Like I said that's temporary. It's janky and wonky but it's a stepping stone.

Just look at image generation. Actually factually look at it. We went from horror colours vomit with eyes all over, to 6 fingers humans, to pretty darn good now.

It's only time.

▲

leecommamichael 11 hours ago | parent | next [-]

Why is image generation the same as code generation?

▲

dcw303 10 hours ago | parent | next [-]

it's not. We were able to get rid of 6 fingered hands by getting very specific, and fine tuning models with lots of hand and finger training data.

But that approach doesn't work with code, or with reasoning in general, because you would need to exponentially fine tune everything in the universe. The illusion that the AI "understands" what it is doing is lost.

▲

rvz 8 hours ago | parent | prev [-]

It isn't.

Code generation progression in LLMs still carries higher objective risk of failure depending on the experience on the person using it because:

1. They still do not trust if the code works (even if it has tests) thus, needs thorough human supervision and still requires on-going maintainance.

2. Hence (2) it can cost you more money than the tokens you spent building it in the first place when it goes horribly wrong in production.

Image generation progression comes with close to no operational impact, and has far less human supervision and can be safely done with none.

	▲	raw_anon_1111 an hour ago \| parent [-]
		This sounds like every system that I didn’t write completely myself and honestly some that I did

▲

mr_freeman 4 hours ago | parent | prev [-]

> Just look at image generation. Actually factually look at it. We went from horror colours vomit with eyes all over, to 6 fingers humans, to pretty darn good now.

Yes, but you’re not taking into account what actually caused this evolution. At first glance, it looks like exponential growth, but then we see OpenAI (as one example) with trillions in obligations compared to 12–13 billion in annual revenue. Meanwhile, tool prices keep rising, hardware demand is surging (RAM shortages, GPUs), and yet new and interesting models continue to appear. I’ve been experimenting with Claude over the past few days myself. Still, at some point, something is bound to backfire.

The AI "bubble" is real, you don’t need a masters degree in economics to recognize it. But with mounting economic pressures worldwide and escalating geopolitical tension we may end up stuck with nothing more than those amusing Will Smith eating pasta videos for a while.