Stay informed with free updates
Simply sign up to the Artificial intelligence myFT Digest — delivered directly to your inbox.
Artificial intelligence start-up Anthropic is tailoring its Claude chatbot to researchers and life sciences companies, as AI groups race to create specialised applications from the technology.
The San Francisco-based group said on Monday it is integrating Claude into tools that scientists already use, including lab management systems, genomic analysis platforms and biomedical databases, to tackle time-consuming tasks such as data analysis and literature review.
Anthropic, which was valued at $170bn in September, said drugmaker Novo Nordisk has already used its AI model to cut clinical study documentation from more than 10 weeks to 10 minutes, while drug developer Sanofi said the majority of its employees use Claude every day.
The move comes as tech groups are spending billions of dollars on AI products and models, believing the technology can benefit a range of industries from healthcare to energy and education.
This has included a focus on the life sciences industry as top AI companies and start-ups bet on the potential for AI to speed up drug discovery and tackle disease.
OpenAI and Mistral have recently announced new units focusing on scientific research. In February, Google unveiled a “co-scientist” tool that could help scientists come up with new hypotheses, and last week said its open Gemma model had helped discover a new potential cancer therapy pathway.
“What I’m chasing is to bring to biologists the experience that software engineers have [with code generation],” said Eric Kauderer-Abrams, head of life sciences at Anthropic. “You can sit down with Claude and brainstorm ideas, generate hypotheses together.”
The company has seen success with its coding tool, Claude Code, which outperforms those of its competitors. Kauderer-Abrams said this helps give it an edge in the life sciences industry.
“We’re much more focused on amplifying the capabilities of individual scientists and building tools that accelerate the scientists’ workflows than other companies are,” said Kauderer-Abrams. He added that rival groups are trying to do that as well as directly doing science themselves. Some, such as DeepMind spin-off Isomorphic Labs, are trying to find their own drugs.
However, so far, no drugs discovered by AI have been approved and many have failed in clinical trials. One hurdle has been getting enough data to create a general-purpose algorithm that can solve many different problems.
Anthropic said it has made its models suitable for pharmaceutical research by bringing down the amount of times it produces “hallucinations” — or factual errors. It has also offered audit trails for regulatory compliance and the ability to verify every insight against original sources.
Kauderer-Abrams said the company was also banning requests related to prohibited agents, which could be used to make chemical weapons.
The AI group’s push into life sciences follows recent breakthroughs that showed large language models have the potential to help in scientific research. Last month, both Google DeepMind and OpenAI achieved gold medal-level performance at prestigious competitions for coding.
Kauderer-Abrams said language models can take advantage of large existing and publicly available datasets in biology, such as ones on genomics and protein sequencing, which can be used to tailor the models for scientific research.
“In life sciences, that’s one area where pretty much everyone can agree that we can bring things that are unambiguously amazing,” said Kauderer-Abrams.
Read the full article here
