Last week brought the latest in an occasional series of conversations on AI between governmental leaders and Big Tech honchos. For months, the No. 1 topic on the table has been regulation. Businesses are counterintuitively clamoring for it, and the political parties are—for the moment, anyway—united over it.

Held behind closed doors, the Sept. 13 huddle really started the week before. On Sept. 8, Sens. Richard Blumenthal (D-CT) and Josh Hawley (R-MO) laid out a five-bullet framework the two are sure to follow in drafting their subcommittee’s anticipated AI bill.  

“License requirements, clear AI identification, accountability, transparency, strong protections for consumers and kids: Such common sense principles are a solid starting point,” Blumenthal said in introducing the framework.

“Our American families, workers and national security are on the line,” Hawley added. “We know what needs to be done; the only question is whether Congress has the willingness to see it through.”

Here’s a taste of what tech leaders and close observers have been saying about regulating AI in the days since the Sept. 13 meeting.   

1. ‘I agree that Congress should engage with AI to support innovation and safeguards. This is an emerging technology, there are important equities to balance here, and the government is ultimately responsible for that.’—Founder and head of Facebook and Meta Mark Zuckerberg

2. ‘Calls for a new licensing regime or a new regulator follow the approach seen in Europe, where a heavy regulatory touch has produced undesirable economic consequences. A better approach is to build on the success of light-touch innovation that has made the United States a world leader in the internet era.’—Technology policy researcher Jennifer Huddleston

3. ‘Regulate AI risk, not AI algorithms. Not all uses of AI carry the same level of risk. While some might seem harmless, others can have far-reaching consequences.’—IBM chairman and CEO Arvind Krishna

4. ‘It’s important for us to have a referee. I think there’s a strong consensus that there should be some AI regulation. … There is some chance above zero that AI will kill us all. I think it’s low, but there is some chance.’—Business magnate Elon Musk

5. ‘[The combined net worth of the Senate room on Sept. 13 was $550 billion, and it’s] hard to envision a room like that in any way meaningfully representing the interests of the broader public.’—AI Now Institute managing director Sarah Myers West

Buzzworthy developments of the past few days.

• It’s common for burgeoning innovations to run into roadblocks on the long bridge connecting early adoption with eventual ubiquity. Economists call such setbacks “switchover disruptions.” They have much to do with humans’ affinity for the familiar and resistance to change. Will healthcare AI stall out in this phase? Or plow ahead unimpeded? It depends on how well AI proponents build trust in the technology among providers, patients and the general public. The business-of-healthcare gurus (and twin brothers) James and Robert Rebitzer chunk up the challenge into three action items in an article published Sept. 14 in the Harvard Business Review. Check it out.
   
• AI advanced quite quickly from classifying input data (first wave) to producing new data (generative wave). The third wave will be interactive. That’s the belief of DeepMind cofounder Mustafa Suleyman, whose current company, Inflection, has amassed a cool $1.5 billion in investments from some of Big Tech’s biggest investors. The entrepreneur describes his forecast in an interview with MIT Technology Review. Interactive AI, he says, “is going to be animated. It’s going to have the potential freedom, if you give it, to take actions. It’s truly a step change in the history of our species that we’re creating tools that have this kind of, you know, agency.” Read the whole thing.
   
• What goes for healthcare AI also goes for … business-to-business marketing AI? Pretty much. As ChatGPT continues to develop, for example, B2B marketers and PR professionals “should learn to use AI as an assisting tool,” not as an alternative to a human, explains PR agency CEO Judith Ingleton-Beer at MarketingProfs.com. However, to make sure their content and strategies hit the mark, communications experts “must harness their own critical and creative thinking capabilities, alongside emotional intelligence. In the end, it’s all in the human touch.” Read the piece.
   
• Johnson & Johnson is tapping AI for help on multiple fronts. These include clinical diagnostics, drug discoveries, patient recruitment for research, supply chain management and OR operations analysis. The company fleshes out each of those in internal coverage posted here.
   
• Oracle is upping its game in healthcare. This week the software behemoth rolled out new cloud-based EHR offerings, generative AI services, public application programming interfaces (APIs) and more—all aimed at healthcare. Announcement here.
   
• Google promises its Bard chatbot will now play well with some classic Google apps. They’re talking Gmail, Docs, Drive, Google Flights and YouTube. And that’s not all the Googleplex denizens have been up to. They’ve also imbued Bard with a way to fact-check its own responses to queries. Yahoo Finance has the story.
   
• Nvidia continues to lead the GPU league in two key metrics of AI-readiness. Those would be overall performance and the number of machines using its chips. In recent MLPerf testing, Intel finished respectfully close to Nvidia, and Qualcomm did well in certain head-to-heads. However, as IEEE Spectrum reports, things are looking up for the GPU giant and may only get better.
   
• IBM has pledged to bring AI training to 2 million individuals, mostly in underserved communities, by the end of 2026. To pull it off, the company is expanding its collaborations with universities around the world. It’s also working with other industry players and building out the AI curriculum in its own educational operation, SkillsBuild. Details.
   
• From AIin.Healthcare’s news partners:
   
  • Radiology Business: GE HealthCare secures $44M from Gates Foundation to develop AI-powered ultrasound technology
  • Health Imaging: AI can generate usable radiology reports, but it requires a human touch
  • Radiology Business: After seeing 17 different doctors, boy with rare condition receives diagnosis from ChatGPT
  • Health Imaging: FDA clears AI-powered brain MRI evaluation software
  • Cardiovascular Business: Strain imaging making a significant impact on the treatment of heart patients