As you may have surmised (or can see on my bio), I’ve been doing this a pretty long time by now. Coming up on 30 years, actually (i.e. being a technology professional, although I was an amateur tinkerer before that). One thing that’s interesting when your career is long enough, is to be able to look back and see patterns and trends. There’s almost always some new technology concept that has appeared and is marketed and touted as a solution to many existing problems. To name a few: blockchain, 3D (as in television), NFTs, cryptocurrency, quantum computing, XML, microservices, etc. Some of these have been “sticky”, and some have…not.

Artificial Intelligence is not a new trend. In fact, I studied AI foundations when I was a computer science graduate student 20 years ago. But in the last year, the commercial success of finally-useful Generative AI, especially now that we have the compute horsepower to handle large-enough parameter sets, has joined the list of “new technologies that are now marketed as the solution to everything”. So I do not say this lightly: why does GenAI feel different? Why does it feel like it might actually live up to the hype?

I spent some time in “stare out the window” mode pondering this. So first of all, one thing about a lot of these ideas is the foundations may not have been there to make the ideas viable at the time, and then eventually the technology caught up. Blockchain might be an example of that. One of the reasons it maybe hasn’t caught on is because when the chains get long enough, it becomes noticeably non-performant. So maybe when our horsepower improves, this is will become viable. But I think another, more prevalent reason, is simply that blockchain doesn’t necessarily solve enough existing problems. What problem does blockchain really solve? And I’m talking about end-user-facing practical problems, not backend implementation ideas?

When you start to look at Generative AI though this lens, it reveals itself in a unique way. First of all, the tech is finally there. We have enough compute and memory to be able to train and query these super-large models with the performance needed for it to be useful. But also, again a more prevalent reason, is it can be used to solve practical, end-user-facing problems. It’s not a “hammer looking at everything like a nail”. It truly can improve the user experience and quality of life for a great many things.

Let’s take a simple example. People used to semi-joke and say things like “I don’t need my refrigerator/toaster/washing machine connected to the internet”. But with GenAI, those are exactly the problem spaces that could drastically improve your quality of life, if implemented with ubiquity. Something as simple as my wife burned her toast in the toaster yesterday because we switched brands of bread and the current darkness setting was wrong. Why not simply implement some AI model with sensors and monitor the toast, and not really have to configure any settings at all?

As of this moment, I’m using the power of technology to manage my grocery list, completely online and real-time cross-referenced against the store’s inventory. Great! But I can’t remember how much milk I have left in the fridge, or if I have a second unopened container stashed away somewhere. Again, ubiquitous AI can help solve these practical problems.

I get it, it sounds incredibly lazy when you say “I don’t want to have to go walk to the kitchen to see if I’m out of milk.” But its not just that one task. Multiply that by the dozens, if not hundreds, of little tasks that you have to deal with every day. If you could relieve some of that friction and background stress in your life, how much would that improve your mental well-being, your stress level, your productivity, your overall health? How many times have you said, “I don’t know what I have in my cupboard, or what to cook, so I’ll just get a fast-food hamburger instead?” But what if GenAI could look at your kitchen, find a healthy recipe based upon your likes and dislikes, and just have that recipe up on your phone so you can instead eat a healthy meal without having to spend all that mental energy?

Needless to say, I spent a long time staring out the window.