It only works for languages and frameworks that are already in the training data (duh). It still is mostly useless when you need to create something from scratch in an unstable language.
That, and you can’t also get the amazing results if you’re poor or have bad internet.
Good thing almost all of programming falls into the former. Most of the economy runs on well defined languages. Billions and billions of dollars.
Opus 4.5 and update your priors. This was certainly true >6months back and is no longer the case
I read the same exact thing 6 months ago.
We are using the latest stuff. Our experience is still not great.
Why do you guys always assume we don't as though the oldest models are easy to use accidentally
It’s an easy deflection. Dismiss any opinions because you’re using it wrong or not the latest.
Good for anything >= 1 month old.
Use other nonsense fear inducing argument in the mean time, continue gathering gobs of VC money, get your bag, continue till the bubble pops.
In all fairness, and putting hype and anti-hype aside, I’m really interested to see the actual value of LLM/agent services after the VC money subsidies dry out. Would people we willing to pay for services at 10x the current price?
Not true. I built some tools in Hare, which almost certainly isn’t in the training data to any significant extent. It was more work than having it build Go or Rust, but it got it done. It had to curl the docs a fair bit.
That's true for most people too. You are trying too hard.