I think it depends whether you can leverage some knowledge. It's possible for a person/LLM to look at a loss curve and say "oh that's undertraining, let's bump the lr" - whereas a Bayesian method doesn't necessarily have deeper understanding, so it'll waste a lot of time exploring the search space on poor options.
If you're resource unconstrained then BO should ofc do very well though.
Yah, I'm a bit skeptical - ime humans tend to under explore due to incorrect assumptions. Often this is due to forming a narrative to explain some result, and then over attaching to it. Also, agents aren't actually good at reasoning yet.
Good Bayesian exploration is much, much better than grid search, and does indeed learn to avoid low value regions of the parameter space. If we're talking about five minute experiments (as in the blog post), Bayesian optimization should chew through the task no problem.
Been in LA for 2 years now and went to college in LA 10 years ago so have some data points to compare.
Startups in LA are interesting. Back in the day, there was a lot of Ad Tech / advent of big data. This is when Snapchat and Hulu were coming up. I’d go on Angellist and see who was hiring and who’d be down to meet.
Now, especially post covid, I feel sparks of excitement. I missed the crypto hype in LA so that was probably wild and weird. a16z opened an office in Santa Monica and do their speedrun accelerator. Focused on games and media it seems.
upfront hosted some cool cowork and mingle events too.
Two meetups I regularly go to is AI Tinkerers and MLOps. Generally it’s the same small crowd. I went to a Ruby meetup which was cool too.
Less of a young startup crowd. Maybe people got older and rich and retired early.
I stay mostly on the westside but id say more density of shops but at the same time less busy? if i had to guess, things were really going well before covid. then add delivery app and now a bunch of coco delivery robots really slowed down foot traffic.
things still go viral - pop ups are very popular with the young crowd. owalla was giving out water bottles and the line was around the block and i didnt see the end of the line
The tech scene is not doing so well imo. people say gaming is cyclical but idk tbh. AI adoption in graphic design and coding is really going to squeeze an already stressed gaming labor pool. entertainment industry is much the same.
i cant name one young rising startup that was hot like snapchat. so that was disappointing when i was job hunting.
Awesome! Coffee + LLM-assisted programming. I had a similar experience coding an iOS app without experience with objective c using chatgpt. Great learning tool
I'd never heard of Nativescript, but have tinkered with React for the React Native functionality, so maybe Nativescript is something I should look into. Is it pretty mature as far as feature parity with React Native?
I started looking at Kaggle again and autoresearch seems to converge to many of the solution vibes there.
Wild ensembles, squeezing a bit of loss out. More engineering than research IMO
reply