- cross-posted to:
- aicompanions@lemmy.world
- cross-posted to:
- aicompanions@lemmy.world
If this is the way to superintelligence, it remains a bizarre one. “This is back to a million monkeys typing for a million years generating the works of Shakespeare,” Emily Bender told me. But OpenAI’s technology effectively crunches those years down to seconds. A company blog boasts that an o1 model scored better than most humans on a recent coding test that allowed participants to submit 50 possible solutions to each problem—but only when o1 was allowed 10,000 submissions instead. No human could come up with that many possibilities in a reasonable length of time, which is exactly the point. To OpenAI, unlimited time and resources are an advantage that its hardware-grounded models have over biology. Not even two weeks after the launch of the o1 preview, the start-up presented plans to build data centers that would each require the power generated by approximately five large nuclear reactors, enough for almost 3 million homes.
In the meantime weasel programs are very effective, and a better, if less known metaphor.
Sadly the monkeys thought experiment is a much more well known example.
Irrelevant nerd thought, back in the early nineties, my game development company was Monkey Mindworks based on a joke our (one) programmer made about his method of typing gibberish into the editor and then clearing the parts that didn’t resemble C# code.
That’s how they made that ET game back in the day
C# didn’t exist in the early 90s, perhaps you’re thinking of another language?
Obligatory: 95% of paint splatters are valid perl programs
It may have been C+ or merely C with OOP features. I was writing the enemy-AI code (not to be confused with actual learning systems) in visual basic (and made some sweet pathfinding algorithms at the time), but took it too seriously and ended up breaking my brain.
We had a publisher and it was going to be awesome and then Windows 95 came out and broke all our code.