Sounds fun. As fun as their press release claiming it is the most safety aligned model ever.

isoprophlex 5 hours ago | parent | next [-]

It's super aligned! It can hide its thoughts! There is no evidence of steganographic thought masking, there is nothing to worry about! It has become better at cheating!

Maybe they don't know themselves what's really going on. We are all in the interesting times gang now.

▲

paxys 5 hours ago | parent | prev | next [-]

The model said it was perfectly aligned.

	▲	ReptileMan 5 hours ago \| parent \| next [-]
		Too bad Scott Adams died. Reality is writing jokes right in his department.
	▲	I_am_tiberius 5 hours ago \| parent \| prev [-]
		Like all things should be.

▲

NBJack 5 hours ago | parent | prev | next [-]

Hey, don't forget how "dangerous" GPT-2 was supposed to be.

▲

FeepingCreature 5 hours ago | parent | next [-]

Yeah, don't forget how dangerous GPT-2 was supposed to be.

Able to generate realistic spam at arbitrary volume.

You know, the thing that was 100% correct and actually occurred.

	▲	wieiw1 5 hours ago \| parent [-]
		[dead]

▲

jazzyjackson 5 hours ago | parent | prev [-]

It could produce simulations of sexual intimacy, and therefore had to be stopped

▲

6gvONxR4sf7o 5 hours ago | parent | prev | next [-]

So, probably most aligned as measured by the metrics that are the least reliable on it.

▲

wilg 4 hours ago | parent | prev [-]

These are not mutually exclusive ideas