Every now and then I have dinner with some old school friends, and inevitably one of them will ask me a pointed question about how it is that I’ve been working in this industry for over fifteen years and yet I haven’t done a startup and gotten rich yet. This is sort of a running joke from our school days when I would tell anyone who would listen that my goal was to own my own tech company. This, of course, was back when I had no idea what running a company even involved, nor what said company would even do.
(Much like everyone else at that age, I also wanted to make video games. This is something I’ve soured on since then, especially after I learned about the awful work conditions at so many studios, and the significant pay difference versus more sedate tech industry jobs. I still want to try it, but I’m not expecting to make a career of it. Anyway, I digress)
I think I’m still open to the idea of doing my own thing – I mean, I did go to business school – but I’ve become much more cynical about the tech industry as a whole, because I started looking a lot closer at what my childhood role models have spent their time doing and saying.
Almost without exception, they appear to be terrible people.
The obsession with shoving generative machine learning models (not calling it AI and you can’t make me) into everything is the most recent example of this, but you can see this cycle repeating time after time. Social networks. The gig economy. Blockchain. NFTs. Subscription services. Large language models. The motivation always seems to be “how can we shove this into the thing we already do in order to make it looks like our company is innovative so we can make a line on graph move up and to the right?” No consideration of ‘is this something that our users actually want?’ or ‘does this actually make our product meaningfully better?’ or ‘hey would this maybe have a detrimental effect on society?’
I remember a few years ago when Kickstarter, of all companies, announced that they were pivoting to the blockchain and while this was met with a lot of breathless media coverage at the time, there were also a significant number of people scratching their heads going ‘WTF?’ A blockchain is nothing more than a huge distributed ledger – the world’s slowest, most inefficient database that you can never delete anything from. Kickstarter’s goal is ostensibly to allow teams with product ideas get early funding from people bought into those ideas. It was never clear how rebuilding the platform on top of a blockchain would serve that goal in any form, and indeed might jeopardize their progress towards that goal by wasting resources on a dumb technology pivot instead of, I don’t know, building new features.
Kickstarter quietly abandoned its planned migration in response to the backlash from its user base, and its CEO resigned a short while later. Subsequent reporting seems to have uncovered the fact that this pivot attempt followed an investment from the Andreesen Horowitz (aka a16z) crypto VC fund. Shocker.
This sort of thing keeps repeating itself with different technologies and paradigms. Large language models are the current hot thing, and predictably, everyone and their mom is jamming them into everything in search of a use case. They haven’t really found one, outside of the things LLMs were already good at (more natural machine translation and generating reasonable boilerplate text).
They have also given birth to perhaps the dumbest job title I’ve ever heard of – the ‘prompt engineer’. I was already a little dubious on the term being applied to people who write software rather than design physical things, but apparently figuring out how to massage your query so that the juiced up nondeterministic autocomplete machine gives you exactly what you think you wanted counts as ‘engineering’ now.
Ultimately, all this comes down to people trying to make fetch happen – to convince people that some technology is the future so they will actively invest in it, ideally with a company that they own so that they get to reap the maximum benefit. And if the technology ends up not being all that? Doesn’t matter, we’ve already made our money on it. On to the next thing! Pay no mind to the people whose jobs got eliminated for no reason, the blatant theft and ignoring of creator rights, the sky-high energy and emissions cost, or the fact that after being directly responsible for so many empty hype cycles, we should have absolutely no credibility left whatsoever.
To me, engineering (software or otherwise) is about solving people’s problems. You start with what someone needs, analyze it, figure out how best to solve it and deliver something that hopefully makes their life better. Starting with a technology and recklessly cramming it into everything in search of a use-case is just ass-backwards, and I don’t think anything will change my mind on that matter.