Anthropic unveils initiative to fund scientific research

Anthropic has launched an AI for Science initiative to provide resources for researchers focused on significant scientific endeavors, particularly in biology and the life sciences.

The program, announced on Monday, will grant selected "qualified" researchers up to $20,000 in Anthropic API credits over six months. Selection will be based on their scientific track record, the potential impact of their proposed project, and how effectively AI can advance their work. Participants will have access to Anthropic's standard AI model suite, including all publicly available Claude models.

"Advanced AI reasoning and language models can assist researchers in analyzing complex scientific data, formulating new hypotheses, designing experiments, and communicating results more effectively," Anthropic stated in a blog post. "We are especially keen to support projects where AI can speed up the understanding of intricate biological systems, analyze genetic information, accelerate drug discovery for major global diseases, enhance agricultural productivity, and related areas."

Anthropic joins several AI firms championing the 'AI for science' movement. This year, Google introduced an "AI co-scientist" tool designed to aid scientists in hypothesis generation and research planning. Anthropic, along with rivals like OpenAI and organizations such as FutureHouse and Lila Sciences, argues that AI has the potential to dramatically accelerate discovery, especially in medicine.

However, many researchers currently find AI's role in guiding the scientific process limited, primarily due to issues with reliability.

A key hurdle in creating an "AI scientist" is accounting for countless unknown variables. AI can be useful for broad exploration, such as narrowing down extensive lists of possibilities, but its capacity for the kind of inventive problem-solving that leads to genuine breakthroughs remains uncertain.

So far, outcomes from AI tools built for science have often been disappointing. In 2023, Google reported synthesizing around 40 new materials with its GNoME AI. Yet, an independent analysis concluded that none of those materials were genuinely novel discoveries.

Anthropic undoubtedly hopes its initiative will achieve more success than previous attempts.

Exhibit at TechCrunch Sessions: AI

Secure your spot at TC Sessions: AI and present your innovations to over 1,200 decision-makers—without a hefty investment. This offer is available until May 9 or while exhibition space remains.

Exhibit at TechCrunch Sessions: AI

Secure your spot at TC Sessions: AI and present your innovations to over 1,200 decision-makers—without a hefty investment. This offer is available until May 9 or while exhibition space remains.

The company states it will evaluate applications for the AI for Science program on the first Monday of each month, judging them on scientific merit, potential impact, technical feasibility, and biosecurity review (ensuring the research cannot be misused for harmful purposes). Researchers can apply via a form on Anthropic's website, with applications being assessed by the company's team, including experts in relevant fields.