mister automate

You don't have to try every food.

If you are familiar with the ingredients, your brain can crank through the numbers to give you a hint about whether you will enjoy a different mix.

Peanut butter burgers, salted watermelon, Oreos & salsa... Whatever you want.

The prediction won't be perfect, but your taste and level of gustatory pleasure isn't an exact science either.

What intrigues me about flavor math is (well first of all I've never heard anyone discuss it, and secondly) it's a mix of science (particles + cooking strategy) and humanity. Taste.

I'm seeing a similarity in trying to predict the future of everything post AI.

For example, CGP Grey predicted in 2014 that as soon as driverless cars got better than humans, society would quickly swap over.

As it plays out, when hard facts and data meet humanity, forecasting isn't so simple.

What will life be like when having a personal AI team is a societal expectation as much as owning a smartphone is today?

Where will SaaS land when AI builds and uses its own free version?

Will parents really tell their kids to become plumbers?

We've had the human side for a while now, but AI has introduced a whole new flavor which will get mixed with everything.

So if you want to predict, you'd better be good at flavor math.

What if vendors never stated prices, and customers decided how much to pay after the work was done?

Vendors would choose who to serve and when, based on reviews left by other vendors. Pay well, and you'd earn a reputation that made others eager to prioritize your requests.

If someone's driveway was covered with leaves, you might clear it without asking. Worst case, you don't get paid. More often, someone would request the work.

In a way, this is what the tipping economy implies. We're not getting paid enough, so pay whatever you want. But when everyone asks for tips, it gets annoying.

Reverse Economy changes the incentive. Instead of vendors asking customers for more money, customers have a reason to be more generous.

Imagine getting refused service at a restaurant because of your bad reviews from vendors.

It's interesting to think of the customer as the service provider.

What kind of customer are you? How well do you pay? How quickly? How easy are you to work with?

It's a side of the economy we rarely think about.

Even with prices, what if we focused less on getting the best deal and more on being the customer everyone wanted to serve?

Partially inspired by this interesting article from Seth Godin: Hospitality in the Mirror.

Someone had a video playing the other day, which I was within hearing proximity of.

In my half listening state all was well until a simple metaphor raised it's head.

Nothing to the extent that proves the production method, but I was instantly slopspitious and annoyed.

We are now blessed with a new tax by which one must analyse the depth of thought that went into any digital product that we give our time to.

I use AI a ton, and have no problem with others doing the same. What's annoying is when laziness wins over quality... Refinement IS the new work.

The sad part is that it ruins it for everyone else. Someone may have put hours of work into that video, but due to AI's tendency to state the same thing is different ways several times, now a simple metaphor makes an entire video annoying.

This article was written by hand laboriously in a dark cave. Just kidding, but it is 0% AI assistance.

The entire premise underlying figma is that setting up a prototype of the look and feel of an app or website is more efficient than redoing the underlying code every time you want a design change.

For anyone who's noticed the talk around AI coding, 'setting up a prototype' is precisely the part of the process that AI is fantastic at.

The implementation of design feedback can now be completely automated, and nearly at the pace of feedback.

Instead of meeting with humans to discuss the design of your next digital asset, why not put Codex or Claude on voice mode and knock out 20 iterations in one session?

The prototype and the real deal can be the exact same thing nowadays.

Elon Musk has an interesting 'Five step algorithm' where deleting parts or processes is a key part of improvement.

If it were my vote, deleting Figma would be near the top of the list.

Sorry figma. Nothing personal, I promise. I'm just a systems observer.