Humans must not apply

I love AI; it's the most important invention of the century. It has already unlocked so much creativity and technical progress, and what's next is yet to be seen. It's mind-boggling that we have categorized all existing proteins and can create new proteins from scratch. Still, AI brought up existential dread this morning when I watched a video that should have done the opposite.

I use AI daily as a software developer. And no, this blog post isn't one of those "AI is taking our jobs" rants. I'm already aware of the threats and the benefits of the AI advancement. I believe that alarmists are overthinking as long as we don't give AI any access to nuclear weapons (and most of the control centers should be air-gapped anyway). However, there's something deep inside of me that is ringing a bell.

I consider myself a well-versed person. Yet, what do I do that AI won't be able to accomplish better in the next 10 years? Sure, I read a lot (50-100 books per year)—but AI can do this so much more efficiently already. In fact, it's been doing this when humans train it: read books to acquire knowledge. AI is much better at recalling than I am; hence, it doesn't need to run a book review blog.

I code for living (well, not for living; I guess I code more for fun nowadays), and I don't foresee AI replacing this part of my job anytime soon. If it does, I'm all for it—happy to pay $1000/month to get a team of AI developers spitting out crazy ideas of mine. Hell, I've written a blog post on how good it would be when AI replaces politicians!

I write blog posts occasionally and use Grammarly heavily to check my tone, punctuation, and grammar. Even my book was fully translated into English with the help of OpenAI (yet all the research and text were originally written in an old-school way by humans).

Yesterday, after watching the Kurtzgesagt video on the topic, my wife asked whether mussels are safe to eat from a scientific perspective. After all, the source we respect claimed at the end of the video that eating mussels is ethical.

I had zero idea about mussels, only having done research on fishes. My next move? I decided to get the deep research option of ChatGPT for a test drive! In around six minutes, it came back with a list of references and simple conclusions. Yes, technically, it's safe to eat farmed mussels from a safe source in some countries. It still contains heavy metals and harmful chemicals, but our bodies should be able to filter out this amount if we eat mussels only occasionally.

Another concept occupies the back of my mind constantly. Technically, AI still lacks consciousness, yet consciousness might be an evolutionary disadvantage when expanding to outer space. Consider the aliens from Blindsight (please read it if you haven't yet) or Expanse. They have no consciousness, yet they've achieved feats that humans probably won't ever be able to.

If AI can do many things I do daily better than me, and if AI can do many things needed for human survival and expansion better than humans.

Then what's the point of humans besides building better AI?