@MollySOShea I think he also realizes that strict alignment to a questionable vision is far stronger than weak alignment to various “perfect” visions
Large companies have so many people pulling to their own vision they can’t execute
Elon doesn’t abide that
@garrytan He’s being naive in assuming “productivity increase cancels out gap from population”
Why?
Because income drops from tech always land unevenly.
If AI favors “capital” over “labor” then *mean* abundance goes up but only because really rich people got richer… median goes down
@tristanbob I got “local AI fever” briefly
Somehow it feels like, just by spending money, you are being productive and clever
But then you do the hard math and realize it is neither
You guys should learn more about the “technology” that is macroeconomics
His take is bad.
Eg he thinks price decreases are good because he doesn’t unserstand one person’s price is another person’s salary. If both go down, you’re not better off.
Which goes down more is non-obv
Deflation and “lower prices” is not automatically good.
The price of a haircut is a barber’s income. Every price is someone’s salary. If prices go down at the same amount as income, you’re not better off.
In fact, you get economic depression because things like debt contracts are locked in at the higher price
Macro Econ is a “technilogy” worth knowing about
@Regman_ETX@lennysan@pmarca@a16z I get that’s the point he *thinks* he’s making but It’s just not how it works.
The price I pay for a haircut is the salary for a barber. The price of corn is a farmer’s income.
If prices drop, all else equal, so do incomes.
Which drops more is very esoteric.
@mwseibel A good spokesperson for this: Stephen He (https://t.co/LMruGP9plG)
He is famous on Youtube for imitating his Asian dad.
People don't realize his "American accent" was actually the imitation... he spent 2 years learning it.
It glitches your brain to hear the difference
@ThePrimeagen AI is better than 98 percentile of humans at a given job...
You, as a developer, are in top 99th percentile so it doesn't add value to you
If you were trying to write investment copy for a pitch deck, Gemini 3.0 might be better than you now
@paulg Plus, most large companies actually just want to wait for the Microsoft version of your service
Entire startups die trying to jump through the approval processes when large companies lead them on
Does anyone have a framework or questionnaire for interviewing potential co-founders?
Some questions would be start-up specific but handy to see what's out there
@garrytan tae kim is wrong
If you actually build with AI you'd note that Google's data points are steeper than anyone, including OpenAI and Anthropic.
18 months ago Google was behind
6 months ago, they were just behind
They are now ahead of everyone except Anthropic
Next data point...
@DavidSacks This is just naive as to how "truth" works.
"Unbiased" sits at a point equidistant between political sides
"Truth" sits wherever it sits
That means "truth" ≠ "unbiased"
People increasingly arbitrage this by adopting extreme positions