The Future Worth Building Is Human – Thinking Machines Lab

5 hours ago

Thinking Machines aims to develop AI that enhances human will and judgment, ensuring AI serves human purposes shaped by continuous feedback and learning.
Current AI often lacks customization from users; the vision is distributed, diverse AI that learns alongside people, not centralized, frozen models.
Key technical directions include training strong, multimodal models; tools for customization (e.g., training model weights); interactive interfaces for live collaboration; and publishing research for broader understanding.
Knowledge is decentralized and tacit, held by individuals (e.g., chefs, shopkeepers); AI should integrate with this distributed knowledge rather than centralizing or replacing it.
Human participation is essential, requiring technical advances like live multimodal interaction models and evaluation metrics for human-AI collaboration, not just autonomous task completion.
Decentralized alignment is crucial to avoid concentrated power; AI values should reflect diverse users via model weight customization, promoting safety and ecosystem diversity.
The future should balance human and AI contributions, avoiding trade-offs between participation and capability, and enabling organizations to tailor AI to their unique strengths.

Hasty Briefsbeta