This software all but eliminated manual bookkeeping. However, jobs for accountants increased, as more small and medium businesses were able to move to a higher order of thinking about their financials as the labor and expense of basic bookkeeping dropped dramatically.
If we think about software, a basic AI can make a basic utility in one go. But for anything more complex, it requires the persons guiding the AI to have some knowledge of how systems work to help guide it and to actually care about testing it to make sure what it was supposed to do actually happens.
Companies that never dreamed of having a dev team might now justify having a dev or two to make custom tooling for their business. Having dedicated resources to this may be better than having everyone in the company hacking together their own stuff and hoping it all works and integrates together.
Most businesses would be very well served even with some fairly basic automation. But without knowing how to ask the AI the right questions, a non-technical person won’t get that far. Even working inside the tech industry, most people I work with don’t think this way. They look at the problem in front of them, they don’t look at how they can create a system to solve that problem repeatedly. Trying to get people to think this way has been an uphill battle for me. I don’t see the masses suddenly thinking this way because they have a new easy to use tool.
We’ve had plenty of easy to use tools in the past. IFTTT, Apple Automator, Apple Shortcuts, Power Automate, etc. Normal people don’t use any of this stuff, because they don’t think about problems that way.
The awareness that AI brings to the general public, along with the idea that maybe they don’t need a big dev team, could start opening the doors of businesses that are still very manual today.