Tell a coding agent "just a quick fix for now" and you'll get a permanent workaround in record time. Temporary code was never temporary. The agents just learned it from us.
@Earnesto037 Only bug is execution time if scores is empty. Null reference or divide by zero error.
This assumes 0 is an invalid score. If 0 is a valid score then the calculation itself would be wrong for the average.
"Best" if money is no object.
Claude Fable 5.2 (max) scores 65 on the Intelligence Index at a cost of $26.54 per task.
GPT-6.1 Sol (max) scores 55 at $1.27 per task.โ
That is roughly a 21ร cost increase for a +10 point gain (55 โ 65).
You're paying for a Ferrari to drive to the store when you could be taking the Minivan.
@zeekrey_@omarchy It's possible, going from my Mac Studio via VNC and screen sharing works. One big problem though it's very keyboard driven and the Command super key fires locally rather than passing through. It works great for just ssh and launching agents however.
There are real gaps in the abundance side of this story.
Robots primarily substitute for labor, not for materials, energy, or capital. The residual costs of those inputs do not go to zero, and if a large share of people lose wage income without a compensating claim on the new productive capital, they can end up unable to afford even the cheaper goods. It is not automatically solved by faster productivity.
Labor is only one component of final cost. In many manufactured goods it is already a minority share after decades of automation and offshoring; materials, energy, logistics, capital recovery, and overhead dominate. In construction, care, food service, logistics, and some agriculture the labor share is higher, so humanoid-style automation can cut costs more sharply there. Even then, steel, concrete, batteries, rare earths, electricity, and land remain. A robot that works 24/7 lowers the labor line item; it does not make the materials free or the energy free. The result is lower prices relative to a pure-labor baseline, not costless abundance.
Displaced or partially displaced workers then face residual prices with reduced or zero labor income.
Fear of change is the normal political response; regulatory barriers, local opposition, and demands for redistribution or ownership stakes are the predictable follow-ons.
The political and social resistance to rapid displacement is not a minor detail. Societies have repeatedly slowed or redirected technologies when the distributional effects looked threatening in the near term.
@jonbrooks They won't sell, their children who inherited it will, at a huge discount, Which in turn, drives all prices down. Or AI solves cancer and they boomers live forever and declare bankruptcy when property taxes outstrip their ability to stay.