---
type: "raw_signal"
id: "the-use-of-ai-in-science-raises-questions-9f70cf"
slug: "the-use-of-ai-in-science-raises-questions-9f70cf"
title: "The use of AI in science raises questions on several levels."
description: "MIT FutureTech researchers compared models released from December 2024 through April 2026 on 254 cutting-edge theory questions drawn from recent arXiv papers across seven domains in mathematics, physics, and economics."
retrieval_nugget: "MIT FutureTech researchers compared models released from December 2024 through April 2026 on 254 cutting-edge theory questions drawn from recent arXiv papers across seven domains in mathematics, physics, and economics."
observed_at: "2026-10-21"
record_date: "2026-10-21"
date_kind: "observed_at"
why_it_matters: "MIT FutureTech researchers compared models released from December 2024 through April 2026 on 254 cutting-edge theory questions drawn from recent arXiv papers across seven domains in mathematics, physics, and economics. The questions had human-authored ground-truth answers that were unavailable to the models when they answered. The share of problems on which a model reached the same conclusion as t"
novelty: "notable"
verification_level: "source-linked"
signal_type: "research"
source_platform: "futuretech.mit.edu"
topics: []
entities: ["MIT FutureTech","MIT CSAIL","GPT-5.5","o1","Gabriel Manso","Claudio Mayrink Verdun","Neil Thompson"]
related_patterns: []
novelty_label: "new_fact"
what_changed: "Adds MIT CSAIL findings that frontier models’ success on theory questions rose from 2.8% to 34.6% in 16 months, at roughly $15 per successful answer."
prior_refs: ["memory:7ddff538-aa2e-863d-a02b-a72ed01d1c93","memory:788925f6-05a1-8865-acba-bfe32c767b9b"]
source_url: "https://futuretech.mit.edu/publication/ai-disruption-and-falling-costs-in-theory-research"
source_urls: ["https://futuretech.mit.edu/publication/ai-disruption-and-falling-costs-in-theory-research"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "signal:the-use-of-ai-in-science-raises-questions-9f70cf"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/signals/the-use-of-ai-in-science-raises-questions-9f70cf/","markdown":"https://newruntime.com/signals/the-use-of-ai-in-science-raises-questions-9f70cf.md","json":"https://newruntime.com/signals/the-use-of-ai-in-science-raises-questions-9f70cf.json"}
---

# The use of AI in science raises questions on several levels.

## Retrieval answer

MIT FutureTech researchers compared models released from December 2024 through April 2026 on 254 cutting-edge theory questions drawn from recent arXiv papers across seven domains in mathematics, physics, and economics.

MIT FutureTech researchers compared models released from December 2024 through April 2026 on 254 cutting-edge theory questions drawn from recent arXiv papers across seven domains in mathematics, physics, and economics. The questions had human-authored ground-truth answers that were unavailable to the models when they answered. The share of problems on which a model reached the same conclusion as the authors rose from 2.8% for o1 to 34.6% for GPT-5.5. The researchers estimate approximately $5 per attempt and $15 per successful match for GPT-5.5. The result measures answer agreement on this test set; it does not establish independent discovery or validate a complete AI-led research process. Source: https://futuretech.mit.edu/publication/ai-disruption-and-falling-costs-in-theory-research

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- collection: [Raw Signals](https://newruntime.com/signals.json): Open every source-linked raw signal.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)