Twitter @dpaleka
I am a researcher in a field that moves fast. It is hard to overstate how new everything in machine learning is; to paraphrase a famous quote, a DL researcher in 2022 time traveling back to 2012 is Prometheus bringing fire to mankind.
Twitter turns out to be:
- the best way to know what happens;
- the best way to show other people what I’m doing.
Furthermore, it’s a good way to connect with people who could make great collaborators or friends. You should have Twitter. For better or worse, this is where the action is. 1
Of course, Twitter is social media, hence it is full of bad actors and bots. I decided upon two simple rules to make it more bearable:
- mute people who post things you don’t want to waste attention on;
- preemptively block bad people and trolls, so they don’t ever interact with your tweets. Do that liberally: if you are even contemplating blocking someone, you should do it. 2 Your time is precious.
AI/ML researchers
Machine learning theory
AI security and safety
Applied machine learning
- Eric Jang
- Favorites: just ask for generalization, reviewing papers, rome
- Andrej Karpathy
- Davis Blalock
- Zhendong Wang
If you have (or have read) a cool site along these lines, feel free to send me a link.
Academia
Machine learning online resources
Research group blogs
Philosophy and technology
- Nat Friedman
- Sam Altman
- Favorites: Moore’s law for everything, how to be successful
- Paul Graham
- Alexey Guzey
- dynomight
- Favorites: being managed, historical analogies for LLMs
- Ben Kuhn
- Favories: searching for outliers, staring into the abyss
Misc
Out of distribution
- Bits About Money by Patrick McKenzie
- Milan Cvitkovic
- Bartosz Ciechanowski
- ACoUP
- Favorites: the LOTR series, steppe nomads
- Austin Z. Henley
- gwern.net
- Guzey’s Best of Twitter
Evergreen links
Remarks
It’s a sad state of affairs that I have to state it explicitly, but I do not endorse any of the sites above, and especially any people behind those. I haven’t even met most of them, and it is quite possible some of them will end up on the wrong side of history.
A site being here means that it’s either interesting or important. Many things that are important are not good, and many interesting things don’t turn out to be true or useful.
-
This heuristic is useful for some websites too, but not always. A reasonable compromise is blocking in
/etc/hostsand then unblocking if you need to interact with the site. ↩︎ -
The 2022 management of Twitter feels bas in many aspects. But I don’t think it changes the calculus unless the important discussions start happening elsewhere. ↩︎