Very excited to share our interview with @polynoamial on AI for math — the Erdős unit distance problem, saturating the IMO, the future of math research, and more!
Diffusion Without Tears is our attempt to make the score-matching + SDE interpretation of diffusion geometrically intuitive. If you're interested in our upcoming interview with @DrYangSong, I recommend reading this first! Link below.
the reason data quality is bad is because most data teams have zero power over the creation of data. it's like a soviet supply chain, where the end consumer has to accept whatever the factory churns out.
Very excited to share our interview with @DrYangSong. This is Part 2 of our history of diffusion series — score matching, the SDE/ODE interpretation, consistency models, and more. Enjoy!
my advice for data scientists who want to do high-impact work at startups is to learn enough software engineering to be self sufficient
we'll know the data stack revolution is complete once this becomes bad advice (a day I greatly look forward to)
Very excited to share our interview with @jaschasd on the history of diffusion models — from his original 2015 paper inventing them, to the GAN "ice age", to the resurgence in diffusion starting with DDPM. Enjoy!
Excited to share our interview with @sedielem! This is Part 3 in our History of Diffusion series. We talk about diffusion as spectral autoregression, diffusion language models, flow matching, and much more. Enjoy!
waiting for the "engineers shouldn't write etl" moment for user event tracking.
frontend/mobile engineers hate writing & maintaining event tracking code, but data analysts can't do it on their own with current tools. nobody's happy, and we need a new solution.
The problem with vc market maps is that the categories are made up and the labels don't mater.
✨AI solves this✨
Instead of grouping startups by what investors think, now you can group them by what text-embedding-ada-002 thinks. (Interactive version: sl8r000.github.io/startup_embedd…)