This is incorrect. If you take the most basic interpretation of an LLM at temperature 0 as predicting the most likely token, and you run it on, say, 1,000 runs of "complete this Spanish sentence with the word for 'X'", then:
- maybe ALL humans would fail the test in some way, eg. let's say everybody gets at least 10 of those wrong, and the average person gets 100 of those wrong.
- still, as long as most people correctly get each word right, your LLM would get every single response correct (because for each item in the test, 900+ people out of a thousand gave the same correct answer in the training set).
In that sense, it's totally possible for a system trained on a vast vat of average-human input to generate super-human outputs.
But still, the questions in that test are "solved" in the sense of "I can take a dictionary and answers these questions with full certainty". Beyond established knowledge LLMs are monkeys with typewriters, at best.
I’d like to see you ace even a middle-school level Spanish test with just a dictionary (sub Spanish with some other language if you happen to know Spanish).
Let's define "zeta" as a mathematical function ζ(s) which takes a statement "s" as input, where s is a statement of a breakthrough in LLM capabilities achieved relative to the current date and time and ζ(s) is the probability that a given AI skeptic will honestly recognize "s" as a breakthrough,
then our Riemann-Goalpost hypothesis is that ζ has zeros for every "s" which is a negative integer (every breakthrough that happened in the past is null in value) and only has positive values where s is positive.
We can conclude from the above that given a far enough date in the future, any given breakthrough can be spectacular, but once achieved, will be derided as trivial.
To this pedantic point, If the average written intelligence of all humans alive and dead is > the max intelligence of all live humans who are also willing/positioned to do the same task at the same time and at the same place.
But yeah, I don't think LLMs (the current core architecture) can provide super intelligence. I think it needs a bit more than next token prediction architecturally speaking.
I'd say that, at least with Tibetan Buddhism, there are things like deity yoga, wrathful and peaceful manifestations, Dakinis and Dharma protectors, Bodhisattvas, all are essentially theistic in nature. Adi-Buddha in the Vajrayana tradition is also a "primordial" Buddha, so it has many of the hallmarks of theistic religion.
There is no cosmic judge, karma works like Newtonian physics, events propagate endlessly through time and will catch up with you (don't know about effect of black holes on karma)
Right, but certain careers have figured out that women are, on average, better at the socialization game and thus, broadly speaking, a better fit for the job. Ain't nobody "diversity" hiring on the oil rigs or other jobs where the physical act is more important than interpersonal interaction.
However, it is difficult to measure those positive traits for what they are, so employers are selecting based on gender hoping for positive correlation. But that's illegal, so "diversity" hiring was created as a scapegoat to help avoid legal fire.
Are you asserting that companies in a free market are intentionally hiring people unqualified for their jobs and paying them wages as if they were qualified?
No, music is not like math in that sense. Music is like math in the sense that carpentry is like music or math, or sports are like music or math. You can't be a spectator, you have to grind thousands upon thousands of exercises to be good in it.
