algorat➰〰➰ currently at @mit @medialab @futuresketches && doing math at the club && taking care of a warehouse @hexh0use past: @nvidia @carnegiemellon
Brooklyn, NY
Joined March 2010
- Tweets3K
- Following1K
- Followers3.6K
- Likes10.3K
Pinned Tweet
Last weekend I made a fantasy IDE. It had a moving windowing system to livecode multiple shaders, audio code view, webcam feed etc.
Includeding a chatbot & QR code to a group chat w it. It could give a track ID, tell you to dance, or explain the current code on screen if asked!
char retweeted
the most effective lies are the ones the teller believes, so it follows that the most effective liars are those who construct a different reality around themselves to facilitate that (often with the help of others)
Math drama has existed before written history, and math drama will persist long after we are gone.
When all else crumbles, there will always be math drama.
cims.nyu.edu/~tristanb/state…
(Unsurprisingly OpenAI regurgitated existing unpublished work by researchers)
ALT I asked whether the model had been trained on, or had access to, our sessions in Codex, into which we had been putting all our drafts for the whole of this project. I was told the model did not look up user data. I asked again, about training, and I did not get an answer.
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.
The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.
The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
“Of course, there are many mathematicians who are more or less honest. But almost all of them are conformists. They are more or less honest, but they tolerate those who are not honest." - Perelman
“But automating mathematics would come at a heavy price. I argue in this piece that AI tools short-circuit human understanding, devalue mathematical knowledge, and threaten the infrastructure that keeps mathematics coherent.”
arxiv.org/html/2608.02859v1
a friend lovingly created an open source ebook reader (really a programmable e-ink touch screen you can write code for it, it comes w beautiful ebook software) if you had a goal of reading more (& want to say fk u to kindle/amazon) this is 4 u (RT pls)! crowdsupply.com/oddly-specif…
char retweeted
if you want to run this yourself, the developer (slvDev) open-sourced the entire project here: github.com/slvDev/esp32-ai
char retweeted
I’m not against AI assistants or figuring out the ethical maze of AI companions
I just think the current focus being largely on outsourcing the need for care/connection is lazy
Why not find thoughtful ways to anticipate logistics/execute tasks that allow one to be more present
This quoted post is unavailable.
char retweeted
biologists: ongoing debates around the safety and ethics of mosquito gene drives for 20+ years
meanwhile... a YC backed startup is building autonomous killer micro drones to "completely eradicate mosquitoes"
there's a lesson in here somewhere
found this v cool piece where thy recreate the first convolutional network by LeCun (from 1989!!) with LEDs on transparent PCBs
Video reddit.com/r/computervision/…
socius-org.github.io/artefac…
I remember maybe @aman_gif showing me something similar a while ago but it was more analog ?
char retweeted
♡(ˊo̴̶̷̤ ᴗ o̴̶̷̤ˋ)⸝* the old Chinese Internet is almost gone, but I wanted to hold on to a piece of my childhood - so I scraped 5000+ gifs from tencent’s CDN + @waybackmachine to make a museum of 2005-2009 qZone (our MySpace).
you can even design your own page! link below :)
char retweeted
The Digital Markets Act was designed to move fast, but Politico’s reporting highlights how enforcement is being slowed by legal caution, limited resources, and the threat of litigation from Big Tech.
🧵👇️ (1/10)
📖Read more:
politico.eu/article/eu-big-t…
