The whole topic of generative AI has been a lot on my mind for the past couple weeks, even apart from just having to sigh and say “this sure sucks.”
I’ve been meaning to write a follow up that goes deeper into a couple of ideas, but like everybody, I’m kind of tired of it. So I’ll take advantage of the fact that it’s Tuesday and I can give the digest version.
First up is Missy Elliott asking “Is it worth it?” on “Work It.” I’m pretty sure that she wasn’t talking about considering the real costs of technology, but it seems apt now. There’s not a lot that we as individuals can realistically do to push back against harmful tech being used irresponsibly, except to refuse to use it ourselves, and to refuse to do the billionaires’ work for them by evangelizing it.
I think a lot of people still using it and defending it believe that they’re taking a reasonable middle ground. Especially in the realm of code generation, I’ve heard several examples along the lines of “I didn’t believe the hype either, until I used it! It’s incredibly powerful!” Some of it, I’m sure, is an earnest attempt to benefit the people who have rejected it, on the assumption that we just haven’t used it recently. Or that we have inaccurate assumptions about how well it works, based on higher-profile examples of stuff like chatbots comically failing to answer the most basic questions.
That’s why I believe we’re at (or more accurately, well past) an inflection point, where we’re more obligated than we’ve ever been to consider the full cost of everything we use. Everything has been allowed to run unchecked and barely regulated, and if it weren’t abundantly obvious before, it’s crystal clear that the people who have spent the last couple decades hoarding obscene wealth aren’t going to be the ones with consciences.
We have to look at everything we use and seriously consider how much it costs beyond what it costs to us. Stop thinking of it as how much you (or your employer) have to pay for tokens, or how fast and/or accurate it is at giving you a result. Instead, think of the cost in terms of everything that goes into training the models and handling the queries, and how that cost has manifested itself in the real world. You’ve got to ask yourself, “is it worth it?” without automatically appending “to me?”
The other aspect is how increasingly I’ve seen people entreating us all to be more open-minded, civil, and respectful when talking to or about people continuing to use and defend tech that we’re vehemently opposed to. The idea that it’s one thing to see some dipshit deep in the throes of AI psychosis going off on rampages, but we need to be more open-minded if we’re going to sway people on the fence.
But do we, though? I’m reminded of the Beatles song “We Can Work It Out.” I’m no longer the Beatles fan that I used to be, but I’ve been repeatedly surprised by just how poorly their songs have aged. And I blithely listened to this one for years before really paying attention to what it was saying.
On the surface, it sounds like a pleasant and positive little song about compromise and cooperation. But it’s actually just a guy saying, repeatedly, that he’s right, and you’re ruining everything by not seeing it his way. The closest he ever gets to genuine compromise is saying “only time will tell” whether he’s right or wrong. In other words: stop being obstinate and just listen to me.
I think it’s a good reminder to pay attention to what actual compromise and open-mindedness looks like. You can be civil, pleasant, and even catchy, while still saying garbage.
Leave a Reply