Hacker News Reader

In my experience, 4.6, together with the Claude Code improvements was a non-linear event. A threshold was crossed that forced me to review my complete model of genAI.

If your ideas/studies/experiences with genAI for software development and engineering were from before januari, basically /clear and re-init.

View on HN

Many say but I don't agree. It is clearly better now but I had basically the same view on code gen-AI a year ago as I have now. It was obvious even then that LLMs were a big deal. They were really cool then and are amazing now. But some issues are undeniably still there. Maybe they are not a question of some simple quality measure, meaning they might not be solved by simply crunching more tokens with larger context.

I actually finally got started with CC after 4.6. ;)

The output is certainly prodigious (I can do things for side projects that I'd be very unlikely to finish on my own), but it's not a coding prodigy so it hasn't dislodged me from the idea that it's outsourcing.

Overall, yes, quite remarkable.

Reply on news.ycombinator.com