Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI-first – We're just 6 months away from AGI

revontulet.dev

Post reply on HN
Print view View original

51–52 of 52 posts

Re: AI-first – We're just 6 months away from AGI

#51

Post by Kiro »

creesch wrote: post #26
Kiro wrote: post #17

Earlier quoted context omitted.

I think the article is way more bullish on AI than what the ";-)" and you imply.

The smiley very much implies that the title shouldn't be taken seriously. I'd say it is right on the mark. To be clear, bullish has a few meanings. But they all boil down to assertively communicating. From your comment I gather you might think it's etymology traces back to "bullying", but it actual relates back to "bull".

I'm talking about "bullish" as in "optimistic about something's or someone's prospects" so literally the opposite of what you think I meant. In other words, the article is more optimistic about AI than the smiley implies.

Re: AI-first – We're just 6 months away from AGI

#52

Post by dinfinity »

creesch wrote: post #50
dinfinity wrote: post #49

Earlier quoted context omitted.

> And yes, the argument is always to let them work on small isolated bits of code. Or that if your requirements are tight enough they produce very good code. You're getting there. The most valuable change is using software (like Cursor) that runs the models in agentic mode so they: 1. find the context themselves (with a basic document with context to point them in the right general direction for the current task). 2.…

Right, so now we have moved to "well yeah, but you need to use LLM agents". Do you truly not see how you are actually continuing the trend of shifting the goalpost every time someone is critical about the use of LLMs? Critical about parts that just a few short months ago had the same promise you now moved to the use of agents? Not to mention that with each iteration the amount of tokens needed goes up substantially.…

> Do you truly not see how you are actually continuing the trend of shifting the goalpost every time someone is critical about the use of LLMs

A late reply, but the point is that the old models are harder to get good results with, even in agentic mode. Therein lies the "massive difference in outputted code quality between models" I mentioned.

So using the newest models with the right approach quite easily leads to good results, which is what I meant with: "At this point, if your coding experience with LLMs sucks I'd say there is an 80% chance that you're just doing it wrong."

This as opposed to the original final claim made by the person I was replying to: "Just roll the dice correctly."

That implies that it is just about getting lucky with almost random output. It hasn't been that way for a long time, unless you really try to get the LLM to fail.

> With LLMs you need to effectively babysit them on each project again.

Only if you're doing it badly. Document and point the AI at documentation properly. Your onboarding process and documentation should not rely on people anyway. Write Once, Read Many.

Post reply on HN