Detail & Related papers  (2025-02-20T06:22:03Z)
• Large language models surpass human experts in predicting neuroscience results [60.26891446026707]
  Large language models (LLMs) forecast novel results better than human experts.<n>BrainBench is a benchmark for predicting neuroscience results.<n>Our approach is not neuroscience-specific and is transferable to other knowledge-intensive endeavors.
  arXiv  Detail & Related papers  (2024-03-04T15:27:59Z)
• Accelerating science with human-aware artificial intelligence [2.7786142348700658]
  We show that incorporating the distribution of human expertise by training unsupervised models dramatically improves (up to 400%) AI prediction of future discoveries. These models succeed by predicting human predictions and the scientists who will make them. Accelerating human discovery or probing its blind spots, human-aware AI enables us to move toward and beyond the contemporary scientific frontier.
  arXiv  Detail & Related papers  (2023-06-02T12:43:23Z)
• Can ChatGPT be used to generate scientific hypotheses? [0.2010294990327175]
  generative AI seems to be able to effectively structure vast amounts of scientific knowledge and provide interesting and testable hypotheses. The future scientific enterprise may include synergistic efforts with a swarm of "hypothesis machines", challenged by automated experimentation and adversarial peer reviews.
  arXiv  Detail & Related papers  (2023-03-30T20:40:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.