Such a great comment, and I agree with all of them.
For me in a similar vein:
- mar ‘24: thinking about how to survey the field and implement a hard research task in Natural Language Processing, and then just approximating it well enough with a prompt and a completions api
- mid ‘25: Llama 3 being able to analyze a good sized codebase I was onboarding onto, and synthesize it into diagrams that matched the quality of ones I’d generated by hand with deterministic tools.
- dec ‘25: opus 4.5 basically generating multi-class modules and tests perfectly (syntactically). Finding that errors were my own under-specification of the prompt. Stopped writing code by hand, mainly because it was good enough and came with tests, docs, build scripts, and other goodies for free.
Could you elaborate? :) Because otherwise, you're possibly sending the same message as what I alluded to (i.e., we feel great because we're expecting a big exit).
I lived there this summer and fell in love with SF, aside from the whole Silicon Valley thing. It's definitely for a certain type of person: you'll find incredible diversity and openness in the people. A couple of small things I noticed that really pleased me: (1) interracial couples are incredibly common, more so than any other city I've visited or lived in anywhere in the world, (2) people are genuinely open and friendly; the GLBT friendliness is just a specific effect of this, (3) you have incredible high-end dining near the Embarcadero, and incredible low-end dining in the Mission, (4) no matter what your "scene" is, it's quite likely that you'll find like-minded people, (5) if you want nature, drive 20 minutes or less; if you want tech, drive an hour; if you want city, drive 5 minutes anywhere; if you don't have a car, use Zipcar, (6) the city is incredibly small, which makes it a lot of fun. And some things I absolutely hate: (1) parking sucks, (2) the cops on the MUNI/BART are a bunch of assholes, (3) cabs are sometimes difficult to get.
People complain about the fog, but there's really never any fog in the south bay area/Silicon Valley.
For me in a similar vein:
- mar ‘24: thinking about how to survey the field and implement a hard research task in Natural Language Processing, and then just approximating it well enough with a prompt and a completions api
- mid ‘25: Llama 3 being able to analyze a good sized codebase I was onboarding onto, and synthesize it into diagrams that matched the quality of ones I’d generated by hand with deterministic tools.
- dec ‘25: opus 4.5 basically generating multi-class modules and tests perfectly (syntactically). Finding that errors were my own under-specification of the prompt. Stopped writing code by hand, mainly because it was good enough and came with tests, docs, build scripts, and other goodies for free.