Phil Heller

The California AI Transparency Act became operative on August 2, 2026. What Matters Today * The first phase of the California AI Transparency Act (CAITA, SB 942/AB 853) is now in effect. (calmatters.digitaldemocracy.org) * It applies to publicly accessible generative AI providers with more than one million monthly users

Libraries worth of physical books should not be fuel for industrial-level destructive scanning by AI companies. Anthropic reportedly spent tens of millions of dollars buying millions of physical books, slicing off their bindings with machines, separating their pages for high-speed scanning, and sending the remains for recycling. The

AI data chains, trusting the process and not just better models, and the case of the missing person who connects the pieces together. Every organization has its own version of a murder mystery. A project dies without one obvious cause. A customer relationship deteriorates. A number changes in a way

For a brief period in the 1990s, CD-ROMs made possible a new form of creative medium from a unique type of creative person. In the 1980s, the personal computer could be used to process words, maintain spreadsheets, organize databases, play games, do early desktop publishing, and access online services.

Netflix’s earnings call this week completes the three-act arc of one of the great media transformation stories of the century. But, it may not be the happy ending that Netflix imagined. Netflix is the Silicon Valley company that invaded Hollywood, won the battle and took the crown, and

Thinking Machines Lab just did something almost no frontier AI company does. It shipped a model and led with a disclaimer. Inkling, the company states plainly, is “not the strongest overall model available today, open or closed.” Instead, the company argues, a combination of qualities makes it a good open-

The AI industry is getting better at showing agents doing things. It is still much weaker at proving that those things were valid, useful, authorized, representative, or safe. That is the problem underneath three recent research threads: Mnemosyne, automated benchmark auditing, and METR’s Frontier Risk Report. The issue is

Business runs every day of the year, and almost all of it is weighing risk. An opportunity comes up. There’s a big new idea. Something could always go wrong. The company estimates how likely, how bad, and for whom, then decides whether the upside justifies going forward. Maybe it

Apple has sued OpenAI, its hardware unit io Products, and former Apple employees over allegations that Apple trade secrets were used to help build OpenAI’s forthcoming AI devices. The claims are unproven, and the lawsuit has only just been filed. But the strategic meaning is already larger than the

OpenAI’s GPT‑5.6 family, Sol, Terra, and Luna, is easy to describe as another model release. Sol is the flagship tier, Terra is the balanced workhorse, and Luna is the fast, economical option. Axios framed the initial release as unusually constrained because OpenAI coordinated its limited preview with