What does “PhD-level” AI mean? OpenAI’s rumored $20,000 agent plan explained.
Silicon Valley may value imperfect virtual PhDs more than universities pay real ones.
It means AI can recite information from a domain that PhD-level people are concerned with. This doesn't mean it can draw correct conclusions, rephrase emails properly or do any heavy-lifting like come up with computer code beyond boilerplate templates and tech-demos. It's mainly just hype. AI is useful. But not very bright as of today.
The funny thing in my field is that real life PhD's tend to be worse than Masters-level staff because they've gone so far into theory that they have problems with practical application.
It’s almost like AI is purely marketing and also has problems with practical application.
It means that it has a rich daddy.
It means it took a lot of time and money to gain vast sums of knowledge, but has no useful ways to apply it.
Doing dual studies (a few months of uni, a few months working, getting paid ~1k€/month every month), and asked my colleagues about certain topics on the first day of work.
"Do we need activity/class charts? Boolean algebra?"
They laughed their asses off, and it felt so good to finally have the definitive confirmation to what I preached my mates all along: 90% of what we learn is useless bullshit.
It means piled higher and deeper will become the official definition.
A couple of days ago another Chinese company dropped their general AI agent which seems to blow openAIs plans out of the water:
It's already hitting the mainstream press: https://www.forbes.com/sites/craigsmith/2025/03/08/chinas-autonomous-agent-manus-changes-everything/
Lots of social buzz already saying this could be bigger than deepseek as Manus says in their video that this will be open sourced. So good luck with your money plans OpenAI!
