Funny that the token cost of solving Navier–Stokes was ~$3M while the prize money is $1M 😀. Crazy feat — and it really begs the question: what does the future of math look like from here?
Great post on it here: https://t.co/uTFKnJEFZw
Accuracy for any production text-to-sql is quite far from acceptable. LiveSQLBench stands at 48% accuracy. Check out how we get to a higher accuracy in production (85% on LiveSQLBench)
https://t.co/fgouQIbsLa
I have always liked analyzing the performance and efficiency of systems. We just shared an interesting story on @golang memory analysis on the @gleanwork blog https://t.co/xCgG4tI5Ri. Feel free to reach out if you are interested in learning more about this or @gleanwork.
Render Caching is a technique I've never seen written about. The @gleanwork new tab page is dynamically built on the client, but has to load instantly. So we snapshot the DOM to the Service Worker for later rehydration.
@workboxjs + @reactjs = 💨
https://t.co/0fvR1Jy1iq
Yesterday was beyond exciting getting to debut Glean and show off the amazing hard work our @gleanwork team has done for more than two years. https://t.co/crPU0VJRmu