This week Mark Zuckerberg’s Meta began offering the latest version of its large language AI model, Llama 2, for free and on an open-source basis. And that’s to commercial concerns as well as academic researchers.

Announcing the move July 18, Meta said it’s working on the giveaway strategy with “preferred partner” Microsoft, a major financial backer of rival model ChatGPT from OpenAI. Plus the model will run in Windows on Microsoft’s Azure cloud service. “By making AI models available openly, [we’re hoping] they can benefit everyone,” Meta explained.

The development has stirred up a feeding frenzy among some top technology reporters. Here are a few highlights. 

• “Meta is sticking to a long-held belief that allowing all sorts of programmers to tinker with technology is the best way to improve it.”—Mike Isaac and Cade Metz of The New York Times
   
• Why would Microsoft support an offering that might degrade OpenAI’s value? Because “giving developers choice in the types of models they use would help extend [Microsoft’s] position as the go-to cloud platform for AI work.”—Katie Paul of Reuters
   
• Meta will not directly make money from the deal with Microsoft. However, “it stands to potentially benefit by getting its homegrown AI software in the hands of more users and developers.”—Jonathan Vanian of CNBC
   
• Meta has received more than 100,000 requests from researchers to use its first model, released in February. “But the open-source LLaMA 2 will likely have a far bigger reach.”—Emilia David of The Verge
   
• “While Meta is making the code freely available, it does have an acceptable use policy guiding how it can be used commercially.”—Ina Fried of Axios
   
• “Meta has a large and powerful AI research organization and ample cash to support its work and recruit the best AI talent.”—Mark Sullivan of Fast Company
   
• “There is a possibility that, by giving all comers the chance to launch a rival to ChatGPT, Bard or Microsoft’s Bing chatbot, Meta is potentially diluting the competitive edge of tech peers such as Google.”—Dan Milmo of The Guardian
   
• Meta is working with Qualcomm to make Llama 2 workable on smartphones, tablets and PCs. “These new on-device AI experiences, powered by Snapdragon, can work in areas with no connectivity or even in airplane mode,” Qualcomm said Tuesday.—Corinne Reichert of CNET

More from Meta: July 18 announcement | Downloadable Llama 2 model | Research paper | Technical details

Buzzworthy developments of the past few days.

• It’s not as though OpenAI is napping while hungry challengers like Meta circle overhead. For one thing, the ChatGPT company is pumping $5 million into the American Journalism Project. CEO Sam Altman’s stated aim is to help the profession leverage AI for “strengthen[ing] our democracy by rebuilding the country’s local news sector.” The move may reflect equal parts PR savvy and business sophistication. OpenAI is dangling the carrot of up to $5 million worth of credits toward advanced API (application programming interface) tools in front of the AJP and its member organizations—any or all of whom “may choose to build and use tools utilizing the technology.” Announcement.
   
• Meanwhile, on the heels of the FTC’s initial investigation into OpenAI over possible consumer protection violations, the CMIO blogger Dr. Jayne of HIStalk dishes some dirt. One physician she heard yapping about ChatGPT “boasted about how she was able to create lab results letters for her patients, sending a full day’s worth of letters in under three minutes,” Dr. Jayne writes in a “Curbside Consult” column. “The ability to create and proofread those letters in the cited timeframe is questionable at best. Based on the looks on the faces of some of the colleagues she was speaking to, I wonder if they were questioning her professional judgement.” Read the rest.
   
• A well-connected consultancy is forming a consortium to advise healthcare providers on sound adoption of generative AI. The firm, Chicago-based digital transformation shop Avia (the healthcare consultancy, not the sneaker maker), says it’s already rallied more than 20 health systems plus the American Hospital Association and others. Avia says the group’s first big project will be producing a “board-ready, network-validated expert playbook” to share among participants. Topics will include priority use cases, ethics and transparency, privacy and security, and data ownership. Announcement.
   
• Patient-facing healthcare workers in any setting could use comfy scrubs and “chillout pods.” Both stress relievers are being 3D-printed in Australia thanks to researchers at the Royal Melbourne Institute of Technology. Story and photos.
   
• Big doings with healthcare AI in the Peach State. Researchers from Emory University’s School of Medicine and the Georgia Institute of Technology are working on the use of natural language processing to accelerate interventions following diagnoses of COVID and, potentially, other diseases. Also in the mix is Switchboard MD, an AI startup founded by Emory doctors. The team piloted a protocol for turning patient messages in the EHR into action on the ground. Peer-reviewed study here, Emory news item here.
   
• Anti-aging science and AI development seem a natural fit for hand-in-hand advancement. They may yet be. Researchers at MIT, Harvard and the Broad Institute, together with colleagues at biotech AI company Integrated Biosciences, have published a paper in Nature Aging confirming this matchmaker’s hunch. The scientists describe their use of AI to identify submolecular senolytic compounds that can be coaxed to inhibit aging-related processes like fibrosis, inflammation and cancer. Lead author Felix Wong, PhD: “These data demonstrate that we can explore chemical space in silico and emerge with multiple candidate anti-aging compounds that are more likely to succeed in the clinic, compared to even the most promising examples of their kind being studied today.” Full study. News summary.
   
• Three publicly traded healthcare companies have been named frontrunners in the race to use AI for revolutionizing medicine while rewarding investors. The three: GE Healthcare, Medtronic and McKesson. The outfit doing the naming: InvestorPlace. Story.
   
• Know any mental health researchers looking to publish on the use of generative AI in their specialty? Send them to JMIR Mental Health. Better yet, send them the relevant call for papers.
   
• From AIin.Healthcare’s news partners:
   
  • Radiology Business—AI trained on adult data set performs adequately when turned toward pediatric population
  • Radiology Business—Imaging AI vendor’s stock leaps 86% after inking ‘strategic commercial agreement’ with Radiology Partners