This is very true, but I don't think this is an AI-only problem. It is fundamentally a human problem.
There are so many things happening around us, and one of the biggest things that separates people is knowledge. A non-technical person who puts in the effort to learn technology becomes a technical person. The only way to become good at something is to actually do it and practice.
Someone who still doesn't know what AI is also doesn't know what is possible in the world today. My mom is a good example. She comes up with a lot of ideas and tells me about them on calls. Recently I told her, "Just tell ChatGPT. It can probably build a lot of what you need." But even that sentence isn't enough for her to understand what I mean, because she doesn't yet have a mental model of what AI can actually do. She simply isn't aware of its capabilities.
I'm going home in a few days and will be staying with her for a couple of weeks. I'm actually going to try an experiment: teach her some of these tools and see what happens. Can someone with essentially no technical background learn to use AI effectively and actually build something meaningful within a few weeks?
She already has a good job and doesn't really need to become technical, so there is no practical necessity for her to do this. But I think it would be a fascinating experiment.
Does it really? Or is it that they're all pretty much good enough and so you just go with what's cheapest or what you're used to?
For example, in IntelliJ I can switch between maybe a dozen models with a drop-down but the only time I do it is when whatever model I'm currently using has an outage.
Another example, our product has some ai features (not coding related). I use spring-ai and so I can change models just in the config file. But I stick with deepseek because it's more than good enough and is super cheap.
That's our moat, right there. I'm starting doing clean-up jobs now, advising non-technical users on where they're going wrong. I had a guy recently who'd spent well over a month on Claude, running up the wrong tree on a research question.
AI appears to be empowering non-technical users, but whether it's making them more productive or actually answering their questions is a completely different matter.
This is very true, but I don't think this is an AI-only problem. It is fundamentally a human problem.
There are so many things happening around us, and one of the biggest things that separates people is knowledge. A non-technical person who puts in the effort to learn technology becomes a technical person. The only way to become good at something is to actually do it and practice.
Someone who still doesn't know what AI is also doesn't know what is possible in the world today. My mom is a good example. She comes up with a lot of ideas and tells me about them on calls. Recently I told her, "Just tell ChatGPT. It can probably build a lot of what you need." But even that sentence isn't enough for her to understand what I mean, because she doesn't yet have a mental model of what AI can actually do. She simply isn't aware of its capabilities.
I'm going home in a few days and will be staying with her for a couple of weeks. I'm actually going to try an experiment: teach her some of these tools and see what happens. Can someone with essentially no technical background learn to use AI effectively and actually build something meaningful within a few weeks?
She already has a good job and doesn't really need to become technical, so there is no practical necessity for her to do this. But I think it would be a fascinating experiment.
I see this too. Generally, having a builder mindeset is independent from the tools available.
Does it really? Or is it that they're all pretty much good enough and so you just go with what's cheapest or what you're used to?
For example, in IntelliJ I can switch between maybe a dozen models with a drop-down but the only time I do it is when whatever model I'm currently using has an outage.
Another example, our product has some ai features (not coding related). I use spring-ai and so I can change models just in the config file. But I stick with deepseek because it's more than good enough and is super cheap.
Subtle Lena[1] reference, maybe?
[1] https://qntm.org/mmacevedo
nice catch haha
that's super true, it;s very hard for non-tech people to use ai at its peak power
That's our moat, right there. I'm starting doing clean-up jobs now, advising non-technical users on where they're going wrong. I had a guy recently who'd spent well over a month on Claude, running up the wrong tree on a research question.
AI appears to be empowering non-technical users, but whether it's making them more productive or actually answering their questions is a completely different matter.
asking the AI helps
That helps, but people tend to ask about uses they can already imagine. The hard part is discovering what’s possible beyond that.
[flagged]