I gave Opus 5.5 and Astra access to my computer to play Polytopia. They were both pretty bad (granted, I did have it at max diff and 11 other tribes), but their personalities while playing were pretty interesting.
To start, Opus was much faster at computer use than Astra but wasn’t really making any better decisions. Opus also wanted to repeatedly give up after finding out how hard the game was. There were numerous times when Opus would give up, even just a few moves into the game, saying things like “this isn’t realistic for me to win,” and completely stop playing; then I’d have to repeatedly tell it to get back to playing and stop giving up. It would also try to offer peace to every tribe it knew, even though there was little benefit to doing so, and generally avoided any aggressive tactics.
Astra, on the other hand, was much more aggressive. Astra was way slower, but generally performed better (though this may have to do with other factors like starting position) and would play the games out to the very end. Astra would not offer peace and would say things internally like “I will finish this exposed enemy warrior” or “I will wear down the approaching giant before destroying him”
I attached some of the gameplay, but I had to cut out the parts when Opus would rage-quit the game to avoid showing my desktop.
opus is the left video and astra is the right
if this doesn’t wake you up to the fact that we have officially unleashed a new scaling axis I don’t know what will
whenever log plots break along some stable metric but overall progress feels like it’s improving at same expected rate, this is a good sign you’ve got a new scaling axis. We’ve seen the same thing with Moore’s law over and over