

I agree with most of your points, I’m a software dev, same age and am using AI for writing/outlining boiler plate code, like unit test, translation resource files (which will be later filled by human made translations) and sometimes to do simple task in parallel, while I work on the more tricky parts of the code. So it’s a help, but it’s limited. In some areas where I have very deep knowledge, I immediately spot issues, then take a lot of time to refine prompts and tweak the agent and I still get mediocre results. For someone with average knowledge the results look plausible, even good, but as an expert I always see countless flaws. I don’t think this will improve much as we’re around a local optimum of AI and don’t expect any improvements for quite a while. That’s why I gave up on AI for anything complex and only use it to do the most boring stuff for me.




This means they connect the government ID with the AI prompts. So this is a major privacy fuckup, adding a nice method of surveillance to the toolbox of the government.
If it was really about education they could have created randomly-named accounts and each person would get an account randomly assigned.