Industry Watcher’s Digest

Buzzworthy developments of the past few days.

• Just like that, Anthropic has blossomed into an AI outfit to reckon with. It’s raised more than $7 billion over the past year, along the way winning support from the likes of Amazon, Google and Salesforce. And now Anthropic is signaling its intention to directly compete with OpenAI for generative AI dominance. Why would it not? Its founders, siblings Daniela and Dario Amodei, left OpenAI in 2021 to breathe life into their brainchild. Anthropic is in the business headlines this week because its most impressive model, Claude 3 Opus, can synopsize book-length documents of up to 150,000 words. That’s longer than The Last of the Mohicans by James Fenimore Cooper and only a little shorter than Salem’s Lot by Stephen King. Your move, ChatGPT.
   
• Scientific publishing powerhouse Elsevier has birthed a genAI-based clinical decision support tool. The company’s Elsevier Health division says it put heads together with medical AI company OpenEvidence to design the offering. Also chipping in on the front end were Cone Health, the University of New Mexico and “more than 30,000 physicians from across the U.S.” The result is ClinicalKey AI, which gives clinicians quick, point-of-care access to relevant sections of medical journals, medication guidelines, clinical references and medical textbooks. Full announcement here.
   
• How AI evolves over the coming years will depend on how humans build their relationship with it in the coming years. Does this point strike you as totally obvious yet oddly compelling (thanks to the use of relationship in this context)? If so, you might appreciate hearing from Homero Gil de Zúñiga, PhD, distinguished professor of media effects and AI at Penn State. “AI is going to do what we prompt it to do and what we ask it to do,” he tells the school’s news operation. “We must study how we interact with it, because if you’re not squeezing AI to its maximum capabilities, then AI will stay at the same level.” Q&A here, recent peer-reviewed paper on the same subject here.
   
• Drug development has been one of healthcare AI’s most talked-about uses for years. The leader of a small company that supplies AI-based recommendations to Big Pharma dishes on the dynamic in Fortune this week. “Our clients engage us to give them the insight and convert insight into foresight—in the shortest time possible and in the least expensive way,” Lifescience Dynamics founder and president Rafaat Rahmani tells the magazine. “These AI tools squeeze the most out of our data and bring that data alive.”
   
• Only 7% of psychologists worry about losing their job to AI. But some 62% are concerned about the technology’s potential to misinterpret data. Meanwhile 54% think chatbots will lack sufficient empathy with patients, 41% fret about threats to patient privacy and safety, and 40% are on the alert for biased outputs. The findings are from a survey of 100 U.S. psychologists. Full results from the exercise are posted at PsychologyJobs.com.
   
• Hippocratic AI of Palo Alto, Calif., has lined up more than 40 partner orgs to help it test what it calls ‘the world’s first generative AI-powered healthcare provider.’ Or, more modestly elsewhere in its announcement, “the industry’s first safety-focused large language model designed specifically for healthcare.” Either way, provider partners in the evaluation endeavor include Memorial Hermann Health System, University of Vermont Health Network and Fraser Health. Announcement here.
   
• From the AI research beat:
   
  • University of Rhode Island and UT-Austin: Scientists to study real-world eating behaviors using wearable sensors and artificial intelligence
     
  • University of Arkansas: AI Outperforms Humans in Standardized Tests of Creative Potential
     
  • University of York: New dressing robot can ‘mimic’ the actions of healthcare workers

• From AIin.Healthcare’s news partners:
   
  • Radiology Business: Philips launches new AI-enabled CT scanner aimed at cardiology at ECR 2024
     
  • Cardiovascular Business: Cardiologists partner with imaging AI specialists to improve care for high-risk heart patients
     
  • Health Imaging: AI shows potential to diagnose wrist fractures as well as radiologists