Attention

Substacknewsletter · by Illia Polosukhin · English

Attention is all you need: Illia Polosukhin's thinking on AI, blockchains, and how to build an Open Superintelligence.

Audience
Not shown publicly
We show a dash rather than guess
Observed frequency
No recent posts
Observed from indexed posts
Last post
Jul 29, 2026
Jul 29, 2026
Likely audience
AI engineers
Inferred from post content

Attention is a Substack newsletter by Illia Polosukhin covering AI governance and regulation, AI safety and infrastructure security and Open source AI and model access. Its most recent indexed post is from Jul 29, 2026.

Topics

  • AI governance and regulation
  • AI safety and infrastructure security
  • Open source AI and model access
  • Formal verification and code security
  • Competitive dynamics in AI research

Recent posts

  • Why Pacing…
  • Rethinking the…
  • Agent Harnesses…
  • The Unified…
  • Your Data…

Sponsors detected

No sponsored posts detected in the last 180 days of indexed posts.

Detected automatically from posts marked as sponsored. Detection can miss or misread placements.

See the full picture on Attention

Subscriber momentum over time, every detected sponsor with the source post, companies and people it mentions, guests, and an alert when any of it changes.

Start 7-day Pro trial

Attention alternatives: similar newsletters

Ranked by how many topics they share with Attention.

  • Homo Economicus9K subscribers

    SubstackSeveral times a week · last post 9 days ago

    Likely audience: Academic researchers

    1 shared topic

  • galadrim2K subscribers

    SubstackMonthly or less · last post 14 days ago

    Likely audience: legal and compliance leaders

    1 shared topic

  • Substacklast post Jul 17, 2026

    Likely audience: AI policy analysts

    1 shared topic

  • SubstackEvery two weeks · last post 19 days ago

    Likely audience: security analysts

    1 shared topic

All Attention alternatives →

Compiled from public pages · refreshed Nov 7, 2026 · How NewsletterFIT collects and labels data