Artificial Intelligence (AI) has massive amounts of human data to learn upon.

However, future AI gets a mix of human data and previous AI data outputs. This will progressively get worst and worst over time, till the point that AI is learning from it's own outputs, including mistakes.

You will get to the point of what AI generated with DOGE Cuts and Trump Tariffs, which any logical human would understand is invalid and will just cause major issues but AI does not. It will then depend on the human to double check the AI results.

Same deal with programming or anything else.

AI not only has to deal with human mistakes and disinformation, but also it's own recycled ones.

Then you have AI like Grok from Elon Musk on xTwitter, spewing out thousands of private user conversations with it upon search engine results. Everything from generating a password for me, coding and questionable requests.

What would you do to ensure AI results are valid?

Do you even trust AI, specially if not local?

It seems there's the group that relies on it too much, even going as far as using it in their job (such as the White House) and the other which avoids it completely. Or are you a mix?

I personally like the idea of it, but consider it's doomed to failure. Anything it generates I would consider public access and would completely double check myself to ensure it's valid. It's almost a waste of time, because I could of done it myself a lot cleaner.