FutureHouse Unveils AI Tool for Data-Driven Biology Research

FutureHouse, a nonprofit backed by Eric Schmidt with the ambitious goal of developing an "AI scientist" within ten years, has unveiled a new tool designed to advance data-driven discovery in biology. This release follows just one week after FutureHouse debuted its API and platform.

Named Finch, the tool processes biological data—primarily from research papers—alongside user prompts like "What can you tell me about molecular drivers of cancer metastasis?" It then executes code, generates figures, and analyzes the results. In a series of X posts, FutureHouse co-founder and CEO Sam Rodriques likened its capabilities to those of a first-year graduate student.

Rodriques emphasized that accomplishing these tasks within minutes provides a significant advantage. He noted that Finch often uncovers fascinating insights and has proven highly valuable for the organization's internal projects.

In addition to open-ended analysis, Finch can also do directed data analysis. In this example, we had it do a differential expression and functional enrichment analysis of some RNAseq data (GEO accession GSE87466). It identifies upregulated genes (e.g. inflammatory modulators,… pic.twitter.com/phRybhTVpb

— Sam Rodriques (@SGRodriques) May 6, 2025

Like numerous startups and tech giants, FutureHouse believes that Finch and similar AI tools will eventually automate parts of the scientific process.

Earlier this year, OpenAI CEO Sam Altman suggested in an essay that superintelligent AI could dramatically speed up scientific discovery. Similarly, Anthropic's CEO, whose company recently launched an "AI for science" initiative, has predicted that AI may help develop cures for most cancers.

However, concrete evidence remains scarce. Many researchers currently find AI unconvincing as a guide for scientific inquiry. Notably, FutureHouse has not yet achieved any scientific breakthroughs or novel discoveries using its AI tools.

Biology, especially drug discovery, represents a promising field for AI firms. Precedence Research estimates the market was valued at $65.88 billion in 2024 and could grow to $160.31 billion by 2034.

Exhibit at TechCrunch Sessions: AI

Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last.

Exhibit at TechCrunch Sessions: AI

Secure your spot at TC Sessions: AI and show 1,200+ decision-makers what you’ve built — without the big spend. Available through May 9 or while tables last.

While there have been some notable successes, AI has not delivered an instant, miraculous solution in laboratory settings. Several AI-driven drug discovery companies, such as Exscientia and BenevolentAI, have faced significant clinical trial setbacks in recent years. At the same time, the accuracy of leading AI systems like Google DeepMind's AlphaFold 3 can be inconsistent.

Rodriques acknowledged that Finch still makes basic errors, which is why FutureHouse is hiring bioinformaticians and computational biologists to assess its performance, improve reliability, and train the system during its closed beta phase.

Interested individuals can register here.