Trusting the Process

I’ve come into contact with two types of people angry about AI:

1. Those worried that AI is taking jobs
2. Those worried that AI will ruin the planet

I’ve already talked quite a bit about the first one. It’s true. AI is affecting the job market. No doubt. Just like other technologies. Either figure out how to get AI working for you, or you’ll be working for AI.

Now, the second.

I saw a sign in a coffee shop posted right next to the register. It read “Don’t use AI. AI is literally killing our planet.” Handwritten. Doodles around the words. Artsy.

Independent coffee shop. Another sign saying, “No hate speech.” A community board with notices about yoga and art classes, lost pets, and people looking for roommates. You could buy the locally made art directly from the walls.

You get the picture (or hopefully, buy it).

And also TRUE.

The insatiable appetite for tokens requires more power. A lot more power. More power requires more resources, especially fuel and water.

Microsoft is turning Three-Mile Island back on to supply more power to AI at a cost of over $1.5 billion to restart it and around $15 billion in power purchase costs over 20 years.

So what do we do? Do we stop using AI? If I personally stop, will that help?

I think the path forward has to be a coexistent solution.

How do we find a coexistent solution?

We lean into the tension. Argue about it. Passionately. Hyperbole and protests. Shaming. Doomsday books and books about why it’s all a scam. Dystopian movies and movies about the magic future with AI.

Both sides slinging mud.

I wish that wasn’t the path forward. I wish we could have reasonable and respectful conversations. I wish we could plug an efficient “solve the problem” team into the problem and trust the process and outcome.

Alas, that doesn’t seem to be the way we solve things in our democracy.

Solutions to big problems come from the tension. The inefficiency itself seems to be the track upon which the ideas flow. It’s not good enough. Try again. That’s stupid. Try again. You’re wrong. Try again.

The younger me hates it this way. It’s inefficient. Uncomfortable. Rarely leads to the “right” solution.

But the current me appreciates this process. Everyone gets a say, and what is the “right” solution anyway?

I doubt we’ll ever have the academically perfect solution.

But I’m confident we’ll have the best solution.