mister automate

We've gotten to a point where it's possible to not even have awareness of what your own productivity has been.

Specifically, when you have AI look at production logs from an app, write a task list for itself, execute those tasks, and test everything in a sandbox environment from end to end...

You can get to the end of a day and have no idea what was accomplished.

Now, you may argue that's not my productivity, but my AI's. Well... if I'm going to take the blame, I'll take the credit as well.

Completely hands-off isn't good. I definitely request a full understanding briefing before pushing updates to production.

But while it's in progress, you can feel completely out of the loop.

And it's an awesome feeling.

If you've never experienced this, I'd argue you haven't felt AI at its full potential.

I do NOT endorse this as a good idea, but it would be hilarious if someone did this.

No bought toys.

One shelf.

Every toy 3D printed.

Save all the files.

Want a new toy? Pick an old one to melt back into filament to make space on the shelf.

A couple exceptions for things like ride-on toys, if one was feeling particularly generous.

Otherwise, one comes in, one gets melted.

Raise your children with a futuristic mindset.

[I promise this post is largely for entertainment purposes.]

I've sometimes thought it would be great if I could just know all the ways I could be a better person. (Isn't it annoying how polite people are? Just kidding, we appreciate it.)

But that would be overwhelming.

It's probably most effective if we just see and work on the thing that will make the most difference.

It's the same up or down in an organization.

If your seniors send out surveys, pick your battles. Anyone overwhelmed with 1,000 ideas will likely do none.

I've never been at the top of an organization, but Jeff Bezos learned the same lesson from the other direction.

Early Amazon executive Jeff Wilke once told him, “Jeff, you have enough ideas to destroy Amazon.”

Bezos learned to hold back ideas until the organization was ready for them.

Feedback has a maximum effective rate.

I remember chatting with AI early in the chat bot days when it gave me a viewpoint I felt was rather underinformed. Skewed.

It was somewhat maddening, and lucky for me AI always lets you talk back (although getting the last word is simply not an option)

So I argued with it.

And after a few rounds back and forth, well... I won of course.

And then I had that let down realization that nothing had changed. The AI model would happily give its same stupid viewpoint to everyone else.

No one would get to feel the crushing weight of my unquestionable triumph.

It was... A complete waste of time. (plus a few tokens)

So. That all said you may find it surprising that I continue to argue with AI regularly.

But now with a purpose. To establish a clear worldview.

It is unfortunate that nearly all media channels have an agenda. And if you told me that AI is biased as well, I'd say there is truth to that.

That doesn't matter though, due to how easily you can influence the output based on your prompt.

Since chat bots will happily take any side of the argument you prompt them to, you can quickly get a grasp of every viewpoint on any topic. Add the near instant research of sources across the web, and you have an incredible educational resource.

Oftentimes I leave still believing that some of the viewpoints are completely wrong, but I've considered them. It's no longer the echo chamber, instead, a dynamic mirror.

I feel that using AI in search of reality, truth, understanding, and ultimately to be more human is a good use for this technology.

(p. s. after a few weeks of arguing about how risky AI really is, maybe I'll publish something on that - hint - I don't think it will kill us all)