This past summer McKinsey Global Institute projected almost one-third of workplace activities in the U.S. will be automated by 2030, driving something like 12 million “occupational transitions.”

Most of the change will be facilitated by AI.

Generative AI is quickening the pace, McKinsey noted, and the advance of artificial general intelligence, or AGI, is adding to the anxiety.

However, some jobs are well out of AI’s reach no matter how adept the technology becomes.

This week Yahoo Finance identifies 20 such roles. They include three specific to healthcare, as follows (with comments from Yahoo reporter Afifa Mushtaque):

Nurses.

“Human touch and intuition are vital in nursing, and AI can't fully replicate these capabilities. Besides, when one is sick, the warmth of human support becomes essential. Thus, nursing is one of the jobs AI will probably never replace.”

Surgeons.

“Surgeons can adapt to unexpected complications and make real-time decisions, which is crucial in complex procedures. Furthermore, surgeries often require delicate and precise movements, which are challenging for AI to observe accurately [and for unassisted robots to replicate]. Surgery is thus one of the professions least likely to be replaced by AI.”

Mental health counselors.

“AI lacks the capacity to truly empathize with or understand the complexity of human emotions and the context in which they arise. Moreover, the therapeutic relationship, built on trust and rapport, is a fundamental aspect of successful counseling, and it is challenging for AI to establish such connections. It is one of the important jobs artificial intelligence can’t replace.”

The list includes two jobs that could rightly be called healthcare-adjacent—emergency responders and social workers.

Mushtaque says Yahoo Finance made the present safe-job forecast by identifying work that can’t be done without three attributes: nuanced human sensibilities, original creative thinking and/or adaptable physical dexterity.

Meanwhile, in its report of this past summer, McKinsey predicted automation will affect a broad set of work activities involving expertise, interaction with people and, yes, even creativity.

“Without generative AI, our research estimated, automation could take over tasks accounting for 21.5% of the hours worked in the U.S. economy by 2030,” the McKinsey authors wrote. “With it, that share has now jumped to 29.5%.”

The full Yahoo Finance article is here.

Buzzworthy developments of the past few days.

• The list of FDA-approved medical devices packing AI is closing in on a nice round number—700. As of Oct. 19, the count stood at 692. Curiously, not a single product on the list, which FDA itself maintains, uses generative or large-language AI. Less surprising is that none taps AGI, artificial general intelligence, either. Meanwhile more than 85% of the listed devices are aimed at radiology. Browse the list and get more insights into it here.
   
• GE HealthCare has legit bragging rights regarding that list. The company is spreading the word that some 58 of the 692 cleared devices are its brainchildren.
   
• Generative AI stands to reward adopters with “much more value” than other kinds of AI. So says distinguished Gartner analyst Mary Mesaglio. The thought leader made the pronouncement at a Gartner event this month. “We are moving from what machines can do for us to what machines can be for us,” she told more than 8,000 CIOs and other IT executives. Get more from CIO Dive.
   
• Three of four Americans do not trust AI in a healthcare setting. Yet three-fifths believe the technology could help make up for healthcare staffing shortages. And 61% trust their own physicians to use it properly. Despite the latter, 63% worry the technology will cut patient/doctor facetime. The findings are from registry vendor Carta Healthcare, which surveyed 1,027 consumers in August. The company has posted its key findings in an infographic.
   
• Unethical data collection. Cybersecurity threats. Amplified biases. Distributed misinformation. It’s enough to make a healthcare institution call the whole “AI thing” off. Or, in the hopes of the World Health Organization, enough to spur regulating bodies around the globe to get busy regulating AI—and doing it right. Read the exhortation in WHO’s own words.
   
• A new AI body is born in academia. The AI for Health Institute launched this month under the auspices of the McKelvey School of Engineering at Washington University in St. Louis. The institute’s inaugural director, computer scientist and engineer Chenyang Lu, PhD, calls cutting-edge AI the “new frontier of healthcare.” Basic AI tools are “insufficient” for healthcare’s unfolding challenges, adds Lu, who is also a professor of anesthesiology. Announcement.
   
• AI regulation may be close at hand in Europe. If pending legislation passes EU review, which could happen in a matter of hours, the world will see the first laws guardrailing AI anywhere. “There is a plausible scenario that we keep talking until the middle of the night and close the file on 25 October,” Dragoș Tudorache, a member of the European Parliament and an AI expert, tells The Guardian.
   
• A nonprofit organizer of healthcare breakthroughs has picked Cleveland Clinic to lead new research into quantum computing. The organization, Wellcome Leap, will support the medical institution in a project focused on “identifying, developing and demonstrating biology and healthcare applications that will benefit from the use of quantum computers.” IBM and Algorithmiq have a major hand in the work. The pot will be sweetened by up to $40 million for research and $10 million in challenge prizes. Details.
   
• From AIin.Healthcare’s news partners:
   
  • Radiology Business: FDA highlights radiology's continued dominance in AI-enabled device submissions
  • Cardiovascular Business: AI detects signs of AFib in asymptomatic patients
  • Health Imaging: NewVue launches AI-powered radiology command center
  • Radiology Business: FDA clears artificial intelligence-based MRI assistant from South Korean developer
  • Health Imaging: BabyNet AI model predicts fetal birth weight