Alex Carten
@AlexCarten_ • 1,708 subscribers
Taking complicated things apart. There are always fewer parts than you'd think.
Videos
In 1983, an 80-year-old Navy admiral held up a piece of wire and proved that almost every engineer in the room had been lying to their bosses about the speed of their own computers. Grace Hopper cut the wire to 11.8 inches — the maximum distance light can travel in one nanosecond. Anyone who quoted a chip speed in nanoseconds without knowing that number didn't actually understand what they were selling. The wire wasn't a teaching prop. It was a public test. Generals and executives had been nodding along to "nanosecond" for years without being able to picture one — she made the invisible unit something you could hold, and something you could be caught not understanding in front of a room. 42 years later, computers run in picoseconds, a distance a thousand times shorter than what she was cutting, and nobody hands out the proof on live TV anymore.
Alex Carten458,823 views • 2 days ago
Richard Feynman spent 4 days listening to NASA explain why the seal would have worked at that temperature. Then he reached for a glass of ice water. The night before the launch, the ice team measured the left booster at 25 degrees and the right one at 8. The engineers who built them had spent that evening arguing that nothing below 53 was safe. They were overruled on a 30-minute private call. The manager on site refused to sign the launch recommendation, so a vice president in Utah signed it instead and faxed it to Florida. Challenger lifted off at 36 degrees and came apart 73 seconds later with 7 people aboard. The report called this a failure of communication. That label is the least interesting thing about it, because the exchange that settles the whole question is 38 seconds long and contains no argument at all. A NASA official is asked whether anyone understood the seal would stop working at some temperature. He answers in the passive voice and closes on 5 words that hand the decision to nobody in particular. Then Feynman, who has said almost nothing for hours, runs the only experiment anybody performs on camera during the entire investigation. He clamps a piece of the seal, drops it into the ice water in front of him, waits, and takes it out. The rubber stays squashed. It does not spring back. He gives the temperature at which this happens, and finishes on 9 words that are the most polite sentence ever used to end an official version of events. The number underneath the footage is worse than the footage. NASA management had been putting the odds of losing a shuttle at 1 in 100,000. Feynman polled the working engineers anonymously and got answers between 1 in 50 and 1 in 200. Then he found where the management number came from. It was never calculated forward from the hardware at all. It was chosen first, because the vehicle carried people, and the odds of the individual parts were fitted backwards until they produced it. The shuttle went on to fail twice in 135 flights. The chairman of the commission described him privately in 6 words, and it was not a compliment. The idea to look at the cold was not even his. It reached him through an Air Force general who asked what cold does to rubber. Who gave it to the general stayed hidden for 27 years, and it turned out to be the person on that commission with the most to lose by saying it out loud. His own conclusions were kept out of the report until he threatened to take his name off it. They were printed at the back as an appendix and end on a 16-word sentence about the one thing engineering cannot negotiate with. He died 2 years and 4 days after the hearing. In 2003 a second orbiter came apart on re-entry, and the board that investigated it found the decision process behind the first accident had never been fixed. The engineers were right, and they said so out loud the night before. It changed nothing, because 1 man in Florida refused to sign and 1 man in Utah agreed to. Would you have been the one who refused to sign?
Alex Carten189,962 views • 14 days ago
Everyone assumed string theory was finally dying. It's not dying. It's worse. Physicists have written more papers on string theory than on any other idea in the field's history. Fifty years of work. It has never once produced a testable prediction. Last week an essay asked the question directly: is string theory still our best hope for a theory of everything? The answer given was yes. Here's the actual argument behind that yes. String theory is "inevitable" — as long as you first assume a universe that doesn't exist. That's not a rebuttal to the critics. That's the entire case for the defense. 50 years. 100,000+ papers. Entire careers built on a framework nobody has ever tested, defended by physicists who now have to assume a fictional universe just to keep it alive. Is this still physics, or is it the most expensive tautology in the history of science?
Alex Carten38,340 views • 19 days ago
No more content to load