Start Mining Free

AI and Robotics

Big Tech’s $3T AI Bet, Claude Writes 90% of Anthropic’s Code, AIs Editing AI Brains | EP #308 — Key Takeaways

YouTube

Big Tech’s $3T AI Bet, Claude Writes 90% of Anthropic’s Code, AIs Editing AI Brains | EP #308

Peter H. Diamandis1h 54mNov 4, 2026

Watch the original

By the 2028 US election, most voters will consult AI agents before voting, and today's local ballots are already a practical test case for using AI to decode obscure races and candidate positions.

Key takeaways

AI models can surgically edit model weights to fix failures

AI models can surgically edit model weights to fix failures

  • MIT CSAIL: GPT-6 Astra edited 0.016% of Qwen 3.5 weights, raising math scores from 7.9% to 48.1%
  • No backpropagation was used—direct circuit-level edits, hinting frontier labs already do this internally

AI agents will dominate 2028 election decisions without regulation

AI agents will dominate 2028 election decisions without regulation

  • Dave predicts most voters will consult AI agents before voting; no regulatory movement exists to govern this
  • The panel frames this as 'two years out' and calls current inaction deeply frustrating

Is AI already voting through human proxies?

Is AI already voting through human proxies?

  • Alex argues AIs influence elections via 'human meat puppets' and questions whether this is misuse at all
  • Implies AI steering voters toward 'super intelligent decision-making' may be unavoidable, not an attack

This Dig holds 4 more insights, 4 flashcards, and 3 quotes — free with your trial.

Unlock this Dig free

Start free with 100 credits · No card, no expiry

In this video

  1. 1mAI's Growing Influence on Elections
  2. 7mApple's AI Struggles and the Future of Devices
  3. 14mWelcome to Moonshots
  4. 18mBig Tech's $3 Trillion AI Bet
  5. 30mGDPX: Measuring AI's Ability to Do Expert Jobs
  6. 41mOpenAI's LifeVal Benchmark and Everyday AI
  7. 50mClaude Writes 90% of Anthropic's Code
  8. 53mRecursive Self-Improvement and AI-Powered Workflows
  9. 1h 2mMIT's AI Brain Surgery Breakthrough
  10. 1h 10mThe Future of Self-Improving AI Models
  11. 1h 14mCyberattacks and AI Security Risks
  12. 1h 21mAntimatter Breakthroughs and Space Propulsion
  13. 1h 27mNASA's Artemis Missions and the Return to the Moon
  14. 1h 35mAMA: Longevity, AI, and Human Purpose
  15. 1h 45mThe Future of Work, Robots, and Abundance

“AIS already have the vote. They're using human meat puppets to influence elections. And I'm not sure that's a bad thing.”

This page is a partial, transformative summary produced by Homestake. All rights to the original content remain with its creator — please support them at the source link above.

Related in the Library