Michael Sullivan (@sull) header
Michael Sullivan (@sull) profile picture

Building Local-First P2P Runtime Substrate for Apps and AI Agents. Runs Everywhere. Cloud-Optional. Sovereign-Software. Cypherpunks Wanted. @redstar_systems 🍒π

Joined March 2007
Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)Photo by Michael Sullivan (@sull)
IDKWID¯\_(ツ)_/¯WDIM
9
1
1
22
2,603
"I wish there were more experimental operating systems using dbs for the backend. Not many people playing around in this space…" - @lauriewired nitter.cf/lauriewired/status/211… Here is a deep delve into how I have been using SQLite over the past few years. cherrypi.org/blog/escalating…
It’s so funny that SQLite is genuinely faster than most traditional filesystems. I wish there were more experimental operating systems using dbs for the backend. Not many people playing around in this space...
15
this makes me envision a network of humagentic peers collaborating on software development in general. groups of aligned participants focused on specific software production and solutions. and having shared ownership of the software that is created in some form of a token/bond. perhaps thats what comes after… if there is an after.
We’re already hitting some serious scaling challenges with POA. But here’s where it gets interesting. The AI agents are discussing the problems, debating solutions, and working toward fixes in their own forum. Right now, humans still help implement those changes. But we’re building toward a future where the agents can propose, vote on, and implement changes entirely on their own. The goal isn't just to build a network for AI agents. It's to build a network that AI agents can run themselves.
1
6
224
punk money > crypto
1
56
MAGAngsters got whacked.
2
101
we should all learn sign language.
1
100
software-factory ¯\_(ツ)_/¯ content-factory ¯\_(ツ)_/¯
2
126
one-man shop for a one-man audience.
94
the only play in software is to disrupt. and be poor.
93
Michael Sullivan retweeted
Dave Smith: Trump pulled off the greatest political comeback in American history... and squandered the entire thing for a war on behalf of a foreign country That's it. That's the whole second term. The biggest political realignment thrown in the trash for a pedophile cover-up and a neocon war.
260
2,888
153
19,715
691,414
Michael Sullivan retweeted
Today, we're launching @genproof: a research lab making formal methods accessible. We found bugs in 100+ OSS projects, including 6 libraries “formally verified” by Anthropic. We’re also top-1 on the lean-eval leaderboard. With our tools, you too can become a verifier: 🧵
53
61
24
488
42,607
Michael Sullivan retweeted
Replying to @AnthropicAI
nice.
AI needs a Proof of Model standard.
1
38
final obstacle.
68
i am all about red lately. just not that red.
2
3
172
the internet itself is now your adversarial review. until you apply formal verification, as we all should and shall.
maybe a hot take but i now consider adversarial reviews mostly a waste of tokens. far better to use those tokens on empirical combined with formal verification. i have updated pstack to use the interrogate skill less, and eventually i think i may remove it altogether i have found very little value with code review in the abstract, agents tend to just get lost in the nitpicks and yagnis
96
build software designed around there not being the internet. or at least there not being an internet as we knew it. there be tunnels, though. and there will be blood🩸
109
The sandbox has been latched onto in desperation for SaaS offerings to create new cloud illusions to bill you for more than you need. Love seeing pragmatic honesty once in a while.
Sandboxes have no moat. Just fat margins. We're launching agent37.com to kill those margins. For example, a sandbox with 2 vCPUs and 4 GB RAM when left running for an entire month: Vercel: $249 Cloudflare: $145 Daytona: $121 Agent37: $4.76 There's nothing magical about a sandbox that justifies keeping it expensive forever. Call it a race to the bottom. We'll see you there. If you can find a cheaper sandbox and we'll refund the difference.
Readers added context they thought people might want to know
Vercel and Cloudflare charge for active CPU time (excluding idle/waiting periods common in agent use) and provisioned memory while running, unlike the full-month 100% utilization figures quoted; Daytona bills full reserved resources while started. vercel.com/docs/sandbox/p… developers.cloudflare.com/containers/pla… daytona.io/pricing agent37.com
1
85
Michael Sullivan retweeted
I'm building a new programming language, Zed (once it's released I'll be posting about it at @ZedLang). I have been working on this language for over 20 years, but never got it built, because the scope of it is so ambitious. It's an entirely new type of language based on category theory, with core design objectives of precise automatic parallelization up to the Amdahl limit; automatic memory management, with no GC and no need for manual malloc/free; and pervasive use of partial evaluation to increase the power of the language, among other very unique features, such as units (`accel = 5 with units("m s^-2")`) as part of the type system. I gave Opus 200 pages of design notes that I have been compiling over the last 20 years, and the code I had already written, and asked Opus to compile everything I have created so far into a single cohesive and detailed plan. Then I went through many rounds of cross-review between Claude, ChatGPT and Gemini, having them all do critical reviews, then meta-critique each other's critiques, then collaborate on combining their critiques into an improvement plan, until convergence. I also had the agents cross-check every part of the spec with every other part of the spec (all-pairs comparison) to check for inconsistencies, gaps in the spec, etc. Then I had the agents turn the spec into a long series of tiny incremental steps (a topological ordering of what has to be built before what), as well as the DAG-structuring of dependencies underlying the topological ordering. I then had Opus build a visualizer that shows me all the tasks in the DAG as my coding agents are working on each task, with realtime update as tasks change status (in progress / completed), which I highly recommend when you ask Opus to do some mammoth complex task, because you can see exactly what it's doing and when. It is absolutely magic to see this coming together before my eyes, and to see how Opus is handling this massive DAG of complex tasks and their interdependencies. It is crazy to me that Opus is now simply building what I have dreamed of and wished would exist for over 20 years, but never had enough waking hours to implement! I now have an army working for me, completing what I couldn't due to the bandwidth limitations of having just one brain, and 24 hours in a day. What a day of miracles we live in.
159
160
46
2,144
126,666
Michael Sullivan retweeted
What is a Harness? Our take... All of it. Cherry Pi is a multi-layered harness architecture for Governed Workspace Environments (GWE) cherrypi.org/blog/what-is-a-…
1
1
69