For three months, until the next SOTA model drops. And with each improved model, the number of people who actually need it declines.

▲

ronsor a day ago | parent [-]

Exactly. Astra will likely get its Mythos moment, and then another Chinese lab releases its equivalent three months later while Anthropic releases a newer, better model.

▲

dakolli a day ago | parent [-]

There's a reason why most businesses don't hire PhDs, they dont need it. PhDs are cheap in many countries, arguably more affordable than a full time mythos lvl agent, yet businesses in India still don't hire 30k a year PhDs to run their businesses. Its simply not necessary. AI labs have lost the plot from a business perspective. If chasing ASI is the goal, that also seems stupid.. Putting all of your customers out of business with ASI doesn't really lead to great outcomes either.

They're commodity fetishism maxxing.

▲

Game_Ender a day ago | parent [-]

The higher level models are designed to tackle bigger problems with less oversight and more accuracy. It’s about entirely changing how you use them to mirror an employee you can trust. A truly good team members everyone trusts to just get things done at high quality with minimal direction for days or weeks. That is the goal.

	▲	dakolli 16 hours ago \| parent [-]
		https://en.wikipedia.org/wiki/Commodity_fetishism