AI Money Should Be Spent on Doing Something Useful

I have talked about this before, but two recent news items bring it back to the fore.  First, OpenAI is on pace to lose approximately 5 billion (with a B) dollars this year.  The Imitative AI industry as a whole shows no signs of being profitable in the near to mid future. On the other hand, California appears to be consistently meeting midday power requirements entirely with renewable energy.  Battery improvements seem to be driving a lot of that success.  Which technology, though, gets the hype?

Imitative AI simply does not deserve the positive attention it gets.  I can run down the list of reasons — the lack of successful products, the environmental impact, the dubious morality of how the models are created, the fact that shoving data into new models does not guarantee those models will improve  — but the simple fact remains that imitative AI has produced no killer apps, nothing that promises to transform, well, anything. Even AI friendly articles admit that companies would have to completely retrain their employees and redefine how their data is stored and labeled to get even moderate use from imitative AI systems.  There is very little actually promise coming from imitative AI.

Batteries and renewable energy, however, have been improving at rates that allow for a complete re-thinking of our electrical grid.  The investments in those areas have produced significant, economically important, payoffs.  Even putting aside the economic benefits, their contribution to helping stave off the worst of global climate change make them morally as well as economically superior investment choices.

But Silicon Valley isn’t great at morality.

Pure selfishness plays a roll, I am sure, but I think the issue runs deeper than that.  These people refuse, on an emotional level, to understand that they got lucky.  The government essentially handed them a platform, the internet, for free.  They happened to be in the right place at the right time to ride the dotcom bubble and the rises of social media.  Someone was going to make money during those gold rushes, and a combination of good timing, possibly good products (though not always) and good luck handed that money to them.  The history of stupid Silicon Valley investments  — things like WeWork, which was literally just office renting with an app — crypto, nfts, metaverse, and now imitative AI stem from the desire to win the lottery once again.

The problem, though, is that there isn’t a lotter being played.  In terms of the internet, there are likely no fundamental paradigm shifts in the near future.  The internet has saturated the world and there are no easy ways to have accelerated growth on it anymore, short of massive monopolies.  Monopolies, as they hate being reminded, are illegal.  But they keep coming back to that well, because it is all they know, really.  They don’t see the amount of good fortune they benefitted from and apparently convinced that they can will into existence the old golden age of software because they are just that smart.  Despite the fact that, collectively, they have crated almost no real value since the early 2000s.  Smarter businesspeople would be looking for newer markets, for riper growth areas — such renewable energy — but Silicon Valley collectively clings to the past with the fervor of desperate nostalgia.

We, of course, don’t have to indulge their fever dreams.  The government can and should use its power to encourage more investment in renewable energy.  It can and should use its power to discourage investing that requires hitting huge on some investments to support all the others.  The government can and should reward companies for making actual money rather than taking in market share.  Just because Silicon Valley collectively cannot face the future doesn’t mean we as a society cannot.