I was raised SDA; this mindset goes back hundreds of years. Leading up to the 1844, "Great Disappointment", people stopped learning, stopped growing, stopped saving, etc.

[0] https://en.wikipedia.org/wiki/Great_Disappointment

Even if AGI is inevitable, its timeline and effects on your life are unknown.

You are alive. You are alive. You are alive.

Concerning AI, how would you know the output is correct if you have no domain knowledge? Even without domain knowledge, how would you think critically about the AI's output if you haven't honed your critical thinking skills?

There are many cases where AI output is nominally correct under casual examination. This makes sense, because it is designed to generate the most probable output. There's a George Carlin quote that applies here.

When crowdsourcing knowledge, we can reasonably expect that the most probable result will be mediocre. Even if we qualify for domain experts as training input, the output will be mediocre among the domain experts. From there we can also ask about how the expertise is qualified or if political or toxic feedback mechanisms like institutional group-think play a role.

Without honing your mind, you'd be unable to best utilize AI as means to achieve your goal. Problem solving as a skill starts with being able to identify and describe the problem coherently. This is a prerequisite to choosing a goal or attempting to develop a solution. Even if AI is an amazing tool, it will only be as valuable as the ends you apply it towards.

"Everybody wants to be a bodybuilder, but nobody wants to lift no heavy-ass weights." -- Ronnie Coleman

* People that can actually program. These are people who can write software.

* People that can do a little bit more than copy/paste. This is most people in software. These people cannot measure things or write software, but they can follow instructions and reproduce patterns provided to them. These are your React developers who perform a lot of mental arithmetic about stupid shit like state management and yet struggle to put text on screen.

* People that cannot write software.

LLMs clearly widen the gaps between those three camps. People that can write software will continue writing software. People that pretend to write software will become more reliant on LLMs. People that cannot write software will continue to make bullshit assumptions that don't apply to anything. AGI will just be LLMs + 1.

Gene line perpetuation. As underwhelming as it is, you still have a penis, which women, try as they certainly will, will have a difficult time replacing with silicon during your lifetime, and probably for a few generations more.