- Mar 22, 2008
- Reaction score
- Vendee, France
Seems OK after a quick look... it's missing the actual numbers, and it used VESSEL instead of VESSEL4, but on the other hand it wrote comments, so that already makes it better than 95% of programmers...
Yeah, it also comments a lot after the code as well. It keeps saying that it can't know the actual numbers, so it can only make so much, but that would be about it. It's weird , because it says it can't remember previous conversations, but when I started one from a different device etc and asked it to code for the Mavic, it started using the code for the Tello, so obviously it still thought it was the same thing and had to be reminded. Then, after it did write the right code, it wouldn't give me all of it exactly because it realised we're talking about a different , bigger quad (as 'big' as a Mini is, it's stil way more capable than the Tello). Now, I'd ask it for more, but it locked me out for one hour when I tried to put it through a Turing test .It failed one of the questions spectacularly, but in away that wouldn't be too surprising even for a human to fail. But asking it iffy stuff often ends up in getting a 'network error' or getting locked out etc.
edit: @Urwumpe Would be nice to know what the output would look like after that feedback, but it may take a while