One of the bottlenecks right now is testing.

The LLMs suck at testing still, so the feedback cycle still requires human input

▲ imdsm 2 hours ago | parent [-]

That's because testing is a layer above, and that requires effort from people, and people don't want to put in the effort or lack the skills to properly test with them. I think LLMs are superb at testing, but I have a lot of machinery in front of them to enable it

▲ ponector an hour ago | parent | next [-]

LLMs are superb at testing easy things. Small website, simple lib, etc.

Same way as LLMs cannot code anything complex, they cannot test complex scenarios.

▲

falcor84 an hour ago | parent [-]

Where is the "cannot" coming from?

The only underlying rule that I see is: the more complex a task is, the lower the probability of success. That's it. And it's the same exact rule that applies to humans. The difference though is that we're seeing an exponential growth in AI capabilities, which are then rapidly disseminated on a scale of months, and we don't see such capability increases in humans.

As I see it, anyone whose mental model is built on what AI can or cannot do is going to have a very bumpy decade.

	▲	satisfice an hour ago \| parent [-]
		AI cannot take responsibility. It will not indemnify you. My decade is going just fine.

▲ throwaway2037 2 hours ago | parent | prev [-]

    > but I have a lot of machinery in front of them to enable it

Can you share some details?