|
Hi! If you’re finding value in our Applied AI newsletter, I encourage you to consider subscribing to The Information. It contains exclusive reporting on the most important stories in tech, like this story from Aaron and Wayne on Apple software chief Craig Federighi‘s cautious approach to improving the company’s AI standing. Save up to $250 on your first year of access. Welcome back! Even as OpenAI begins a herculean effort to generate billions of dollars in chatbot ad revenue in the coming years, CFO Sarah Friar is already discussing its next business opportunity: “value sharing.” Speaking at a panel at Davos moderated by The Information CEO Jessica Lessin, Friar suggested that in the field of drug discovery, her company could, for instance, take a “license to the drug that is discovered” using OpenAI’s technology. In other words, OpenAI would take a profit-sharing stake in the financial upside its AI creates for customers. Friar didn’t elaborate on how OpenAI would help firms discover new drugs. Pharma and biotech firms are already sophisticated in using various forms of AI for drug discovery. For now, such firms could use OpenAI models to analyze lots of data and to propose hypotheses or methods of testing. But OpenAI seems to be developing increasingly sophisticated AI models that understand biology and drugs and could be used to assist pharma firms more directly in the drug discovery process. As Steph and Valida reported, OpenAI recently spoke to life sciences diagnostics vendor Revvity about licensing its data for training models, for example. Friar is no doubt familiar with older AI drug discovery firms such as Recursion that struck deals with pharmaceutical firms to give them big bounties for successful drugs identified by their tech. There aren’t many, if any, examples of such successes yet, though. OpenAI isn’t the only firm eyeing this opportunity. Its rivals Anthropic, Google DeepMind and Isomorphic Labs, an Alphabet subsidiary focusing on using AI for drug discovery, have also held discussions with early-stage biotechnology startups about data licensing or partnerships. Over the weekend, Friar in a blog post suggested OpenAI could strike a value-sharing type of arrangement in the energy and finance sectors too. (OpenAI is trying to convince financial services firms to license their data, as our article mentioned.) “Licensing IP-based agreements, and outcome-based pricing will share in the value created,” Friar said in her post. Large language models are already good at spotting patterns humans might miss. OpenAI’s models can sometimes connect dots between concepts from different fields to suggest new types of experiments involving anything from nuclear fusion to pathogen detection, as we reported last year. Scientists seem pretty gung-ho about these models, despite their many limitations and mistakes. While it may seem fantastical to think about OpenAI generating more revenue from IP licensing or royalties than from ads, Friar’s comments are a crystal clear signal of what it wants to do. The question is whether she is still talking about it after OpenAI raises the tens of billions of dollars it is currently seeking to raise from investors. Anthropic’s Relatable Cloud Cost Rise In the cloud world, computing becomes more efficient over time. But it’s not always a straight line, as Anthropic’s situation shows. While the maker of Claude Code is absolutely crushing it in terms of generating revenue, its cloud costs for running its models (i.e. inference) ended being 23% higher in 2025 than it had previously forecast, even as revenue projections stayed the same, Sri reported Wednesday. If that sounds familiar, it’s because many businesses have had similar experiences with cloud bills for general computing. As AI inference becomes a bigger line item for chief information officers, surely they’ll want such costs to be more predictable than what Anthropic went through! Apple’s AI Wearable Pin It’s never too soon for brands and businesses to be thinking about how to engage with or develop apps for AI wearables, especially now that Qianer and Wayne scooped Apple’s upcoming AI pin, which has multiple cameras and is due out in 2027. Meanwhile, OpenAI executives are openly talking about prototype devices the firm is developing. They aren’t saying what the devices are, but my colleagues have reported OpenAI is also working on a pin, glasses, earbuds and a desktop speaker. Not to be forgotten, Meta already sells smart glasses with a built-in AI assistant. As models get better at understanding people’s surroundings, presumably they’ll be helpful in navigation, translation and other tasks such as talking someone through their homework as they work on it. (As always, it’s worth watching Spike Jonze’s “Her” if only to see these kinds of wearables in action. That film is CEO Sam Altman’s North Star, as we have reported.) We can’t say what the application ecosystem around such a device would be like, but it’s increasingly likely there will actually be one. Humane and Google Glass may have just been ahead of their time!
| 
