Co-founder of Dynamik Orbits - Entrepreneur and aviator

Joined January 2012
98 Photos and videos
My 3 yo was born in France but he is struggling with the R too. Cannot pronounce his last name 🤣
フランス語のRの発音がむずくて脳が融ける🫠
9
His legacy still inspires many today
On June 13, 1777, a 19-year-old French teenager landed on a beach in South Carolina, uninvited, to fight in someone else's war. He would become one of the most important men in American history. The Marquis de Lafayette was one of the richest young aristocrats in France. He had a beautiful wife, a fortune, and zero reason to risk any of it. But he believed in the American cause so fiercely that when the French king forbade him from going, Lafayette bought his own ship and sailed anyway. He literally went AWOL from a life of luxury to bleed for a country that didn't exist yet. Congress was annoyed at first. Another foreign officer looking for a paycheck? Then Lafayette offered to serve for free and pay his own way. That got their attention. He met Washington and the two formed one of the great father-son bonds in American history. Washington had no biological children. Lafayette named his only son George Washington Lafayette. He took a bullet in the leg at Brandywine and kept rallying the retreat. He was instrumental at Yorktown, the battle that won the war. He went home a hero on two continents. A foreign teenager believed in America before America did. 249 years ago today.
2
50
L'erreur de raisonnement se situe là: "exactement ce que l’on essaie d’éviter"
Il n’y a pas de “refroidissement gratuit” dans l’espace, au contraire, c’est probablement l’environnement le plus contraignant pour évacuer de la chaleur. On n’y dispose que du rayonnement thermique, dont la densité de flux reste très faible aux températures d’un chip. À ces niveaux de puissance, dissiper efficacement devient un problème de surface radiative, donc de taille et de masse… exactement ce que l’on essaie d’éviter. Je sais que ça va à contre-courant de la hype actuelle, mais les data centers dans l’espace restent, à ce stade, une très mauvaise idée, à la fois techniquement et économiquement.
38
Je ne sais pas s'il faut rire ou pleurer: "Par ailleurs,@DarioAmodei, si vous et vos équipes voulez développer Anthropic en Europe, dans un cadre qui respecte votre liberté entrepreneuriale"
La décision de Washington de couper l’accès aux modèles les plus puissants d’Anthropic doit nous réveiller. Dans la course à l’intelligence artificielle, une nation qui dépend des autres pour sa technologie est une nation qu’on peut débrancher du jour au lendemain. Il faut faire avec l’IA ce que nous avons fait avec le nucléaire : il faut la penser comme une part de notre souveraineté. La maîtriser ou la subir : il n’y a pas d’autre voie. La France possède des atouts uniques en Europe : une électricité nucléaire, décarbonée et souveraine, des ingénieurs parmi les meilleurs du monde, mais aussi des entreprises comme Mistral, OVHcloud, Scaleway et ChapsVision, capables de rivaliser avec les géants américains. Encore faut-il en finir avec la naïveté et décider, enfin, de réarmer notre puissance technologique. Trois décisions à prendre pour faire naître nos champions : • Réorienter la commande publique vers des solutions souveraines • Structurer notre écosystème et bâtir nos infrastructures • Libérer les chercheurs et les entrepreneurs des réglementations qui les étouffent Je ne veux pas d’une France vassalisée. Je veux une France qui reste maîtresse de son destin. Par ailleurs, @DarioAmodei, si vous et vos équipes voulez développer Anthropic en Europe, dans un cadre qui respecte votre liberté entrepreneuriale, vous êtes les bienvenus à Paris. Nous proposerons à tous les Français experts de l’IA qui travaillent dans la tech, chez Anthropic comme ailleurs, un programme très attractif de retour au pays.
49
Your wish has been granted
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
29
“You can be wrong, just not confident and wrong”
Today, Neil deGrasse Tyson can go shit in his hat: "The delusion is thinking that SpaceX is going to lead the space frontier. That's just not going to happen, and it’s not going to happen for three really good reasons: One, it is very expensive. Two, it is very dangerous to do it first. Three, there is essentially no return on that investment."
48
Matthieu Richard retweeted
Jun 12
Liftoff! First $SPCX trade complete 🚀
3,735
6,097
71,855
3,668,724
"1M inhabitants on Mars by 2035-2040"
Key Gwynne insights from this… - no top number for Starlink sats. (Despite what others might say) - capacity constrained in several key markets. - flight 15 may be first Florida launch. - we will not sell compute that we need for our internal AI. - there will be likely more M&A particularly in AI (in general and SpaceX) - Elon the best CEO now maybe ever. - 1M inhabitants on Mars by 2035-2040.
1
47
Matthieu Richard retweeted
Time to celebrate SpaceX IPO with some s'mares. Those are Mars s'mores, made with Starship-flamed Marsmallows!
Replying to @whoisheartbreak
Also had a Marsmallows booth with Starship flaming
2
1
468
Matthieu Richard retweeted
fitting I think that SpaceXai is itself a multi-stage rocketship
2
3
103
6,653
Facing what he describes as the dominant source of all future military and economic power ("like nuclear weapons, but potentially even more so"), the remedies are: an FAA-style agency, wage insurance (so we can expand the share of bullshit jobs), and a very post-war coalition of nations sharing values (whose values?). I don't expect him to solve the deeper political crisis the West is facing. But man, reading this, you realize there isn't even the beginning of a solution to the problem. The problem isn't that Treebeard is slow. It's that the forest no longer follows him (at least in part of the West).
Today I'm publishing a new essay, Policy on the AI Exponential. AI is progressing extremely fast—much faster than the policy process was built to handle. The essay lays out where I think the technology is now, and the action needed to close the gap: darioamodei.com/post/policy-…
54
Future execution = fundamentals in Musk companies
$SPCX - SPACEX IPO VALUATION BUILT ON HYPE, SAYS CHANOS Short seller Jim Chanos argues SpaceX’s planned $1.75 trillion IPO valuation is disconnected from fundamentals and driven by overly optimistic future expectations. Speaking ahead of the record-breaking $75 billion listing, he said investors are pricing in ambitious narratives rather than realistic earnings prospects, adding that SpaceX’s valuation multiple far exceeds that of Tesla.
1
81
for better or for worse
21
Even if the physics is hard, it yields to engineering and iteration. Building anything on Earth however is becoming harder everyday and in many cases impossible. So it’s not that space is ideal, but it’s the only way to achieve those goals at scale.
I’m old enough to remember a few months ago when the timeline was filled with “data centers in space don’t work because physics” takes
49
We've been building this for months with @julienvm at @dynamikorbits: a full interactive model of the SpaceX business, ahead of the IPO. The most valuable part for me was being forced to understand how this company actually works. What I learned 🧵
We finally released our SpaceX model! 🚀🛰️ This helps calculate the estimated company value and share price based on your assumptions.
1
1
3
978
9/ Nobody can forecast these cash flows ten years out. You triangulate: DCF, Monte Carlo, multiples. What you're really pricing is execution, the proven ability to break the bottleneck, again and again.
1
1
28