Notes fromthe work
Longer thinking from the Atlas team: what we are learning as we scope risks, de-risk interventions, and hand them to the people who will run them. Everything here is also published on Substack.
- 15 Sept 2026 Atlas Computing is now Atlas Ignota A new name for a new chapter.
- 8 Sept 2026 Applications Open for Atlas’s Incubator on AI Infrastructure Security Now hiring 10 Founders-in-Residence — plus Field Strategists — to secure AI infrastructure against nation-state-level threats.
- 25 Apr 2026 Your Solution Doesn't Know Your Problem Exists If your problem is structural, your approach must be radical.
- 24 Dec 2025 CSLib: Lean’s Formal Software Foundation Bottom line up front: If you love Lean and care about software, you’re likely to be excited about the progress on CSLib and might be interested in contributing to one of these two tracks present below.
- 16 Dec 2025 An alternative to "fund people not projects" Our catechism for creating impact-maximizing organizations places finding a founder *last*
- 31 Oct 2025 Post-FMxAI 2025 newsletter Takeaways from Formal Methods x AI conference 2025 @ SRI, Menlo Park: atlascomputing.org/fmai25
- 17 Oct 2025 Civilization's maintenance backlog A few dozen organization-shaped holes to be filled before powerful AI arrives
- 2 Sept 2025 Website updated! to match our focus on mapping and addressing neglected catastrophic risks
- 12 Jul 2025 Daniel Windham: Passing the Torch Reflections on our work as I transition out of Atlas
- 3 Jun 2025 A refinement-based paradigm for code generation Consider this to be an extended answer to the question “why would you build the IDE for specification” described in our previous blogpost. The questions in this document are a modified+truncated Heilmeier Catechism that were part of a (rejected) proposal, but I wanted to share it as a public artifact.
- 22 May 2025 IDE for validating specifications A reminder of our overarching vision
- 22 Apr 2025 Progress in autoformalization experiments Can today's AI systems generate formally verified code?
- 1 Apr 2025 Govern AI with Rules, Not Values A Vision of Specification-Driven AI
- 6 Aug 2024 Our Report On AI-Enabled Tools For Scaling Formal Verification Scaling Human Understanding and Review Capacity
- 31 Jul 2024 Announcing Flexible Hardware-Enabled Governors Imagine if every uranium atom pulled out of the ground had its own international atomic energy inspector tasked to follow it, report on its usage, perhaps track its location.
- 7 May 2024 Retrospective on Mathematical Boundaries Workshop Primarily written by Evan Miyazono (with help from Manuel Baltieri and others) - mistakes my own
- 9 Apr 2024 Evan's thoughts on boundaries (Apr 2024) AI Guardrails may benefit from formalizing the concept of boundaries
- 4 Apr 2024 Welcome to Atlas Computing's blog! A brief summary of where we write about our work
GetInvolved
Atlas is a non-profit organization. Send a risk, bring your skills, or fund the work.
Fund anExploration
Some funders come to Atlas with a concern: cyber-resilience, information integrity, or another AI risk no one clearly owns. Atlas maps concern to a concrete intervention: scoped, tested, and ready for the right owner.
Bring YourSkills
Apply to join Atlas full-time as a Field Strategist or support an existing effort. Atlas needs people who can scope problems, pressure-test plans, and convene the stakeholders who make a fix real.
Send aNeglectedAI Risk
Tell us about a society-scale AI risk where the owner is missing. What’s the clearly desirable outcome that needs Atlas to find a path forward?