AI in Healthcare

Today's Stories

1. 1

   Google's medical AI matches real doctors at managing complex diseases, study finds

   Google's conversational medical AI system, AMIE, performed as well as primary care physicians when managing patients with complex, ongoing health conditions, according to research published in the journal Nature on June 17. The study tested AMIE in realistic clinical scenarios and found it could ask the right questions, interpret symptoms, and suggest appropriate next steps just as effectively as trained doctors. This is not a chatbot answering simple health questions — it was tested on the kind of nuanced, multi-condition cases that typically require years of medical training.

   If this technology reaches patients, people in areas with doctor shortages — or anyone waiting weeks for a GP appointment — could get reliable medical guidance much sooner.

2. 2

   General AI chatbots beat purpose-built clinical AI tools in head-to-head medical test

   A study published in Nature Medicine found that general-purpose large language models (LLMs — AI systems like ChatGPT that can read and write text on any topic) outperformed specialized clinical AI tools designed and marketed specifically for healthcare settings. The research compared AI tools across a range of real medical tasks and found the off-the-shelf AI came out ahead. This matters because hospitals have spent significant money purchasing dedicated medical AI software, and this study questions whether those purchases were worthwhile.

   Hospitals and clinics may need to rethink which AI tools they are actually paying for — patients could eventually benefit if healthcare providers shift to cheaper, more capable general AI instead of expensive specialized systems.

3. 3

   Merck bets up to $510 million on AI startup Protillion to discover new drugs faster

   Pharmaceutical giant Merck announced a collaboration with AI drug discovery startup Protillion on June 16-17, with total payments reaching up to $510 million if agreed milestones are hit. Protillion uses AI to identify promising drug targets — essentially, to figure out which molecules in the body are worth trying to treat with a new medicine. AI drug discovery works by analyzing enormous amounts of biological data to find patterns that human scientists would take years to spot.

   If AI-assisted drug discovery delivers, new medicines for diseases like cancer or Alzheimer's could reach clinical trials significantly faster than the traditional decade-long timeline.

4. 4

   Pfizer licenses AI drug design technology from startup Chai Discovery

   Pfizer signed a licensing agreement with Chai Discovery, an AI startup that specializes in predicting how proteins (the biological machinery that drugs interact with) fold and behave — a crucial step in designing new medicines. The deal, reported June 9, gives Pfizer access to Chai's AI tools so its researchers can design better drug candidates without relying solely on slow, expensive lab experiments. Chai's technology builds on advances in protein-structure prediction that have transformed biology over the past few years.

   Faster, AI-guided drug design at a company the size of Pfizer could mean more potential treatments entering clinical trials, eventually broadening the options available to patients.

5. 5

   AI startup Radical Numerics raises $50 million to reinvent how biology is studied using AI-generated DNA

   Radical Numerics, a startup whose researchers previously built AI systems capable of generating synthetic DNA sequences (artificial genetic code designed by an AI rather than evolved in nature), raised $50 million in new funding, Fortune reported on June 15. The company argues that understanding biology requires modeling entire living systems — not just individual molecules — and that AI can do this at a scale no human team could manage. Its work spans drug discovery, bio-defense (protecting against biological threats), and fundamental biological research.

   This kind of research could eventually lead to medicines designed from the ground up by AI, or faster responses to newly emerging biological threats — though practical treatments are still years away.

6. 6

   AI safety startup Pramaana Labs raises $27 million to make AI reliable enough for doctors and lawyers

   Pramaana Labs secured a $27 million seed round (early-stage investment) led by Khosla Ventures on June 17, with a mission to bring 'formal verification' (a mathematical technique that proves software behaves exactly as intended, with no hidden errors) to AI systems. The startup is targeting sectors where AI mistakes carry serious consequences — specifically healthcare drug discovery, legal work, and tax preparation. Current AI systems can confidently produce wrong answers, which is manageable in low-stakes tasks but dangerous in medicine or law.

   If Pramaana succeeds, AI tools used by your doctor or lawyer could become far more trustworthy — reducing the risk that an AI-generated recommendation leads to a harmful medical or legal error.