build good, ship to serve, rejoice always
Joined April 2021
- Tweets 2,279
- Following 747
- Followers 402
- Likes 4,375
707 Photos and videos
2h
Built a “Fable-like” two-brain loop.
Fusion (Opus 4.8 GPT-5.5 Pro Gemini 3.1 Pro) plans in a Fable spec voice and verifies. Executes via Codex. Compound-engineering loop: plan → delegate → verify → iterate.
Distilled and fused (ultraplan, fable-prompting, follow-goals, agent-loop) that shape each brain.
Pointed it at a real codebase and it surfaced a product seam I’d missed.
53
lxs retweeted
Jun 15
A rare Super Delta formation featuring the Thunderbirds and the @USNavy Blue Angels soared above @UFC fans to kick off #UFCWhiteHouse. #Freedom250
🎥 by @whitehouse
509
3,462
21,252
785,928
lxs retweeted
Jun 15
Patriotmaxxing.
Very well done by the UFC
23
406
3,366
139,020
lxs retweeted
Jun 15
USA.
Jun 14
250 years old. Rocket ship IPOs. New York is winning again. Trillionaires. American soccer looks good. European minds are visiting, comprehending.
How can you be pessimistic about the USA?
23
41
992
87,310
Jun 14
time to use those credits
Jun 13
Introducing the Fusion API, the smartest compound model in the market.
Fusion achieves Fable-level intelligence at half the price.
How it works 👇
15
Jun 12
Impressive! This is where the web gets better.
Jun 12
Ask our developer docs. They’ll show you the way
The new docs agent on 🔗developers.openai.com helps you find answers about OpenAI products and takes you directly to the relevant documentation.
1
44
lxs retweeted
Jun 12
Ask our developer docs. They’ll show you the way
The new docs agent on 🔗developers.openai.com helps you find answers about OpenAI products and takes you directly to the relevant documentation.
53
60
957
177,026
Jun 12
Gemini models are good at UX/UI yet GPT has the better super app, Codex than Antigravity. Claude’s output has more personality yet OpenAI’s team on X has far better “tweet” engagement.
🤔
50
You have Claude Fable for only a few days. Here's how to make the most of it.
Introducing /improve: use your most capable model to audit your codebase and write plans for cheaper models to execute later.
Studies your code, figures out bugs, perf, tech debt, missing tests, what to build and writes plans any agent can run.
179
383
6,268
760,478
Jun 10
Anyone want to try Fable 5 in Claude Code? Drop a comment and I'll reply to your comment with 'Give Claude, get more Claude code'
No gimmicks, just quickness as these go fast.
2
1
88
May 28
gemini live API is 🔥
i’ve been cooking this past two days and just launched an amazing sales coach for my team.
gemini 2.5 flash for rag/brain train
gemini file search
gemini live for voice to voice
as much people hate on antigravity. all built with AG 2.0
1
77
May 25
When the wife tells me that i’m spending too much time with the clankers. I agreed and now touching grass.
1
16
May 25
I’m hooked on using /goal in AntiGravity but now in need of Google releasing a Codex style mobile remote control to check in.
@_mohansolo
44
May 24
inside antigravity. working with gemini 3.5 and opus 4.6 agents optimizing my sales coaching chat agent and after rounds of RAG and prompt shaping. new benchmark is taking the model back in the past. this is why it's important to benchmark and not chase shiny new items. see below
1
55
May 24
gpt-4.1-nano is blazingly fast (1.7s) but quality drops too much (60%) — misses Brain knowledge retrieval and key guardrails like manager approval for tiered billing.
gemini-2.5-flash wins clean — best quality, middle speed, great cost.
1
56