@1dolinski
Seems we’ve achieved error handling, job oriented, recursive AI
ATM seems most measurements flex total time running
The ideal measuring stick is token count, speed and capability
Parallelize and increase compute to buy time
0 reply
1 recast
3 reactions