I doubt that we are entering a world where people will be using AI to write their own little apps, and that is coming from someone who does it.
All you need to do is look at history to realise that. We had things like BASIC, HyperCard, and the early web that attracted droves of people who were excited by the opportunities. While a lot of those people went into industry to write software for other people, a lot did not. Most people, I suspect, didn't even explore the potential because they simply don't think of computers as programmable machines -- even if they were aware that there are people who program computers for a living.
The barrier to entry is lower now because of AI. Even though I code, I ask the AI to write a lot of little helper apps for my wife who homeschools our kids, and I don't code review many of them. I only do enough user testing to ensure they function as requested.
Armed with this knowledge, I have tried in vain to get various family and friends who are not technically minded to do anything like this. In some cases, they just tell me that they can't think of anything the AI can help with.
Many things that seem simple to them are more complex, and they can't tell the difference between simple apps that the AI can knock out in a few minutes and larger projects that require internet hosting and/or vastly more AI credits than they would ever be willing to pay for.
By far the biggest problem is that many of them are not very good at describing what they want. For simple apps, you can go back and forth with the AI to ask for changes, but the more complex the solution, the more its becomes like a traditional engineering design project where carefully though out design of features and their interactions are critical for success. The new "code" may be more like human language, but the user facing details must still be specified.
II2II · · focus · HN ↗
All you need to do is look at history to realise that. We had things like BASIC, HyperCard, and the early web that attracted droves of people who were excited by the opportunities. While a lot of those people went into industry to write software for other people, a lot did not. Most people, I suspect, didn't even explore the potential because they simply don't think of computers as programmable machines -- even if they were aware that there are people who program computers for a living.
wisemanwillhear · · focus · HN ↗
The barrier to entry is lower now because of AI. Even though I code, I ask the AI to write a lot of little helper apps for my wife who homeschools our kids, and I don't code review many of them. I only do enough user testing to ensure they function as requested.
Armed with this knowledge, I have tried in vain to get various family and friends who are not technically minded to do anything like this. In some cases, they just tell me that they can't think of anything the AI can help with.
Many things that seem simple to them are more complex, and they can't tell the difference between simple apps that the AI can knock out in a few minutes and larger projects that require internet hosting and/or vastly more AI credits than they would ever be willing to pay for.
By far the biggest problem is that many of them are not very good at describing what they want. For simple apps, you can go back and forth with the AI to ask for changes, but the more complex the solution, the more its becomes like a traditional engineering design project where carefully though out design of features and their interactions are critical for success. The new "code" may be more like human language, but the user facing details must still be specified.