@bigdavetalks@WW3_Monitor Sure sure comrade ๐คญ you are doing good job defending ruzzia but they may not have much cashflow ๐ธ soon. ruzzians were targeting cities from day 0 and it's a delusion that something can "provoke" them
Sonnet 5 is a great model if you want something less smarter than Opus 4.8 but for the same price. Heiku 5 can be less smarter than Sonnet 5 for the same price also. This would make model selection simpler. Sonnet 5.1 can be less smarter than Sonnet 5 and can be a bit more expensive.
Your kid asks what to study. You don't know. Nobody does.
AI rewrote which careers are safe. Three paths still work:
โ Hands. Plumber, electrician, wind-turbine tech, solar installer.
โ Care. Nurse, teacher, doctor, counselor.
โ Kid + AI. Learning to use AI as a multiplier.
Data behind the caveman bit:
BLS 2024โ34 projections (Aug 2025) โ wind turbine techs +49.9%, nurse practitioners +40.1%, both top-3 fastest-growing US occupations. WEF Future of Jobs 2025 โ AI is the #1 fastest-rising skill 2025โ2030.
what's your plan for your kid? Caveman wants to know.
Nobody knows the right way to use AI.
I have been coding 20 years. I use Claude every day. I still don't know if I am doing it right.
And nobody does. Not the AI gurus. Not OpenAI. The industry is two years old.
If you use AI every day at work, you are already ahead. Stop comparing yourself to a fantasy.
Most people skip this AI trick.
Add this one line at the end of any prompt. The model drafts โ reviews itself โ gives you the polished version first.
"Refine result in 3 Rounds: Create draft โ Improve through review rounds to fix weak points โ Present final result first โ Then briefly summarize changes from each round."
Works in any AI.
@brivael The most important component for it is many non-educated naive people to support it at the initial stage so it can grow into humans-killing machine.
Opus 4.7 crashes on commands that worked like a charm on 4.6 - not just performing worse, but simply exit team leader process with red internal exception.