Gary Basin
56.5K posts
@turtlelambvase @norvid_studies @lu_sichu @max_spero_ i had the same intuition as norvid's op and my sense is random text would bias towards human because pangram is calibrating for *specific* basins of ai post-training text where human is the default / low false positives. some evidence
Monte Carlo Tree Search training corrects the model move by move, while current LLM training only tells it whether the whole trajectory worked. MCTS is preferable if you can get it. But nobody's managed to get MCTS to work for language models. In his blackboard lecture @ericjang11 talked to me about why:
Using ai to visualize removing the ugliness from our world
26 is the ultimate coming-of-age age. the age of breakups (or engagements). the reality of your late 20s hits you like a bus. you try a new routine. new hobby. new city even. chasing the missed early 20s. at 26, you reinvent your life. my favorite thing to watch.
After 15 years of investing, we realised that truly exceptional founders have something impossible to fake: deeply unconventional lives. We analysed 15,000 founders using five binary signals to measure this: odd hobbies, early signs of exceptionalism, extreme life choices, unusual geographies, non-linear careers. These sum to give a 0-5 score per founder. Whether someone started coding at 10, speaks five languages, climbed Everest or quit a safe job to live in Chile, the signal was deviation from the mean. Rather than focusing on IQ or EQ, we call this metric the Outlier Quotient, or “OQ”. When forecasting founder success, it turns out that OQ was the single most predictive variable in our entire classification model, trained on ~70 different factors. Our OQ score had zero correlation with having worked at a top-tier company or attending an elite university. The signals most VCs rely on aren’t just noisy, they’re blinding. The best founders don’t signal like everyone else, they don’t think like everyone else, and they certainly don’t build like everyone else. If you want to spot breakout talent before the rest of the market, stop screening for conformity. Back the founders the system was built to filter out.
I finally found *the* solution I wanted to the old/new editing problem. And it is a solution that at the same time works extremely well, is quite elegant I believe, and can't be implemented if you don't build something like DwarfStar. Thread (but check [upto] in the screenshot).
The Eternal Sloptember geohot.github.io//blog/jekyll/u…
