App-ification or Swiss Army knife for GenAI tools?

App-ification" or “Swiss Army Knife”?

This article reminds me of a core idea of mine, one that has been brewing over the course of the last couple of years.

Will AI tools become “appified?” Meaning, will models like GPT and Claude get wrapped into a million little apps that each solve a specific problem? This means that most people won’t be concerned with prompts and chats, but will just use apps the way they always have, this time with a layer of Generative AI as a feature.

(This, by the way, is the current state of the EdTech industry, AI tools and all)

Or…

Will AI tools be more like a swiss army knife? Will we learn to think for ourselves when a problem comes up, the way a developer might write a quick script to scratch their own itch?

Here’s an analogy: for years, I have taught people how to use spreadsheets more effectively. Most people know the very basics before they are trained; they don’t know enough to solve complex problems.

If I wanted to lose weight, I could use Excel to track a moving average of my weight, record calories and set up equations, etc… (swiss army knife model, Excel can do a billion other things)

or..

I could download an app like MyFitnessPal or HappyScale (or both). No need to write my own formulas; the app does all of that for me (appified model).

I believe the world would be better served by the swiss army knife mentality. If an app doesn’t exist to solve your problem, you can “think” your way to a solution, knowing what you know about the all purpose tools such as (Chat)GPT or whatever it may be. We’ll all become more nimble thinkers as a result and less reliant on an app store to solve our specific problems.

But… I’m somewhat certain we’ll take the appified approach, since it requires the least resistance from our current model.

Time will tell.