Ask HN: I'm feeling awe, unsettled and disoriented by GPT-4. How about you?

I am feeling relieved that there's a shiny new thing out there to distract investors, opportunists, and potential competitors, so that the rest of us can continue building in relative peace, solving immediate problems that sound less sexy. The innovation windfall will benefit our startup because it will bring better hardware and services at lower cost.

I remember learning that the introduction of the automatic paintball guns dramatically decreased the price of paint. This benefited all players but not equally. The price of paint for players who did not switch to automatic and kept using manually reloaded paintball guns was reduced to virtually zero. New players sprung for the shiny new automatic paintball gun and usually only played a few games before moving on. The regulars improved their skills with manual reload. When they upgraded to automatic they would have excellent paint conservation discipline and arguably would have a lot more fun for much less cost.

I'm trying to avoid giving into hyperbole and over-stating the situation, but clearly something significant is going on. So while I'm not starting to build a bunker in the mountains and stockpiling thousands of rounds of ammo because I fear some "machine takeover" ala The Matrix, I am spending a lot of time thinking about possible scenarios and trying to work on some contingency plans.

Anxious (of the cascading effect on society, geopolitics and our business).

Yeah, that's mostly where I'm at. Definitely concerned about the economic impact, especially when you get into the idea of possible rampant "technological unemployment" and what-not. But again, not trying to be too reactionary, or rush to any conclusions. There have been other technologies that were predicted to lead to rampant unemployment, and things didn't work out like that. So far the old adage has held true that "technologies may destroy jobs, but they usually create more new ones for a net gain in the end." But I'm not convinced that it must be so.

So yeah, I'm not running down to the local community college to sign up for welding classes because I'm worried that all the programming jobs will be gone a year from now. But I am thinking about classes I could take, things I could do, etc. to position myself for other roles in the future. The truth is, despite my job title I don't do much actual hands-on coding these days anyway, and the value I create isn't so much in my ability to code. But I am thinking about how I can lean into that more in the future. Taking classes, reading books, studying things that relate more to the industry I work in, thinking about things like maybe getting a "business analyst" certification, or doing some management training, etc.

And in the end, I did grow up in an environment where my dad built and raced stock cars from the time I was a small child, and a lot of my early memories involve being in a shop, working on cars and doing hands-on stuff. And I did take 2 years of welding in high-school. So if I have to go out and take those welding classes, or auto repair classes, or electrician training or something, it wouldn't be the worst thing I could imagine.

I’m feeling shocked.

This is the biggest technological advance in my lifetime, and it’s clear that we will get several-fold advances in the near term, just from obvious improvements(eg llms reading about how to use themselves based on all the experiments humans are doing now). That’s not even including upside surprises which there will doubtless be. LLMs gaining knowledge of the world through next word prediction was an example of this kind of surprise.

Also, I feel…strangely at peace, and less bothered by the things that usually stress me - achievement, health, the economy, whatever. The moments we’re living in and about to live through are going to be so unimaginably great(good and/or bad) those all seem so inconsequential.

I posted in length in another thread... to sum it up, I ended with "It's a big deal."

I'm also in my 40s, and I have similar thoughts to yours.

My latest thoughts have been about thinking that we need to take a hard look at what software teams are made off nowadays.

Redraw software teams from scratch: I think 2-3 devs can make the most complex project happen with AI assistance now. I never understood startups with 10 front-end devs, wtf

And another "lead dev" needs to be dedicated to meetings/requirements gathering and stakeholder management. BAs don't work. They can't tell a mouse from an elephant in meetings and most solutions end up strapping more rockets to the pigs until they fly.

Large teams don't work. It doesn't make the software better. With AI's assistance, we can hopefully ditch large teams.

I don't think developers have to worry, they'll just become less specialised as they can rely on AI for a lot of context switching.

What I'm glad I have is my design skills. I've learnt design early and what good UX means down to the gestalt. That is skills that will never leave me and always have provided value... Humans don't like change, so design at the base, is very much linear growth in scope.