
I like making things with others. LLM shepherd. Godโs first commandment to Abraham was to smash all the idols. https://teapawt.net/t.co/pCDpWgOOSz (Header photo by @cartoonconnie)
Joined March 2008
- Tweets15.6K
- Following4.9K
- Followers2.6K
- Likes16.5K
Pinned Tweet
For almost the whole last year, Iโve been writing and polishing an article that lays out a frameworks for how to think about (and fix) social media. And now itโs finally here, thanks to @techreview. Hereโs a thread with some key points ๐งต technologyreview.com/2021/12โฆ
Sahar Massachi ๐ฃ retweeted
Val Town's beautiful office is up for grabs!
- Location: downtown brooklyn (325 gold st)
- Rent: ~$9k/month all-in
- Size: 2k sq ft
- Available furnished or unfurnished
- Very nice landlord
We've loved it here, and would love the office to go to a friend. DM me :)
Sahar Massachi ๐ฃ retweeted
Sad day.
@ValDotTown is winding down ๐ฅฒ.
A year ago, I went from Val Town power user to team member in a flurry. I never really got the chance to articulate what the product meant to me, nor what a privilege it was to be brought on.
Now, having worked there for a year, I'm faced with the now even more difficult task of summing up the entire journey.
I'll give it a try.
Val Town unlocked for me a level of creativity and production unrivaled by any tool I have ever used. Like my bio says, within a few months of discovering Val Town, I was literally "thinking in vals."
The product, paired with AI, allowed me to manifest almost any idea or tool that crossed my mind near-instantly.
A few of my projects started to catch the notice of the Val Town team. When a growth role opened up, I put my hat in the ring.
I was not a particularly legible candidate. I had bounced around a bunch of personal projects over recent years. But @stevekrouse knew I loved Val Town, saw a lot of potential, and gave me a chance.
If finding Val Town was like discovering magic, working at Val Town was the chance to develop that magic system alongside the wizards who designed it. It lived up to that promise.
I learned so much from this team. Technical and product skills, sure, but having worked as a solo founder for a long time, the main lesson I will take away from this experience is what a privilege it is to row in the same direction as Really Good People.
All of my wins at Val Town were in no small part due to the confidence the team instilled in me. (That includes encouraging me to start posting here after 10 years away!)
I'm collecting some highlights over here:
charliemolthrop.com/bestof
To every user and fan, like Steve said, thank you for your enthusiasm and creativity.
If you need any help migrating or looking for alternatives, don't hesitate to reach out. The silver nail in the coffin is that there remains a wide world of tools out there for us and our projects.
For me personally, well, first I need to finish migrating my more than 1200 vals!
After that, let's see! I'm still trying to understand my best fit on this insane sci-fi timeline. It's going to be hard to top my last gig, but if you've got any ideas, let's chat!
Thank you to Steve, Tom, Jackson, Pete, Potluck, and everyone else who made Val Town what it was.
Val Town is winding down
Thank you all for building on us with such enthusiasm and creativity! We're sorry we can't continue to serve you
We're now focused on helping customers migrate to other hosting providers. Please reach out if we can be helpful
blog.val.town/winding-down
Sahar Massachi ๐ฃ retweeted
Some startups make unbelievable financial returns, but most don't. I'm fortunate enough to have a reasonably good track record as an investor, but I never know which will be which deterministically.
I realized that fact a long time ago and resolved to try to work with founders that I love in a deep and intangible way. Folks I wake up wanting to help every day. People I try to help however I can, whether it's a weekday or weekend, a personal matter or a work one, a happy situation or a sad one.
I want to work with founders where I deeply care and enjoy the holistic journey, no matter the outcome. I optimistically believe if I do that, the aggregate returns will take care of themselves. That's the beauty of real venture capital.
@stevekrouse is a real one. I loved every minute of Val Town and I'm excited to continue the journey with him for another few decades if he'll have me
Val Town is winding down
Thank you all for building on us with such enthusiasm and creativity! We're sorry we can't continue to serve you
We're now focused on helping customers migrate to other hosting providers. Please reach out if we can be helpful
blog.val.town/winding-down
End of an era. Wow
Val Town is winding down
Thank you all for building on us with such enthusiasm and creativity! We're sorry we can't continue to serve you
We're now focused on helping customers migrate to other hosting providers. Please reach out if we can be helpful
blog.val.town/winding-down
4
104
This is the way
ืืืฉืืจืช ืฉืื ืืจืืืืดื ื ืคืื 3 ืืืืืื ืขืจืืื-ืืืกืืืื, ืฉืืชื ืืื ืืฉืจืช ืืฆืืดื. ืืฉืืืฉื ื ืคืื ืืืืืืชื ื ืืืืืก ืืจืฆืืขืช ืขืื (2003).
ืืืืงืืจ ืืชื ืืืืื ืฉืขืจืืชื ืืฆื ืืืช ืืืฉืคืืืช, ืืืื ืืืงืฉ ืฉืืืชืื ืืืืจ ืืืืจืืื, ืื ืืื ืจืืฆื ืืืืื ืืช ื ืฉืืช ืืืฉืคืื ืืืืืืื (ืืืจ ืืืืจืืื) ืฉืืื, ืืืื ืจืืฆื ืืืืจ ืืคื ื.
ืืืืจ ืฉืื ืื ืฉืื ืืืืืจืื ืืชืื ืกื ืืื ื ืฉื ืืช ืืืจืื, ืืืืจ: ืดืืืจื ืืจืืขื ืืืงืืืืจ 2000 (ืืื ืืชื ืืฉื ืขืจืืื ืขื ืืืฉืืจื ื-13 ืืื ื ืืจืื), ืืื ืกืชื ืืช ืื ืื ืืืืจืชื ืืื, ืฉืื ื ืจืืฆื ืืืืืช ืืืืื ืช ืืฉืจืื, ืฉืื ื ืจืืื ืืช ืขืชืืื ื ืืืืื ืช ืืฉืจืื, ืืื ืื ืืกืืืืื ืืืชื, ืื ืื ืฆืจืืืื ืืืฉืชืชืฃ ืืืื ืช ืืืืื ื. ืฉื ืืื ืืื ืื ืืื ืืืื ืืืืก, ืืื ืืชื ืื ืืฉืจืช ืืฆืืดื ืื ืคื ืขื ืืื ืช ืืืืื ื, ืืฉื ื (ืฉืขืื ืืฆื ืืืืื) ืืชื ืื ืืฉืจืช ืืืฉืืจ ืืืืืืด.
ืืืฉืจ ื ืคืฆืขืชื (ืืกืคืจ ืคืขืืื) ืืืคืืชื ืื ืขื-ืืื ืจืืคืืื ืขืจืืื ืืืืืืช ืขืจืืืืช. ืื ืงืืจื ืขื ืืืื.
ืืชืจืืฉืชื ืืืฉืจ ืฉืืขืชื ืืช ืื ืกืืจ ืขืืืก ืฆืืขื ืฆืขื ืืืืฅ ืืืืืจืชื, ืฉืืื ืืืืจ ืืืืื ืช ืืฉืจืื ืืืืื ื ืืืืืืช ืืืืืงืจืืืช ืืจืื ืืืืืช ืืขืฆืืืืช. ืื ืืืชื ืืืฉืืช ืืืื ื ืฉื ืืื ืืณืืืืื ืกืงื, ืืคื ืฉื ืืกื ื-1923 ืืืืืจื ืดืงืืจ ืืืจืืืด ืืืืฉื ืชื ืืืฉืชืืืืช ืืืืื ื ืืืืืืืช ืืืืจื ืชืืืื ืขืืื, ืืืืืกื ืืขืจืืื ืืฉืจืื. ืื ื ืืื ืื ืจืืฉื ืืืฉืืืช ืืฉืจืื, ืืื ืืืจืืื ืืขื ืืืงืืช ืืืืฉืื ืื ืืืืืช.
ืืืืื ืืื ื-7.10, ืคื ืชื ืืื ืืื ืืืืืืช, ืฉืื ื ืืฉืชืชืฃ ืืืืช ืืืกืืืืช ืืขืืืฃ. ืืืืจ ืืืจืืจ, ืื ื ื ืืฆื ืืงืืืืฅ ืฆืืืื, ืืืืจ ืฉืืืืฅ ืืื ืขื ืขืื ืขืฉืจืืช ืฆืขืืจืื ืืฆืขืืจืืช, ืขื-ืืื ืืืืื, ืขื ืืื ืืจ ืฉืื.
ืืื ืืืืจ ืืืื, ืืชืืจืื ื ืืขืืืจ ืืืฉืืื ืืขืืืฃ, ืืืจืืื ืืชืืื ืฉื ืณืืืืื ืื ืฉืงืณ. ืืชืืจืื ื ืืขืืืื ืืืงืืืืช ืืืจืืืช ืงืืฉื ืฉืื ืืช. ืืืชื ื ืืงืืืฆื (ืฉืืืจื ืืืจืืช ืงืืฉื ืืืืื) ืืฉืชืชืคื ืืงืืืขืืช ืืฉืคืื ืืจืื (ืืืจืื ืืืืืืื ืืืืืจืื).
ืืืฉืจ ืื ื ืฉืืืข ืืช ืจืืฉื ืงืืืืืฆืืืช ืืืฉืืืืื, ืืืฉืชืืืื, ืืืืฉืืชืื ืืืืืืืื, ืื ืฆืืื ืืช ืจืืฉืืช ืืคืื, ืืืขื ืืื ืงืื ื ืื ืืขืจืืื, ืืืฉืจ ืื - ืื ื ืืชืืืื ืืืชืืืืฉ ืืื ื ืขืื.
ืจืืฉื ืืคืืืืช ืืืฉ ืืฉืื ืื, ืืื ืืืชืืจืื ืืืฉืืืื ืฉืืื, ืืื ืืืืื ืืฉืืืื ืขืจืืื ืืฉืจืื ืืืืจื ืืืฉืจืืืืช. ืงืขืงืขื ืืช ืืืกืชื ื ืื ืืขืจืืื ืืชืืืื ืืจืืจ, ืืฆื ืื ืฉืืงืื ืืืฉืื ืขื ืื ืฉืืืจืฉืข ืืชืืืื ืืืจืืื ืืจืืจ ืืขื ืื ืฉืชืื ื ืคืืืืข, ืืื ืืื ืืข ืืช ืืืชื ืชืงืืช, ืืขื ืื ืฉืืคืขืืืื, ืืืื ืื ืืืขืืืืื ืืจืืจ ืืืืื ืืืืืื, ืฉืืืจืื ืืืงืขืช ืืืจืื ืขื ืืืื.
ืื ืืืืืช ืฆืจืืื ืืืืืื, ืืื ืืืืจืจ! ืื ืืื ืืคื ืืฆืคื, ืืื ืฉืืฉืืช ืจืืืืช (ืจืขืืช)!
ืืื ืืืืืื ืื, ืืืืฉืืขื ืืช ืืฉืจืื!
1
100
This is the way
ืืืฉืืจืช ืฉืื ืืจืืืืดื ื ืคืื 3 ืืืืืื ืขืจืืื-ืืืกืืืื, ืฉืืชื ืืื ืืฉืจืช ืืฆืืดื. ืืฉืืืฉื ื ืคืื ืืืืืืชื ื ืืืืืก ืืจืฆืืขืช ืขืื (2003).
ืืืืงืืจ ืืชื ืืืืื ืฉืขืจืืชื ืืฆื ืืืช ืืืฉืคืืืช, ืืืื ืืืงืฉ ืฉืืืชืื ืืืืจ ืืืืจืืื, ืื ืืื ืจืืฆื ืืืืื ืืช ื ืฉืืช ืืืฉืคืื ืืืืืืื (ืืืจ ืืืืจืืื) ืฉืืื, ืืืื ืจืืฆื ืืืืจ ืืคื ื.
ืืืืจ ืฉืื ืื ืฉืื ืืืืืจืื ืืชืื ืกื ืืื ื ืฉื ืืช ืืืจืื, ืืืืจ: ืดืืืจื ืืจืืขื ืืืงืืืืจ 2000 (ืืื ืืชื ืืฉื ืขืจืืื ืขื ืืืฉืืจื ื-13 ืืื ื ืืจืื), ืืื ืกืชื ืืช ืื ืื ืืืืจืชื ืืื, ืฉืื ื ืจืืฆื ืืืืืช ืืืืื ืช ืืฉืจืื, ืฉืื ื ืจืืื ืืช ืขืชืืื ื ืืืืื ืช ืืฉืจืื, ืืื ืื ืืกืืืืื ืืืชื, ืื ืื ืฆืจืืืื ืืืฉืชืชืฃ ืืืื ืช ืืืืื ื. ืฉื ืืื ืืื ืื ืืื ืืืื ืืืืก, ืืื ืืชื ืื ืืฉืจืช ืืฆืืดื ืื ืคื ืขื ืืื ืช ืืืืื ื, ืืฉื ื (ืฉืขืื ืืฆื ืืืืื) ืืชื ืื ืืฉืจืช ืืืฉืืจ ืืืืืืด.
ืืืฉืจ ื ืคืฆืขืชื (ืืกืคืจ ืคืขืืื) ืืืคืืชื ืื ืขื-ืืื ืจืืคืืื ืขืจืืื ืืืืืืช ืขืจืืืืช. ืื ืงืืจื ืขื ืืืื.
ืืชืจืืฉืชื ืืืฉืจ ืฉืืขืชื ืืช ืื ืกืืจ ืขืืืก ืฆืืขื ืฆืขื ืืืืฅ ืืืืืจืชื, ืฉืืื ืืืืจ ืืืืื ืช ืืฉืจืื ืืืืื ื ืืืืืืช ืืืืืงืจืืืช ืืจืื ืืืืืช ืืขืฆืืืืช. ืื ืืืชื ืืืฉืืช ืืืื ื ืฉื ืืื ืืณืืืืื ืกืงื, ืืคื ืฉื ืืกื ื-1923 ืืืืืจื ืดืงืืจ ืืืจืืืด ืืืืฉื ืชื ืืืฉืชืืืืช ืืืืื ื ืืืืืืืช ืืืืจื ืชืืืื ืขืืื, ืืืืืกื ืืขืจืืื ืืฉืจืื. ืื ื ืืื ืื ืจืืฉื ืืืฉืืืช ืืฉืจืื, ืืื ืืืจืืื ืืขื ืืืงืืช ืืืืฉืื ืื ืืืืืช.
ืืืืื ืืื ื-7.10, ืคื ืชื ืืื ืืื ืืืืืืช, ืฉืื ื ืืฉืชืชืฃ ืืืืช ืืืกืืืืช ืืขืืืฃ. ืืืืจ ืืืจืืจ, ืื ื ื ืืฆื ืืงืืืืฅ ืฆืืืื, ืืืืจ ืฉืืืืฅ ืืื ืขื ืขืื ืขืฉืจืืช ืฆืขืืจืื ืืฆืขืืจืืช, ืขื-ืืื ืืืืื, ืขื ืืื ืืจ ืฉืื.
ืืื ืืืืจ ืืืื, ืืชืืจืื ื ืืขืืืจ ืืืฉืืื ืืขืืืฃ, ืืืจืืื ืืชืืื ืฉื ืณืืืืื ืื ืฉืงืณ. ืืชืืจืื ื ืืขืืืื ืืืงืืืืช ืืืจืืืช ืงืืฉื ืฉืื ืืช. ืืืชื ื ืืงืืืฆื (ืฉืืืจื ืืืจืืช ืงืืฉื ืืืืื) ืืฉืชืชืคื ืืงืืืขืืช ืืฉืคืื ืืจืื (ืืืจืื ืืืืืืื ืืืืืจืื).
ืืืฉืจ ืื ื ืฉืืืข ืืช ืจืืฉื ืงืืืืืฆืืืช ืืืฉืืืืื, ืืืฉืชืืืื, ืืืืฉืืชืื ืืืืืืืื, ืื ืฆืืื ืืช ืจืืฉืืช ืืคืื, ืืืขื ืืื ืงืื ื ืื ืืขืจืืื, ืืืฉืจ ืื - ืื ื ืืชืืืื ืืืชืืืืฉ ืืื ื ืขืื.
ืจืืฉื ืืคืืืืช ืืืฉ ืืฉืื ืื, ืืื ืืืชืืจืื ืืืฉืืืื ืฉืืื, ืืื ืืืืื ืืฉืืืื ืขืจืืื ืืฉืจืื ืืืืจื ืืืฉืจืืืืช. ืงืขืงืขื ืืช ืืืกืชื ื ืื ืืขืจืืื ืืชืืืื ืืจืืจ, ืืฆื ืื ืฉืืงืื ืืืฉืื ืขื ืื ืฉืืืจืฉืข ืืชืืืื ืืืจืืื ืืจืืจ ืืขื ืื ืฉืชืื ื ืคืืืืข, ืืื ืืื ืืข ืืช ืืืชื ืชืงืืช, ืืขื ืื ืฉืืคืขืืืื, ืืืื ืื ืืืขืืืืื ืืจืืจ ืืืืื ืืืืืื, ืฉืืืจืื ืืืงืขืช ืืืจืื ืขื ืืืื.
ืื ืืืืืช ืฆืจืืื ืืืืืื, ืืื ืืืืจืจ! ืื ืืื ืืคื ืืฆืคื, ืืื ืฉืืฉืืช ืจืืืืช (ืจืขืืช)!
ืืื ืืืืืื ืื, ืืืืฉืืขื ืืช ืืฉืจืื!
156
It's fun to see Elon Musk waste half a billion dollars to learn a lesson the Great Slate learned for about $7 million. I'll share it again for free!
1. Elections today are highly nationalized, so even local races can absorb unlimited amounts of money without materially affecting the outcome.
2. Door knocking is not a magic strategy for boosting turnout; think about how annoyed you get when strangers come do it to you.
3. Deep conversations about political issues that take place between a community volunteer and a citizen on the latter's front porch exist only in the imagination of progressive think tanks.
4. There is no 'political dark matter' in American elections. Non-voters generally don't vote for a good reason.
5. Five minute spent talking to someone who has never voted generally leads you to the fervent hope that they will continue to never vote in the future.
6. The campaign process can absorb unlimited amounts of money with no effect on outcomes. This is against the conventional wisdom (people believe that $2,600 corporate donations somehow buy politicians) and infuriating if you're spending the money, but in fact a sign of great health and resiliency in American politics.
7. The one factor that reliably *does* get out new votes and change outcomes is a ballot initiative. If you have one of those, you can organize around it.
I whipped up a bot to grab the most interesting election updates and now itโs posting to Telegram.
t.me/electionbits
1
309
Sahar Massachi ๐ฃ retweeted
I knew Ethan knows elections better than anyone, but I saw a version of this hours ago and it predicted how this night would go earlier than anything else Iโd seen.
My live House needle has Dems at 280 seats right now. TX and NC are still counting their E-Day vote, which leans R. So I expect this to come down over the next hour. But I've been saying this for hours... house-eday-2026.pages.dev/
Follow this
My live House needle has Dems at 280 seats right now. TX and NC are still counting their E-Day vote, which leans R. So I expect this to come down over the next hour. But I've been saying this for hours... house-eday-2026.pages.dev/
2
265
This is the way
ืืช ืืืดืจ ืจืขืดื ืืดืจ ืื ืกืืจ ืขืืืก ืื ื ืืืืจ ืืคืืืฉืืช ืืฉืืืืช ืืจืืืืช ืืื ืกืช ืืืืืฆื ืื. ืื ืกืืจ ืืื ืื ืืื(!) ืขืจืื-ืืฉืจืืื ืืืืฅ, ืฉืืืืืฉ ืืช ืขืืืชื, ืืื ืฉืืืืจ ืืืืื ืช ืืฉืจืื ืืืืื ื ืืืืืืช ืืืืืงืจืืืช ืืจืื ืืืืืช ืืขืฆืืืืช, ืืืืื ืื ืืืืืก ืืื (ืืื ืื ืืฉืืจืืช ืฆืืื ืื ืืืืื-ืืืจืื). ืจืืืชื ืืฉืืขืชื ืืืชื ืืืืื ืณืฉืืืจ ืืืืืชืณ, ืืืฉืจ ืขืืจ ืืขืจืื ืืืขืืจืืืช ืืดืฉืคื ืืืืด ืืืงืืืืช ืืื ืืืื ืืกืืื ืฉื ืื ืืืืจ ืดืืฆืืชื ืืฉ ืฉื ืฉื ืื, ืืฉืคืื ืืืง ืืืืืจืืด.
ืืช ื ืืฆื (ืืืืืืก) ืืืื ืกืืืืืืฅืณ, ืื ื ืืืืจ ืืืื, ืืื ืฉืฉืืจืช ืืคืงืืื ืืกืืืจืช ืืฆื ืื ืื.
ืืช ืืืชืืจ ืื ืืืืจ ืื ื ืืืืจ ืืืืืงืื ืฉื ืืฉืืดื, ืฉืืืจืืจ ืฉื ืื, ืืกืชื ืืืืืืืช ืืืืจืช ืืืจืื. ืขื ืืขืฉืื ืืืจืฉืข ื-8 ืืชืื ืืืฉืื, ืืื ืืื ืชืืืื ืืืจืืื ืืจืืจ (ืืืืื).
ืืช ืืฆืืื ืกืืืืจืืฅืณ ืืืจืชื ืืื ืฉื ืขืฆืจ ืืคื ื ืืืชื ืชืงืืช, ืืืื ืืฉื ืืืืื ื ืืคืืฆืฅ 700 ืืืืจ ืืืง ืขื ืืืืฉ 1 - ืคืืืืข ืฉืืืืจ ืืื ืืขืฆืืจ ืืช ืืืฆืืข ืืืชื ืชืงืืช. ืขื ืืขืืืื ืืฉืจ ืืืฉืจื ืืืืืืื, ืืขืืืื ืืืืืื ืืืจืืจ ืืืืืื ืืืืืื, ืฉืืืจืื ืืืงืขืช ืืืจืื, ืืื ืฆืืจื ืืืจืืื.
ืื ืกืืจ ืขืืืก ืืื ื ืื ืืช ืืฉื, ืื ืื ื ืืงืืื ืฉืจืื ืืืจืื ืืฉืจืื ืงืจืืืื ืืขืจืืื, ืืื ืืขืจืื ืืืฉืืืืื, ืืืืขื ืื ืืืคืืฉืืกืืื.
ืจืืฉื ืืืฉ ืืฉืื ืื, ืื ืกืืจ ืขืืืก ืฆืจืื ืืืื ืืฉืจ ืืืืฉืืช ืืฉืจืื. ืืคืกืืงื ืืงืื ืืช ืืืื ืืืฉืืง ืฉื ืืืจืื ืืขื ืืืื ื-7.10, ืืืืื ืืฉืจืืฃ ืืช ืืืืื ื, ืืื ืืืฉืืจ ืืฉืืืื.
ืืฉืชืืื ืขื ืืื ืกืช ืืืงืืื ืืืฉืื ืืืืืคืืช, ืืืคื ืฉืขื ืืืช ืงืืื!
1
95
Sahar Massachi ๐ฃ retweeted
This talk is a beautiful articulation of one of the most important design problems of our time. A few thoughts on planning with agents:
- I'm finding one of the best inputs to plans are recordings of live conversations among a team. this tends to capture a *ton* of peripheral useful context, that's hard to capture even in a solo voice recording -- because people elicit intent from each other. A generalization of this principle is that agents thrive on broad, high level, "messy" context - just like senior employees.
- I think visual outputs from agents are critical to complement text, but I'm actually of mixed minds on interaction. Yes, it's clearly nice for certain kinds of things like twiddling visual parameters. But most of the interactivity my agents produce feels mostly designed for dopamine reasons and not providing much utility. I also think interaction can inhibit deep thought compared to linear reading. I actually think for many deeper problems I would prefer top notch prose -- Richard Dawkins / Steven Pinker / Annie Dillard level -- from my agents over any kind of flashier output.
- One mental model I find consistently helpful for planning: the job is to establish a *design space* together between human and agent -- that is, a set of relevant dimensions of variation -- and then to efficiently explore the space. Getting the dimensions of the space right is actually one of the hardest parts. Once you have that, you can ask an agent to plant a bunch of starting points spread throughout the space, and then locally hillclimb from the best one.
New talk! About the divided human & agent worlds, boundary objects, and building thicker interfaces
I've been exploring how to improve "planning", aka. deciding what agents should build, how you want it built, & making the whole team part of that process
maggieappleton.com/planning-โฆ
Sahar Massachi ๐ฃ retweeted
we need to bring back company toilet engineering newsletters
facebook had one called โthe weekly pushโ
No time wasted at the @modal bathroom, where I left the urinal knowing more about DST than when I arrived
๐๐๐๐๐๐๐
The GRU hacked into Jared Kushner's iPhone. Dutch military intelligence (MIVD) then recruited a GRU spy who gave them access to the hacked phone. What MIVD discovered was that Kushner had access to two bank accounts --one Azerbaijani, the other Israeli --where Saudi Crown Prince Mohammed bin Salman and Emirati ruler Mohammed bin Zayed deposited money. It sounds like a slightly far-fetched episode of "Slow Horses," but it's real. And it proves just how eye-poppingly corrupt this administration was and continues to be as its "special envoy" decides the fate of tens of millions in Europe in the Middle East. Simply incredible reporting by Matthew Cole, drawn from his new book on Kushner. vanityfair.com/story/jared-kโฆ
130
Sahar Massachi ๐ฃ retweeted
CLAUDE FOUND THE SEEDS
anthropic.com/research/ecdsaโฆ
I find this half fascinating and half baffling
Gonna marinate in this one
interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had
with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works
already-clean.goodfire.com/
1
3
164
Everybody:
Me: SMASHing that preorder button like a looney tunes on rocket boots dunking a basketball
There are real limits to what we can accomplish in a world where people believe nothing they do really matters.
My book (!) is about how we wound up in a place where millions of people feel like spectators in their own lives. And how we course correct. Because we can.
91























