1380
Ansgar
@ansgar.eth #1380
Ethereum protocol R&D. Admirer of emergent order.
189822 Follower 250 Following
Gm Starbase! Today is going to be my first ever time watching a rocket launch in-person - looks like I picked a good one! https://x.com/spacex/status/1845424870655398102
The timing on this is ridiculous - my first ever larger size polymarket bet, and that just so happens to be seconds before someone comes in with almost $1M to shift that market 4% in the other direction, in an (unsuccessful) attempt to win a secondary market…
https://x.com/jon_charb/status/1832112496968171679
https://x.com/jon_charb/status/1832112496968171679
New LLM release. Inner monologue as intermediary output seems very promising - my feeling is that LLMs are effectively "system 1" thinking, but so far we have been training them mostly on "system 2" output (because that is more readily available).
https://x.com/mattshumer_/status/1831767014341538166
https://x.com/mattshumer_/status/1831767014341538166
absurd - Microsoft makes it impossible to disable the "Email a code" login option. Someone keeps trying to guess this (6 digit) code for my account. At this rate they have a ~1% brute force chance per year...
The ability for countries to make credible commitments to investors across changes in government is good, actually.
https://x.com/slatestarcodex/status/1831405516326891549
https://x.com/slatestarcodex/status/1831405516326891549
If you are a rationalist and not willing to "question" democracy, you are not a rationalist.
https://x.com/samoburja/status/1830481350488637805
https://x.com/samoburja/status/1830481350488637805
Goodbye SF, I’ll be back!
TIL about the term "neuro-symbolic AI". Always good to have a term to put to an interesting concept.
https://en.wikipedia.org/wiki/Neuro-symbolic_AI
https://en.wikipedia.org/wiki/Neuro-symbolic_AI
Always a bit alien to me to see so many laptops in a lecture setting. During my 4 undergrad years (maths, then econ) I didn't even own a laptop. Only bought one (a used MacBook) when switching over to CS for my master's, and even then mainly for homework after class.
https://x.com/vikhyatk/status/1828653905968079225
https://x.com/vikhyatk/status/1828653905968079225
When Tesla first released the Model 3, I remember visiting California and being excited the few times during the trip I spotted one. Then on my next visit a year later, it felt like they were ubiquitous.
I get the same feeling about Waymo in SF now. Last year, it was exciting to spot one. Now they are everywhere!
I get the same feeling about Waymo in SF now. Last year, it was exciting to spot one. Now they are everywhere!
Even pretty censored local LLM models are mostly following instructions if one manually writes the first 1-2 words of the response and only lets them complete from there.
Would be interesting to use a small (uncensored) model to automate that process.
Would be interesting to use a small (uncensored) model to automate that process.
PODs are the future.
iykyk
iykyk
Had a very SF moment just now:
I was in a self driving car and passed by a Party for Socialism and Liberation demonstration with people holding a big "the people voted, Maduro won" banner.
The two sides of SF…
I was in a self driving car and passed by a Party for Socialism and Liberation demonstration with people holding a big "the people voted, Maduro won" banner.
The two sides of SF…
Seen in New York
Exciting to see this collaboration between Perplexity and Polymarket. This year (and obviously the US election cycle) really seems to be the year of prediction markets reaching critical mass!
https://x.com/perplexity_ai/status/1823029449534615705
https://x.com/perplexity_ai/status/1823029449534615705
I’ve seen a few people on here with their own personal channels. So I’ll give it a try as well.
Created /ansgar, and will set my casts there to "channel followers only". That way it’s truly opt-in and I won’t have to worry about spamming people’s feeds.
See you over there (maybe)!
Created /ansgar, and will set my casts there to "channel followers only". That way it’s truly opt-in and I won’t have to worry about spamming people’s feeds.
See you over there (maybe)!
Agree with Barnabé. Braid looks interesting, but I have seen it referred to as multi-proposer, which seems misleading (there it seems to be very similar in spirit to FOCIL). To me its main contribution is in exploring a deterministic ordering mechanism for L1.
https://x.com/barnabemonnot/status/1820912429418811594
https://x.com/barnabemonnot/status/1820912429418811594
I see that "channel muting" ended up happening the other way, with the "channel followers only" cast option. This effectively gives control to the poster instead of the reader - interesting choice!
Reasonable point by Anatoly, but of course that's what Ethereum's L2-centric approach is about:
Any user should be able to frequently and trustlessly (!) use L2s. Trustless means it must be affordable to run an L1 full node and L2 lightclients.
https://x.com/aeyakovenko/status/1820288161769202141
Any user should be able to frequently and trustlessly (!) use L2s. Trustless means it must be affordable to run an L1 full node and L2 lightclients.
https://x.com/aeyakovenko/status/1820288161769202141
I decided to start the next season of trying to make farcaster my primary social network home. Prediction is that it will take a few more attempts for me, but I am very impressed by how far farcaster has already come and very optimistic for its future progression. @dwr.eth and team are continuing to do an amazing job!
This might be a very weird flex, but on arrival at SFO (from Europe) just now I made it from touchdown to leaving the airport by train in 24min. While even volunteering to give a nose swab (plus questionnaire) in between. And without global entry.
Exhilarating feeling.
Exhilarating feeling.