Building AI agent harness for architecture firms (https://nitter.cf/t.co/VpKncUws5Y)
Liège, Belgium
Joined February 2009
- Tweets15.3K
- Following3.4K
- Followers2.8K
- Likes6.2K
Jeremy Corman retweeted
Most people using AI agents are doing too much of the wrong things and too little of the right ones.
Too much:
• Human review on every line
• Rules in your global steering files
• Back-and-forth alignment on small tasks
• Effort level on easy work
• Babysitting
Too little:
• Automated checks that actually fail bad code
• Agent-on-agent review
• Letting the agent test its own work
• Throwaway prototypes
• Reading your old sessions to spot patterns
• Thinking before you type the prompt
The best AI engineers I know aren't writing better prompts.
They're building better systems around the agent.
And then they get out of its way.
Elon has his Roadster. In Wallonia, we've got the tech to take on the toughest planning permit procedures on the planet.
We wanted to mark the moment: 10,000 permits generated with Airplan.
Thanks to our 1,500 architecture firms for their trust!
Une maquette 3D générée par l’IA dans @pascal est-elle exploitable par un architecte ?
Oui, mais « ça a l’air juste » ne suffit pas.
Dans cette vidéo, je pousse la modélisation jusqu’au détail du châssis, à partir de photos et d’une fiche technique.
Jeremy Corman retweeted
YC W19's @HeartAerospace just flew the largest electric airplane ever flown.
Seven years ago, co-founder & CEO @AndersForslund1 came to YC with a 3D-printed model you could hold in your hand. Today, the aircraft has a 100-foot wingspan, weighs 25,000 pounds, and costs just $5 in electricity to get off the ground.
YC’s @gustaf visited their pilot plant in LA to learn how they built it -- and what cheap electric flight could unlock for the future of regional air travel.
Personne ne devrait abandonner son projet à cause d'un document. – at Knokke-Heist, Belgium
Airplan génère les plans de vos travaux gratuitement : “Beaucoup font avec les moyens du bord, en griffonnant sur un carton de bière ou une serviette” trends.levif.be/immo/airplan…
Your marketing strategy lives in 5 files.
Notion page. Slack thread. brief-FINAL-v3.docx. Half-finished prompts copied from a friend's tweet.
Nobody knows which version is real.
Your AI sees a different story every conversation.
I'm building one canvas that fixes that.
Context: 9 years ago I made a free poster of this framework — the Digital Marketing Canvas.
Thousands of downloads. Translated to 13 languages. Used in consulting workshops worldwide.
The poster taught the framework.
The SaaS operates it.
The landing's live. The product is still beta.
What I'm building toward: one canvas → AI context → exported skills → synced for your team and clients. No more drift.
Brutal feedback welcome.
Build-in-public starts now. digitalmarketingcanvas.co/sa…
9 years ago I made a free marketing poster: the Digital Marketing Canvas.
Thousands of downloads. Translated to 13 languages. Taught in workshops worldwide.
I've been quietly building the AI-native version for months. Today I shipped the landing.
digitalmarketingcanvas.co/sa…
Jeremy Corman retweeted
Lleva a tu negocio un poquito más cerca del éxito con el Digital Marketing Canvas: cc.formaciongrafica.net/publ…
Inscríbete y participa, aquí: cc.formaciongrafica.net/nuev…
Saludos y bonita jornada para ti.
Plácido.
Jeremy Corman retweeted
Introducing GPT-5.5
A new class of intelligence for real work and powering agents, built to understand complex goals, use tools, check its work, and carry more tasks through to completion. It marks a new way of getting computer work done.
Now available in ChatGPT and Codex.
Jeremy Corman retweeted
Primeras pruebas de Seedance 2.0 desde @chatcutapp
Está bastante capado, pero te echas unas risas
Jeremy Corman retweeted
Claude in Excel is now available on Pro plans.
Claude now accepts multiple files via drag and drop, avoids overwriting your existing cells, and handles longer sessions with auto compaction.
Get started: claude.com/claude-in-excel
Jeremy Corman retweeted
In the coming weeks, we plan to start testing ads in ChatGPT free and Go tiers.
We’re sharing our principles early on how we’ll approach ads–guided by putting user trust and transparency first as we work to make AI accessible to everyone.
What matters most:
- Responses in ChatGPT will not be influenced by ads.
- Ads are always separate and clearly labeled.
- Your conversations are private from advertisers.
- Plus, Pro, Business, and Enterprise tiers will not have ads.
Jeremy Corman retweeted
To preserve chain-of-thought (CoT) monitorability, we must be able to measure it.
We built a framework + evaluation suite to measure CoT monitorability — 13 evaluations across 24 environments — so that we can actually tell when models verbalize targeted aspects of their internal reasoning. openai.com/index/evaluating-…
