[Nevermark]

>> It is more probable than not that, within the twentieth century, an ultraintelligent machine will be built and that it will be the last invention that man need make, since it will lead to an “intelligence explosion.”

> Well, that didn't happen.

Predictions are not false because their claims are not in the past yet. That bar would make all predictions wrong.

The faster things happen, the higher the speed-of-progress expectation bar gets raised. This is how objectively compounding progress gets interpreted subjectively as linear, or even as a stall. Despite the dramatically increased rate of progress compared to the context of decades or centuries of speculations about cognitive machines.

Models being used to write a lot of the code for new models is a strong suggestion of compounding capability. With new models achieving higher scores. Not proof, but a high bar for evidence that we may be in that explosion now.

The fact that models trained to match SOTA model behavior (i.e. distillation), now learn much faster and more cheaply, than models trained on human behavior, is also strong evidence that capabilities are compounding.

[AnimalMuppet]

>>> It is more probable than not that, within the twentieth century, an ultraintelligent machine will be built and that it will be the last invention that man need make, since it will lead to an “intelligence explosion.”

>> Well, that didn't happen.

> Predictions are not false because their claims are not in the past yet.

"Within the twentieth century" places it definitely in the past, and definitely false.

[Nevermark]

> It is more probable than not that, within the twentieth century

Ouch! Well ok then!

Zoom out in time, and the explosion started in the 20th century. But the pop, while near, has yet to come.

Mea culpa.

[sethev]

I don't think it's a knock on the original prediction to say it didn't turn out to be true. In fact, at the time it probably felt like a pretty conservative prediction. That being said, if you think back to the year 2000, we were pretty far from building such a machine at that time.

To me it's an interesting lesson in how hard it is to predict technology advances which still applies to our own predictions of the future.

[NitpickLawyer]

> Not proof, but a high bar for evidence that we may be in that explosion now.

I agree, that's why I said "we're seeing things we have working towards this". I think that the "jagged intelligence" that is often used to describe our curent models is confusing a lot of people. On the one hand you have models failing basic "trick" questions that a 5yo would get, but on the other hand you also have models that write better/faster kernels that can run faster/better models, and are currently used to improve the next generation (via dataset prep, filtering, RL environment creation/curation and so on).

[wasabi991011]

The twentieth century is in the past.

[bigmadshoe]

Are you confusing 20th century with 2000s?