Discussion about this post

User's avatar
Alice's avatar

Wonderful article! I would like to point out the huge problem with many chuds optimism towards Ai. I have spoken with fellow young chud colleagues that are convinced about Ai not replacing them and not leading to economical collapse. Instead chuds are convinced that Ai will lead to a wonderful universal income where they don’t have to do any work and can just run local LLMs that do their work for them. This is a common tech bro talking point. I think this view is quite unrealistic due to the historical fact why workers were given better conditions and rights in the past. When society industrialised workers were eventually given rights because the country with more content workers could produce more and better resources and in turn MORE workers by having time for children. I just don’t see how this logic can apply for a post Ai fully automated society where workers productivity and the amount of human workers means nothing for economic gain…

Benjamin Eskilstark's avatar

Good essay. I went to Yale for CS (and a double major in History) back in 2012-2016 and was on the leading edge of the new wave of CS-as-lucrative-career rather than CS-as-science: when I was looking at colleges as a high schooler, I saw that Yale had graduated 11 people with CS degrees in 2011, but by the time I graduated in 2016, CS was in the top 3 biggest majors with almost 200 of us. At the time I felt like this was a major bait-and-switch as I'd imagined having lots of 1-on-1 access to professors and peers who were in it for the pure love of algorithms (alas) rather than just startup opportunities or career ladder credentialism. But then again I probably was only even admitted in the first place because Yale made a conscious decision in 2012 to admit way more STEM/CS-oriented people and I likely would not have been in their top 11 applicants. So you take what you can get.

Regarding the history of CS and AI, I have bad news for you: it's not just AI that is this enmeshed with the military industrial complex. Aerospace Engineering might be the only other field that even comes close to being as pure a child of the MIC as Computer Science is. There was basically nothing to do with computers at all other than missile guidance and nuclear bomb simulation between 1945 and 1960, not that people back then didn't see the potential for broader applications, but the DoD was THE customer for software and it only branched out slowly from there: NASA, airlines, IBM. Cybernetics, as you referenced above, was also heavily concerned with military chains of command.

What can we do with this knowledge? Well for myself, at first, nothing. After I graduated I worked at Facebook for 3 years and saw first hand how targeted ads win "hearts and minds" waaay better than guided missiles, and both are just computer programming. But even during that time (2016-2020) I saw a change in the demographics of new engineers there away from people who just really liked computer programming and towards ambitious people who just wanted the fastest track to six figure salaries. (As an aside, how much should we lament that a place like fucking Facebook is now staffed by mercenaries rather than true believers in programming? From what I've heard, the people who work on nuclear missiles LOVE their jobs because if you don't think very hard about what happens after the missiles land, it's a fascinating and stimulating set of problems to work on, and since it doesn't pay ludicrous SV salaries, they maintain a small supply of true aerospace/nuclear "talent". Whereas if your company goes mercenary, like IBM did in the 80s and Oracle in the 90s, then real innovation becomes impossible and you can only continue to milk the inventions of yore until you fade into obscurity. This is my optimistic take on what will happen to fb/goog/etc)

I also have thoughts about the unique appeal of AI to right-wingers, but am having trouble articulating them now and this is already long.

15 more comments...

Ready for more?