Industry

First state law barring AI from independent therapeutic decision-making, with fines up to $10,000 per violation. This is the policy component of your knowledge infrastructure argument actually moving in real time, not a citation about AI but a citable regulatory fact.

Gates argues the AI transition will rank among the most turbulent periods in human history and that governments, industry, and communities are nowhere near ready. His risks are labor displacement, empowerment of bad actors, and the erosion of social and cognitive development; his remedy is governance, meaning new institutions, protected categories of human work, and a tax on AI. Useful here less for its argument than for its author and reach: the clearest high-profile signal that the shock is felt at the top of the industry, and a map of the policy scramble the knowledge-infrastructure argument anticipates. Where it parts company is that it never reaches the document. Gates frames the entire event in the register of scale and consequence, the volume-and-harm story Information Shock concedes to the previous shock rather than treats as the new one. Even the passage on weakening critical thinking, the one place he brushes the cognitive turn, stops at a harm to be regulated and never reaches mediation. Best read as the discourse naming the size of the event without locating the break, and as an interested source: Gates's ties to Microsoft and OpenAI sit behind the call for rules. The cognition point is covered more sharply elsewhere in this collection (the NYT "Making Us Dumber" piece and Rundell).

Brilliant piece that shows the tradition of documented centered information science and how AI is breaking it. This feeds perfectly into the Information Shock we are currently expereincing.

The AI induced information shock is based on behavior, not an endorsement. Findings such as these are stark reminders that generative AI must be shaped and built with the public good in mind.