Humane Tech Layoffs Signal Shift in AI Ethics Funding

The story of the day

The Center for Humane Technology, a long-time critic of attention economy excesses, is laying off most of its staff to go "founder-led," according to Wired. The move suggests a pivot in AI ethics funding: donors may now prefer lean, direct-action groups over broad advocacy. For builders, this means less institutional pressure on responsible AI—but also fewer resources for ethical frameworks. The irony: as AI adoption accelerates, the organizations pushing for guardrails are scaling back.

New tools & launches

ClawCall — AI agent that dials, holds, and reports back phone interactions. [Product Hunt]

Tonefold — Generate editable MIDI from song descriptions. [Product Hunt]

pmtui — Terminal-based autopilot for long-running AI sessions. Open source. [Product Hunt]

Claude for Google Workspace — Integrates Claude into Docs, Sheets, and Slides. [Product Hunt]

Models & research

Researchers have already improved on OpenAI’s recent math proofs, with verified results in Lean. The work, shared on GitHub, demonstrates how open collaboration can outpace closed research—despite claims that "slop proofs" offer no value. Open-weight, community-driven.

Epoch AI’s latest analysis shows LLMs still lag far behind human researchers in innovation. The gap isn’t just technical; it’s conceptual. Proprietary models remain the bottleneck.

Industry pulse

Australia plans to require AI companies to prove their safety systems work, per a government outline. The move mirrors EU-style compliance but targets operational validation over theoretical risk. Builders take note: expect more audits, not just disclosures.

Stat of the day

The Center for Humane Technology, once a darling of tech ethics funding, is reducing its team to a skeleton crew. Wired reports internal tensions over strategy as the culprit.

Get the full analysis — subscribe to Gradient Pro.