@nickywonka Yes, and no insurance. Only broke the screen once since the 3G and it wasn’t that expensive to have Apple replace it. Less than 18 years of insurance for sure.
@liljamesjohn@ryanvogel You can invoke this in the middle of a conversation sub agent style (even telling the main agent when it should invoke the sub agent) and then it does it in a token cache efficient way.
@liljamesjohn@ryanvogel For the second one, you could configure an existing or custom agent to use a particular model and then @ that agent specifically. I’ll stick an example opencode.json in another comment that does this with the built in explore agent.
@dixiidev@thdxr@ryanvogel A safety razor polishes bald skin way more that non-follically-challenged people imagine. No need for buff like you do for a car.
@StefanTMD To be fair, the limits on Go are quite a bit higher than the free tier, so it might be easier for a smaller number of people on Go to use more tokens than a lot of people on the free tier.
@Lost_Signal@eastdakota I bet overall CPU utilization at typical hyper scalars isn’t over 50% even with big consumers being more efficient. There is likely a billion CPU cores worth of idle compute sitting out there.
@dillon_mulroy I’m a no-lean-back guy for this reason. BUT I was HORRIFIED recently when the steward told me to put my seat back up for landing. I had apparently accidentally reclined somehow by pushing that little button on the armrest. I had become the thing I despised!