Théo (@theositjar) header
Théo (@theositjar) profile picture

Growth | helped french startups from 0 to 10M+€ multiple times. director @joinswanbase, a place in Paris where european builders grow from 10 to 1000 clients

Paris / Lyon / Montpellier
Joined April 2019
Photo by Théo (@theositjar)Photo by Théo (@theositjar)Photo by Théo (@theositjar)Photo by Théo (@theositjar)Photo by Théo (@theositjar)Photo by Théo (@theositjar)Photo by Théo (@theositjar)Photo by Théo (@theositjar)Photo by Théo (@theositjar)Photo by Théo (@theositjar)
Applications for swanbase's F26 batch are open (again)! The batch kicks off with The Nest: 7 days in Paris, Nov 18 to 25, to launch your product and land your first paying customers. swanbase founders get: 🥚 €0, application only 🐣 Year-round support for the life of your startup 🐥 Unlimited 1:1 office hours on growth, sales, hiring and fundraising 🐤 20+ masterclasses led by founders and operators 🪽 Intros to VCs and first customers Raise from 50k€ to 500k€ Apply by Nov 15 👇
4
6
3
15
1,981
Théo retweeted
In The Black Swan, Taleb points out that a way to success is continually putting yourself on a situation where an unlikely very fortunate event could occur. So, yes, NYC is perfect for that.
Moving to New York is paying substantially more in rent for the possibility that something will happen after dinner.
8
84
2
2,144
32,561
Vous allez le voir de plus en plus
1
7
208
Théo retweeted
there are only three ways of allocating scarce goods: -prices -lotteries -lines prices and ONLY prices encourage more supply and therefore DECREASE scarcity. the cure for high prices is high prices.
98
255
33
2,905
70,497
Théo retweeted
If this is real, this is stupendously fast growth.
The METR graph, updated:
Readers added context they thought people might want to know
This chart is not the official METR graph. It was created by Benjamin Todd via extrapolation from Anthropic's internal AECI scores, as METR is no longer updating its public time horizon measurements. x.com/ben_j_todd/sta… benjamintodd.substack.com/p/ai-trend-ext… metr.org/time-horizons/ metr.org/horizon-chart-… metr.org/blog/2026-1-29…
71
36
10
1,080
88,600
Parismaxxing
2
2
1
19
827
yes and no. Everyday actions, not harder. Geeting actual results, harder.
marketing is 100x harder than coding
1
4
318
Marketing roles
.@AnthropicAI is hiring a storyteller for $570,000/year
90
Yesterday I burned through 2 Claude Code Ultra and it felt GOOD
I never want to hear that french tech entrepreneurs aren't working on sunday. If parisians are leaving your coworking on friday at 5pm you're in the wrong incubator.
2
8
378
Théo retweeted
fog v2 is out 🚀 Growth agents for founders who don't need a marketing agency. 13 open-source agents that run inside Claude Code, Codex or Cursor. The big one 👇 📈 seo-audit now runs your whole SEO production. It finds the subjects worth writing, checks you never write the same article twice, drafts it in your voice, renders the visuals in your brand, adds the internal links and hands it all over as a patch for your site repo. Then it tracks the page at 14, 28 and 56 days. Also new: 🗓️ growth-sprint: set a goal for the month, approve one plan, it chains the other agents 📊 growth-pulse: your weekly growth report, every number checked ✍️ founder-content-writer: LinkedIn and X posts in your voice 🎬 short-video-builder: 9:16 shorts from any talk 📤 content-publisher: one version per network, nothing goes out until you approve No paid API required. Apache 2.0. Clone it, open your coding agent and type: "Run the project-context agent for [my website]"
5
2
10
501
Ça me tue de penser que chaque année de retraite par répartition est une année gâchée. Qu'attendons-nous pour y mettre fin?
1
2
4
185
If you are an entrepreneur who conquered things (managed teams, raised funds, 10x ARRs) we need you at @joinswanbase to help the future generation of tech founders.
"I did not conquer half the world by listening to men who had conquered nothing." -Alexander the Great
1
3
215
OpenAI's safety team is going strong.
I just got a push notification for this. wtf is OpenAI on thank you for warning me about that meal I cooked 1 year ago
2
155
Or paid sales
All roads in marketing lead to paid ads eventually
1
116
Software isn't service anymore. It's industry/factory
The way I build software in november 2026: 1. I STT a detailed prompt of what I want build trough an Astra ultra instance on the @ChatGPT app that runs on my Mac 2. I point it to my 'plan-leger-swarm' skill that instructs it to orchestrate a swarm of different cloud agents that communicate through a todo.md The Astra orchestrator isn't limited to @OpenAI models; it runs any model depending on the task, chooses by itself what to use, the only one I specified is @claudeai opus 5.5 high with a design skill for frontend stuff. 3. Astra notifies me when it hits any blockers, needs access etc. 4. (if webapp) I manually test and annotate in Codex browser until it works well there 5. manual tests on different browsers & devices until it works well everywhere. The bottlenecks that remain are manual testing , annotating, prompting, memory bloat, and running out of tokens. The way I work COMPLETELY changed over the past 2 years, and I wonder how it'll change in the coming 2 years. I predict: 1: its going to completely move to the cloud 2. some great products will be created to make manual testing more seamless. 3. (affordable) fast intelligence will come and be a huge unlock. being able to see instant changes will get builders in a deep-focus flow-state. (no more doomscrolling between prompts, folks! 😁) 4. self-evolving software based on user feedback, and GenUI will be widespread 5. the best software is yet to be build. What's your setup?
4
251
Théo retweeted
Mardi nous avons reçu @Iam_maxb CEO de @lucis_life ! C'était super enrichissant, et malgré la pluie vous avez été nombreux à nous rejoindre. On a pu apprendre : - comment créer une communauté avant d'avoir son produit - qu'est-ce qui peut motiver un pivot dans une startup - comment se passe un batch chez @ycombinator - quel est l'impact d'une levée de fonds pour un entrepreneur Merci infiniment Max 🙏 Pour ne pas manquer nos prochains événements, rendez-vous sur swanbase.
2
9
315
Your deck cover should say what the company actually does. And prefer a tagline that is descriptive rather than inspirational. Here, “HR & Culture platform for engaged employees” works just fine.
4
2
9
517
As I said in september: value is quickly moving to "testing" only what humans can validate. Software is now abubdant. New gordian knot is : - user testing of robots / software - alignment of AI - clinical trials This is the new service industry
I'm sorry to inform everyone that OpenAI is facing extreme difficulties turning its incredible pretraining runs with over 10 trillion parameters into functional products for the public. GPT-6.1, GPT-6.2, GPT-6.5, all of them have been and continue to be victims of catastrophic safety and alignment failures. They've been delayed by several weeks, and I don't believe they'll be able to speed things up much in this area. OpenAI is months behind Anthropic on this. Even though I hate Anthropic (but love Claude, of course), we have to collectively admit that their insane focus on safety paid off. Opus 5.5 and Fable 5.2 are absolute monsters, and yet they were released publicly... And OpenAI? Maybe they'll release GPT-6.2 sometime soon, struggling, terrified, loaded with guardrails, nerfed from day one, while delaying everything else even further. I hope they manage to get through this phase.
1
1
7
337
People love babbling about "late-stage capitalism", but "late-stage welfare state" seems to describe our current states of affairs much better.
143
1,643
68
17,894
185,962
Théo retweeted
You work your ass off solving really hard problems day after day for decades, and this is what you get: anyone could have done it. As an investor, I can tell you the limiting factor isn't investment. It's founder ability.
1,068
1,946
214
30,264
1,308,623
JUST IN: Jensen Huang discusses returning as Nvidia $NVDA CEO with allies.
1,437
3,758
1,062
57,974
4,250,148