New AI chatbot equips doctors with the latest research – Interesting Engineering

By subscribing, you agree to our and You may unsubscribe at any time.
In order for chatbots to be useful to doctors and other health professionals, they are going to need access to the latest research. But current models simply don’t have access to data beyond their latest update. Daniel Nadler has been working to resolve this issue with his new startup OpenEvidence.
He plans to achieve his lofty goal by “marrying these language models with a real-time firehose of clinical documents,” Nadler told Forbes on Thursday. He claims that his new model “can answer with an open book, as opposed to a closed book.”
How can medical professionals separate credible, actionable information from the noise? This problem was especially apparent during the pandemic, claimed Nadler. “The fundamental construct of the problem was identical,” he argued. “An information overload and a need to triage that information and a need to use computers to do so.”
And it seems Nader’s solution is working. In March, OpenEvidence was selected to participate in a Mayo Clinic Platform accelerator, a move that saw more than 10,000 clinicians sign up for early access to his platform.
Today, OpenEvidence is interactive allowing users to tailor their questions to precise patient scenarios and ask follow-ups. It can also scan tens of thousands of journals from a document pool that includes more than 35 million journal articles. And the documents are up to date: there is but a mere a 24-hour lag time to process the new journal articles and get them into the retrieval pool.
OpenEvidence is also providing citations to the journal articles that it is pulling from guaranteeing that there are less faulty answers. 
To drive interest in his new platform, Nadler has made OpenEvidence free to use for early adopters who are licensed medical professionals. Antonio Forte, a professor of plastic surgery at Mayo Clinic who is on OpenEvidence’s medical advisory board, told Forbes he uses UpToDate regularly and boasts he can get an answer “within 30 seconds, not within 10 minutes.”
As for revenues, Nadler says he hasn’t decided on a revenue model yet, debating between subscription-based and ad-based. One thing he is certain of however is that he does not wish OpenEvidence to become a chatbot for the average patient. 
“That is not a technical problem. That is a regulatory and ethical problem,” said Nadler. “[There] is a very firm limit to any conceivable harm that could come from the usage of the technology to a patient, because it is always being intermediated by a professional.”