**Response to @paulg:**
Corollary: If you can imagine someone saying "I think we should say 'Here's how we're improving Airbnb' instead" that's the sort of person who kills the startup spirit in companies. Keep them out for as long as you can.
As long as you can still use phrases like "what you don't like," you still have the core of being a startup. Imagine Google or Apple being so candid.
nitter.moomoo.me/bchesky/status/1653735980875018246#m (https://nitter.moomoo.me/bchesky/status/1653735980875018246#m)
Someone who knows about college admissions told me why some universities are abandoning standardized tests. They know they're going to be under more scrutiny, and perhaps even subject to lawsuits, and standardized tests make it easier to see what they're doing.
**Response to @paulg:**
Ensuring the truth of your axioms would be a great thing if you could do it, but few people can. They're axioms; they already seem true. Whereas demoting some axioms to provisional beliefs is a practical possibility.
If you have false axioms, they'll warp all your further conclusions, as you seek evidence that confirms them and ignore evidence that contradicts them. Most people think the solution is to ensure the truth of their axioms, but a better one is to decrease their number.
**Response to @paulg:**
Incidentally, my answer to my question is fusion. Wouldn't that be the icing on the cake of the surprising 2020s?
**Response to @paulg:**
When something has a history of never having been good enough, your natural inclination is to treat this as evidence that you can ignore it. But it might be better to treat it as evidence that you shouldn't.
**Response to @paulg:**
As these examples suggest, "What's been around for a long time, but has never been good enough to take seriously?" is a good heuristic for predicting important new developments. When people have been trying to do something for a long time, there's usually a reason.
Electric cars, fake meat, AI. All three have been around since I was a kid. They were never good enough, so we learned to ignore them. But all three have now crossed the line where you can't. What's next?
**RT @blader:**
After 20 years of interviewing candidates, I've found this to be the single most useful interview question:
"What is your greatest strength ... that you are most worried about not coming across in an interview setting?"
https://nitter.moomoo.me/blader/status/1653491665527783424#m
**RT @jesslivingston:**
The thing that always strikes me about the @Airbnb (https://nitter.moomoo.me/Airbnb) founders is the way they kept going despite a whole year's worth of evidence that the idea wouldn’t work. Hear the story from Brian Chesky himself:
pod.link/1677066062/episode/… (https://pod.link/1677066062/episode/df3e0c417390af992b16a946bcb5651b)#socialradars (https://nitter.moomoo.me/search?q=%23socialradars)
https://nitter.moomoo.me/jesslivingston/status/1653452359479984132#m
**RT @amasad:**
Chris Lattner of LLVM and Swift fame just announced a new programming language for ML that is high-performance and backwards compatible with Python (works with Python libraries). Could be a game changer.
https://nitter.moomoo.me/amasad/status/1653447664816783361#m
"If I want to travel somewhere after I have kids, I'll just take them with me."
I've heard this sentiment expressed fairly often in the future tense, but almost never in the present or past tense.
AI will dramatically increase the already large gap between the information haves and have-nots, because the groups currently most likely to be fooled by fake news are also those with the least understanding of what AI can do.
**RT @granawkins:**
STANFORD TORUS MEGATHREAD
You’ve seen the pictures, but you probably don’t know the full story.
It's the result of a 10-week engineering study at Stanford University in 1975, funded by NASA and led by Girard O’Neill - the most thorough study on free-space settlement to date.
1/
https://nitter.moomoo.me/granawkins/status/1653223967631769600#m
**RT @cremieuxrecueil:**
Tests created in different times, by different people, from different test-making traditions, for different purposes, with different target constructs, produce a normal distribution when their items are scored right/wrong and summed.
Normality isn't baked in.
https://nitter.moomoo.me/cremieuxrecueil/status/1653198944464171010#m
**RT @Replit:**
1/ Announcing Replit Extensions, a major milestone in our mission to empower a billion developers.
Extensions enable our developers to customize their Replit environment and take control of their tools.
Watch @masadfrost (https://nitter.moomoo.me/masadfrost) build an extension to use the Monaco editor on Replit.
https://nitter.moomoo.me/Replit/status/1653070412077752321#m
**Response to @paulg:**
Potential startup opportunity: an online bookseller that specializes in selling new books that are actually in new condition when you get them. It's a good beachhead in the sense that word would spread quickly among potential users. But what do you expand into?
**Response to @paulg:**
Interestingly, packing books badly was a conscious decision by Amazon. I know this because for the first few years they packed them well. Then they switched to just dumping them in a box, presumably because it was cheaper and most buyers didn't care.
**Response to @paulg:**
When I want to buy a book that's still in print, I begin by looking for a used copy in good condition on ABEbooks, since I know a reputable used bookshop will at least pack it well. (Yes, I know ABE is owned by Amazon.)