> But there was one downside to AI. AI was like a human which is why you could talk to it like a human. But because it was like a human, it could also be wrong like a human. So I had to make sure that it wasn't getting the answer wrong. For which I would sometimes use Google Search. I really don't buy that this was written/created by an actual 9yr old. But this might just be my unhealthy pessimism/skepticism when it…
Both parents are programmers, and have also written blogs. While the motivation/drive is his own, he has helpful guidance to accomplish what he wants to do.
The blog is his own words. We helped with the outline, and also provided multiple rounds of feedback to ensure good clarity of what he's trying to express. We tried not to interfere too much with his thoughts. The quoted thought "because it was like a human, it could also be wrong like a human" is something he was telling us when he discovered it hallucinated. But he doesn't understand what an LLM hallucination is, that was his words. I asked him to share that in his blog post.
The code is written on his own. But when he gets stuck, he has us to give him hints. As programmers, we can speed up him significantly by steering him in the right direction.