You don't need a Ferrari team. You need a Toyota Camry team.
A CTO asked me on a feedback call whether the two engineers on his pilot were our A team.
I told him no. They're a Camry team.
Normal, good engineers, with the fundamentals in place before they joined us.
No brilliant hire we parachute in to win the account and pull out after signature.
Then he asked the question every buyer should ask a vendor: is your process repeatable enough that you have five more teams this good?
Yes.
That's the point of a Camry team.
A Ferrari team is one genius. Expensive, fast on a good day, impossible to replace, and the day he leaves, your velocity leaves with him.
A Camry team runs on the process instead of the person.
Eight o'clock every morning: what changed, what improved, who owns it. The same rhythm a plant manager runs on a factory floor.
Prompting is the new coding. Owning the spec end to end is the job. We have uncomfortable conversations with every engineer about that, and we have them daily.
That's harder than hiring a genius. It's also the only version that scales.
He'd run a parallel pilot with a firm ten times our size. Four weeks in, he said the Camry team was doing the better job.
Stop asking vendors for their A team. Ask what happens the day the A team leaves.
Then ask for one month to prove it. Ours costs you nothing if you don't find value in it.
Fable 5.1 just refactored my entire codebase in one call.
320 minutes. 55,000+ new lines. 34 brand new files.
It modularized everything. Broke up monoliths.
Cleaned up spaghetti. None of it worked.
But boy was it beautiful.
🌊 That drop is not a waterfall. It is a CLIFF eating the mountain.
Brown wall goes over the edge and turns into a smoke cloud. Rasuwa, Nepal.
📽 AccuWeather / @abcdgfmgrrr
The most important type of intelligence assessment is telling the difference between a surface appearance of intellectualism, and the actual capability to understand complex relationships between cause and effect.
Because it's easy to spot when someone is dumb.
What's harder to discern is when someone is pretending to punch above his weight class.
Go watch a recorded speech given by Barrack Obama, or Christopher Hitchens. If you don't listen carefully, you could easily be fooled into thinking these men were intelligent.
They speak in ways that sound like how smart people speak.
But, for a moment, imagine you were to take the ideas in their eloquent speeches, and restate them using the phraseology and vocabulary of a stereotypical construction worker... would either of them still appear intelligent?
No. They would not. The ideas, stripped bare of rhetorical frills and fancy presentation, reveal themselves as simple restatements of bumper-sticker level political and social commentary.
You see, plenty of people are clever enough to watch the habits of genius and imitate them.
But geniuses are not geniuses because they are eloquent. They are eloquent because they are geniuses. Eloquence, and other such habits, are a side effect of genius, not its essential nature.
The side effects of genius can be imitated with study. The core functions cannot. Because anyone who can is, by definition, a genius.
And now, for those of you who have been more patient with me than I deserve, and are still following along, we have at last arrived at a general principle we can use.
If you want to know how truly intelligent someone is, take him out of his comfort zone.
Make him do something new, learn something he's never learned, try a skill he's never practiced, solve a problem he has never faced before.
How does he adapt? How does he think? How does he learn?
Without his thousands of hours of practice, without all his familiar tools neatly stashed in his comfort zone, a man is forced to fall back on who he is, on the genetic recipe that makes him.
This doesn't mean that he will make no mistakes, or that he will magically know what he could only learn from experience. But it means that he will learn quickly, understand each mistake as soon as he makes it, and never make the same error twice.
This is why we prize genius.
Not because it speaks pretty words or covers whiteboards in complex mathematically formulae, but because genius learns.
And then genius teaches.
Genius captures knowledge from the world for the rest of humanity.
So if you want to spot it, ignore the habits, ignore the frills and tells and tics, and watch how it engages with the unfamiliar.
Matching the payload capacity of ONE Starship V3 flight would take 200+ Falcon 1 launches
That’s hundreds of rockets, hundreds of separate launches and an enormous amount of time just to move what one Starship is designed to carry
Orbital AI at serious scale needs a vehicle like Starship
And SpaceX now has exactly that
Starship V3 is designed to carry 100+ tons to orbit per flight with full reusability
That means:
• Huge batches of compute satellites launched at once
• Massive power and networking hardware deployed in a single mission
• Starlink connectivity + laser links for moving massive amounts of high-bandwidth data across the orbital network and back to Earth
• Rapid reuse for continuous scaling
• Direct access to enormous amounts of solar energy in orbit
• Orbital infrastructure expanding launch after launch
This is where Starship changes everything
Orbital AI stops being limited by how little mass we can get to space and starts becoming infrastructure humanity can build at enormous scale
And it’s happening much sooner than most people realize
यो दृश्यले मन नै चिरिएको छ… रसुवाको मैलुङमा बाढी–पहिरोले मच्चाएको यो विनाश देख्दा सारै दुःख लागेको छ । 🙏😭
कति परिवारले आफ्ना प्रियजन गुमाए होला��्, कति घर–आँगन उजाडिए होलान्… हे भगवान्! 😭💔
सबै जना सुरक्षित रहनुहोस्। उद्धार तथा राहतमा खटिनुभएका सम्पूर्ण सुरक्षाकर्मी र उद्धारकर्मीहरूलाई शक्ति मिलोस्। 🙏🇳🇵
The number that matters here isn't $100B. It's "thousands of flights a year."
This weekend, one weather system pushed the most ambitious lunar mission ever assembled into 2027. That's what single-site, single-window spaceflight costs.
A second Starship spaceport isn't expansion — it's redundancy. Elon says the goal is making spaceflight as routine as flying on an airplane. Routine is the whole ballgame: a civilization stops being hostage to a forecast the day a scrub costs it a Tuesday instead of a year.
Beautiful view of FORTE now it has arrived at Christmas Island. It is preparing to partially submerge, have Ship 40 floated on top, and will rise again to come back to Texas. Bring her home! 🫶
Big thanks again to Chris Bray out there on site for us!
https://t.co/H4M9h14Wh0