I fully get and agree that there are dangers to AI; I just don't see how it is going to physically kill us all. Can someone please enlighten me? Thx everyone π !
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
About a year ago, a Princeton student named Kyler Zhou was sitting in traffic in New Jersey. At some point he was so bored that he put on a podcast withβ¦ me π
It was my usual talk about the talent crisis. About the smartest kids of a generation going to work for McKinsey and Goldman Sachs, while the biggest problems in the world go unsolved, blablabla, you may have heard it by now ;)
Whatβs interesting is what happened next. As soon as Kyler got home, he emailed me and The School for Moral Ambition (the organization we started to get talented people working on the world's biggest problems).
Kyler wanted to bring this to Princeton.
Well, we told him no π . We had just launched our first university fellowship at Harvard, and there was no capacity for a second campus...
Kyler decided to ignore us. Three days into the semester, he launched the first* Moral Ambition Chapter at a university in the United States. In just a few days, more than 120 Princeton students signed up. In November of last year I visited, and the energy was simply AMAZING.
Since then, everything has been escalating. Students have started Moral Ambition Chapters at Brown, Stanford, Vanderbilt, Georgetown and 5 other campuses β often before we've heard of them.
Which brings me to today. After the launch of our Harvard Fellowship (8% of ALL Harvard juniors applied to the first cohort!), we're now also launching Moral Ambition Fellowships at Princeton, Brown, Vanderbilt and Stanford. The program is simple: twelve super talented students per campus, a $15,000 stipend, and a summer working full-time on the most important causes of our time, like lead poisoning or child poverty or ending factory farming or AI safety etc. β instead of polishing a boring corporate pitch deck.
It's just incredible to see how quickly this movement is growing. Every year, thousands of brilliant graduates are funneled into corporate careers that add little to the world. More and more of them are starting to walk the other way.
(For students: applications for Harvard, Princeton, Brown and Vanderbilt are open now. Stanford opens October 6. https://t.co/3dM17vtNbC)
* Brown University claims that they were actually the first β I've decided to stay neutral in this quarrel π
Something I think Elon does exceptionally well is delegate without sacrificing his understanding of the details
In SpaceX's early days, the company needed an actuator to steer the upper-stage engine of its Falcon 1 rocket and a supplier quoted $120,000.
Elon understood the mechanism well enough to argue that it was "no more complicated than a garage-door opener" and gave an engineer a $5,000 budget to build one instead. All said and done, it ended up costing $3,900 to make. Without a deep understanding of the mechanism, this doesn't happen.
Point is, in almost every organization, leaders gradually lose touch with exactly how the mechanics of the work inside their organization gets done. They might broadly remember what they were doing five years ago, but over time the details become more and more abstract. What they clearly know is what the outcome they'd like to see from the work would look like, but the mechanics of how it happens? Not always.
What happens as a result is that such leaders become fully reliant on their teams for updates. They can't meaningfully audit the work, challenge assumptions, or help solve a problem when someone gets stuck.
Elon's approach to operating his businesses is different because he works to understand the details at every level. That allows him to collaborate with the experts on his team and solve the biggest problems his companies face.
The core idea here is simple: delegate the work, but never delegate your understanding of it.
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
@hlovo_@MKBHD@Mrwhosetheboss These guys are great! But I love how @SnazzyLabs goes another layer deep, rather than just recapping event. IDK if anyone has mentioned him here yet.
Exciting news!
I was accepted to several incredible institutions including @Stanford , @Caltech, @BrownUniversity, & @UCBerkeley EECS, each of which I would have been proud to attend. Ultimately, I committed to Stanford for Physics & Computer Science. π²
Go Cardinal!!!
@BocaBola_@altstoreio@sidestoreio Also, emulators are now allowed on iOS thanks to the EU! https://t.co/gshGaEbpje
Some good ones are Delta, Providence, PPSSPP, Retroarch, Dolphin, Manic, and also Gamma. I have tried many of these, and have heard positive things about the rest.
Welcome to iPhone lol
@BocaBola_@altstoreio@sidestoreio Also worth noting that you can set up a system-wide adblocker using something like adguard pro which creates a vpn that can reroute your dns through their adblocking dns servers. https://t.co/quG9CmpvDE
This one is a one time fee $10. Since I bought it alternatives have come tho.