Ask HN: How are you dealing with AI FOMO? | Hacker News Reader

Ask HN: How are you dealing with AI FOMO?

3 pointsTeMPOraL5 comments

Is anyone else feeling the immense Fear of Missing Out with respect to the breakneck pace of progress in ML tools?

This has been weighing on my mind for the past couple weeks - ever since I got API access to GPT-4 (and simultaneously, access to test GPT-4 deployment at work that's company-approved WRT. handling of confidential data), and found myself... not making much good use of it. Not because I don't know what for, or how - but rather I hardly have time and emotional space for it.

Mind you, I use GPT-4 daily. Mostly through chat interface. I have hundred ideas for how to do better, and am aware of many tools making progress along those lines - but I don't have the time/means to develop or test them.

Thing is, with a full-time job, two small kids, and after a rather tough last year, I'm still buried under a small mountain of personal and professional obligations (and I'm not even counting any self-care / self-maintenance here). I'm slowly digging myself up, but right now, I'm seriously worried that by the time I clear enough mental and temporal space to pause and reflect on the new AI tools, everything interesting that could've been done will have been done by someone - and/or, we'll accidentally launch a proto-GAI and the world will turn upside down.

Are you also feeling like this? How do you cope?

I am not able to deal with it at all. Before the Coronavirus lockdowns I have chosen to divert my career from Software to Hardware design and now, when I am at the end of my academic introduction, I keep on getting distracted by the overwhelming world of Chat AI. I feel like the written word by students has completely lost their meaning in academia, as everyone is using chat for their assignments, so the teachers would rather work with the AI directly than by reading pages of hallucinated sentences submitted by a student.

Focus on yourself and your family. AI will be there to learn about later if you are interested.

And inoculate yourself against AI x-risk nonsense. That’s not something you should be worrying about.

Thanks!

> And inoculate yourself against AI x-risk nonsense. That’s not something you should be worrying about.

I'm actually a "believer" in the "AI x-risk nonsense", having self-converted over a decade ago into a sympathizer / non-practitioning follower of the cult of Yudkowsky and his LessWrong disciples. Religious metaphors aside, over the past 10 years I figured out a (what I think is) healthy attitude to this - I am worried at the intellectual level, but I'm not feeling anxious or terrified of it on a daily basis.

I suppose it's the same attitude as everyone has to the threat of nuclear war.

There's a temptation to try to control things you have no power to change.

I suggest reassessing the reasons you hold your beliefs regarding Yud/LW's AI x-risk extremism, especially in light of the fact that nearly every prediction they've made has been wrong. They have a stellarly bad track record for AI related predictions, and base their conclusions about inevitable AI doom on incomputable, unrealistic models of how AI works. All their predictions about a singularity stem from a hidden integral over infinity within Solomonoff induction, and not realistic models of how AI actually work. Get rid of the bad model and so too go the repugnant conclusions.

We've had artificial general intelligence since 2017. It is changing the world, but not in any way resembling the Big Scary Idea.

Reply on news.ycombinator.com