FORD: They make a big thing out of the ship’s cybernetics. “A new generation of Sirius Cybernetics Corporation robots and computers, with the new GPP feature.”

ARTHUR: GPP? What’s that?

FORD: Er ………. It says Genuine People Personalities.

ARTHUR: Sounds ghastly.

F/X DOOR HUMS OPEN WITH A SORT OF OPTIMISTIC SOUND

MARVIN: It is.

@airspeedswift @troz The problem with LLMs has NEVER been the simulation of human personality. How will this make their results more accurate or useful?

@michaelgemar @airspeedswift @troz In a way, this is the mistake in the prediction of AI in science fiction, and fair enough. We never had computers up to now being “inaccurate” and “can’t count”. The human fallibility and imitation was the last thing sci-fi authors would have expected to be implemented.

In a sense, OpenAI’s press release reflects the “known” expectations of AI historically, fictionally. As you say, it’s not even the challenge.

@michaelgemar @airspeedswift @troz in fact, we’ve got these weird evolution of tropes now:

1980s scifi: the android takes a long time to search a database

2000s fan: “he’s slower than a search engine, duh!”

2020s fan: “how do you even trust his answers though?”

@whophd @airspeedswift @troz It’s really extraordinary that we’ve traded accuracy for plausibility.