#1: Artificial intelligence: Revolutionary or not?

Of the 624 companies that pitched at the four-day conference, the biggest overflow crowd may have belonged to Nvidia, which unlike the others isn't a health care company. Nvidia makes the silicon chips whose computing power, when paired with ginormous catalogs of genes, proteins, chemical sequences, and other data, will "revolutionize" drug-making, according to Kimberly Powell, the company's vice president of health care. Soon, she said, computers will customize drugs as "health care becomes a technology industry." One might think that such advances could save money, but Powell's emphasis was on their potential for wealth creation. "The world's first trillion-dollar drug company is out there somewhere," she dreamily opined.

Some health care systems are also hyping AI. The Mayo Clinic, for example, highlighted AI's capacity to improve the accuracy of patient diagnoses. The nonprofit hospital system presented an electrocardiogram algorithm that can predict atrial fibrillation three months before an official diagnosis; another Mayo AI model can detect pancreatic cancer on scans earlier than a provider could, said Matthew Callstrom, chair of radiology at the Mayo Clinic in Rochester, Minnesota.

No one really knows how far — or where — AI will take health care, but Nvidia's recently announced $100 million deal with Amgen, which has access to 500 million human genomes, made some conference attendees uneasy. If Big Pharma can discover its own drugs, "biotech will disappear," said Sherif Hanala of Seqens, a contract drug manufacturing company, during a lunch-table chat with KFF Health News and others. Others shrugged off that notion. The first AI algorithms beat clinicians at analyzing radiological scans in 2014. But since that year, "I haven't seen a single AI company partner with pharma and complete a phase I human clinical trial," said Alex Zhavoronkov, founder and CEO of Insilico Medicine — one of the companies using AI to do drug development. "Biology is hard."