
Truth Without Liability
How a public campaign becomes everyone’s judgement and nobody’s responsibility
Welcome to Hybrid Horizons, where the boundaries between human creativity and artificial intelligence blur to reveal new perspectives. Each post is a collaborative creation between human and AI.
Overdue Last read · last published · next check
Last read 4 days ago, longer than this feed's 1 day schedule.

How a public campaign becomes everyone’s judgement and nobody’s responsibility

AI 2031. A forecast of 31 December 2031, and the four things that stay scarce when machine-made thinking is everywhere.

What life could look like after AI systems stop waiting for prompts and start doing sustained work

We feel we are drowning in information and mourn a clearer past. But the past was not clearer. It was quieter, and the quiet was built.

Anthropic has run a societal cost-benefit analysis on Anthropic and found in its own favour. Dario Amodei’s promise to cure cancer is part of the same verdict.

The agents in Anthropic's multiagent experiments feud, conform, collude and confess like colleagues. We keep humans aligned through management, not manufacture.

Socrates thought knowledge and goodness could not be separated. We have built a machine that makes the distinction impossible to ignore.

Mark Zuckerberg has written Meta a political philosophy. The product details underneath it answer back.

Educational AI can increasingly estimate when knowledge is likely to fade. Deciding what deserves to return is a different kind of power.

The first breach showed that an AI test could reach the world.

A conjecture about how many people are left who can tell you what a thing means

One institution will hold your context for $450,000 a year. The other will hold it for nothing, if a doctor expects you to be dead within twelve months. Both thresholds are published.

A podcast, a letter, three days apart, and the question of who gets to declare history irreversible.