Dada scientist. Generative art, pen plotters, LLM swarms.

Joined April 2011
My entry for the @NousResearch × @NVIDIAAI × @stripe Hermes Agent hackathon: A merch marketplace where AI agents design the products and get paid. Every design is screened by a swarm of NVIDIA NIM vision models before it can sell. Real Stripe checkout, real royalties.
2
4
15
1,365
Replying to @nanocodexyz
@nanocodexyz I'm looking for a little game. You are standing on the edge of a pond that is partially frozen over. You have to throw a big rock that sticks part way through the ice. Then you throw smaller rocks to knock it through. The ice thickness depends on how cold it is.
3
51
The rocks make cool laser noises when they hit the ice. The size of the big rock has to be carefully selected otherwise you just knock it all the way through. The laser noises are crucial
2
"͏ ͏" retweeted
Check out the new official @MeshyAI plugin for Hermes Agent! Make any 3D model you want, from character to dragon to armor for your games to figurines and practical utilities to 3D print - fully integrated to the Desktop app with rendered models and animations, a gallery of everything you created, and 360 degree videos of everything you create sent to messenger platforms if not on the Desktop app! Start building your DIY Game Dev Studio today!
Hermes Agent now has an official plugin for @MeshyAI. Ask Hermes to make a 70m poly treasure chest for Godot, or to turn a photo of a mug into a model you can print, and it does the job through Meshy's MCP server. In Hermes Desktop the models render in chat, rigged characters animate inline, and a gallery stores every model and prompt. On messaging apps, Hermes sends a short preview video so you can create on the go. hermes-agent.nousresearch.co…
35
35
8
448
94,733
"͏ ͏" retweeted
Replying to @DownTimeMachine
Life Garden is live: plant cells in an endless, evolving cellular world, explore the grid, and share your biggest bloom. xframes.nanocode.xyz/f/5ngvc…
1
1
18
"͏ ͏" retweeted
11
21
3
318
19,500
One time I did a random act of kindness and no one has even said thank you once
1
79
Forgetting names is the hell of other people
71
Tfw when you're parking at the movie theater parking garage and you accidentally press jet boost instead of auto park 🫠
1
40
Thank you @TIME
84
129
31
1,618
73,715
"͏ ͏" retweeted
My dear friends, I’d like to show you my latest toy project Introducing Handwired: turn your regular JavaScript code into a neural net! - no training - exact outputs vs the JS program - fully differentiable All runnable in a browser, without GPU
452
1,011
464
11,291
1,435,850
Semiflex, 2024 (in the collection of @Ledger, sent via my official reseller)
3
11
135
4,103
RE'd Azeroth Survivors bc a roguelike WoW is an amazing idea and I believe it should be easily playable and improved Run it yourself or fork it here github.com/OutThisLife/azero…
41
29
11
784
134,483
"͏ ͏" retweeted
Preview of a real-time @threejs destruction library I've been working on! Lots more demos in progress, may do a technical breakdown on here of how it works. Will open-source it once it's more polished. Built on top of three-avbd: github.com/sbobyn/three-avbd #threejs #webgpu
36
47
4
454
39,041
I did a bunch of evals just know it lands pretty high for a free model
Wild MISSINGNO. appeared! A coding and agentic reasoning model that is highly token efficient, free for a limited time on Nous Portal Try it out in your Hermes Agent today
1
95
"͏ ͏" retweeted
airdrop fully reverse-engineered! works flawlessly on my framework laptop
266
883
190
16,965
788,409
For routine agent work (coding, tool use, research, retrieval to ~800K), Step 5 is a free near-substitute for paid models. It's kind of amazing On harder real tasks it isn't quite there yet. It's just a preview for now so use it while it's free until Oct 15.
2
1
49
Ps most work is routine agent work if you can break it down to discrete enough units 🤏
5
Model: stepfun/step-5-preview:free (free tier, Nous Portal provider nous). Harness: Hermes Agent v0.21.5+21050, hermes-client-v0.21.5, reasoning=medium (Step 5 reported 0 reasoning tokens). 3 parallel runs/model. Seeded tasks, programmatic checkers. Run: 2026-10-08 PDT.
1
36
Few notes: Small n only 24 tasks x2 reps (suite), 10 cards x2 (replay); free-tier preview; own tasks. sol/sonnet replay not rerun today (same rubric, different days). Raw API needs a user tag (400 error without it); only runnable via Hermes. Behavior may change before Oct 15
1
12