"The task is, not so much to see what no one has yet seen; but to think what nobody has yet thought, about that which everybody sees." Erwin Schrödinger, the Austrian physicist and Nobel laureate noted this in the context of his work with quantum mechanics. There’s an interesting example of this playing out in hiring right now. As candidates use AI to tailor resumes to every job description, employers are finding themselves with a strange problem: more and more applications l

Harvard Business School’s AI faculty avatars offer an early glimpse of a meaningful shift in the future of work: expertise may no longer be constrained by an expert’s physical availability. HBS Foundry is using AI versions of faculty to help entrepreneurs practice pitches, sales calls, and board meetings on demand. My mother is a teacher, so I’ve seen firsthand how scarce expert time can be. One teacher can support only so many students, in so many classrooms, during so many

What is worth doing? “Go to college, get a good job, have a good life.” That was the formula BUT increasingly, for many, the calculus has changed. The old formula worked for a lot of people. It also gave generations a relatively clear answer to a harder question: What should I do with my life? That answer is getting harder to take for granted and the details have a lot of nuance to them. The U.S. high-school graduating class is projected to shrink 13% from its 2025 peak to 20

Will Anthropic’s IPO eclipse SpaceX’s $75 billion public debut? Maybe. Maybe not. Here’s what I find FAR more interesting. Anthropic CEO Dario Amodei recently said that saying AI will cure cancer has become a cliché. “The thing that will work is actually curing cancer.” That sets a pretty high bar for what AI has to ultimately deliver. And I think we’re starting to see what it takes. Look at biology: 1\ Google Deepmind's AlphaFold helps make biology more legible. 2\ Insilico

Enterprise AI ≠ models, use cases, pilots. It’s a new way of working, embedded into the business. Ode with Anthropic, the enterprise AI transformation company backed by Anthropic, Blackstone, Hellman & Friedman and other major investors, has acquired AI services firm Casper Studios. The strategic fit is straightforward: Ode goes deep -> It builds custom, production-grade AI applications for an organization’s hardest problems. ( + ) Casper goes broad -> It helps companies depl

What if some of the most important technologies of the next decade are already being built in places most of us never think about? The July/August issue of MIT Technology Review is a great reminder of just how much is happening beneath the surface of the technology we use every day. In Assam (India), a machine listens for elephants and plays the sound of a leopard when one gets too close to a village. In the Netherlands, ASML is building $400M machines at the centre of the gl

Google reportedly agreed to pay $10 million for a de-identified package of Spirit Airlines’ internal business data in the airline’s bankruptcy auction. The package reportedly includes about 100 million emails, 500 million Microsoft Teams messages, roughly 30 million lines of code, and years of pricing, revenue, marketing, operational, and other business records. Google has said the dataset will be rigorously scrubbed of personally identifiable information by a third party bef

Google DeepMind’s GNoME generated 2.2 million crystal structures in its search for new materials. The system identified 381,000 as predicted new, stable materials, and researchers found 736 that had been independently realized experimentally. 41 of those have been synthesized by an autonomous robotic laboratory. That’s 800 years of science, compressed. We are seeing the same pattern in cognitive work. The cost of producing cognitive work is falling rapidly. Software, analysis

Anthropic is reportedly in talks to acquire AI startup Decart for approximately $6 billion. Decart is an interesting company because it spans several layers of AI, including training and inference optimization, real-time generative video, and an interactive "world model". Decart solves a practical bottleneck for Anthropic: inference and the economics of AI at scale. For me, though, the fascinating part is the world-model piece. A world model is an AI system that builds a repr