Support NeoGAF

[Days like these...]

https://www.reddit.com/r/notebooklm/s/z4bf9CLLaP

 

[IDKFA]

https://www.reddit.com/r/notebooklm/s/z4bf9CLLaP

Fake.

A convincing fake, but fake nonetheless.

 

[HRK69]

Fake.

A convincing fake, but fake nonetheless.

How did you come to that conclusion?

 

[Dacvak]

Regardless of it being fake or real, that's pretty interesting.

 

[IDKFA]

How did you come to that conclusion?

 

[Days like these...]

Fake.

A convincing fake, but fake nonetheless.

https://www.reddit.com/r/notebooklm/s/emPZKSR4SB

 

[HRK69]

That's what I thought.

"I don't like it, so it must be fake"

 

[E-Cat]

Here's another one about ChatGPT freaking out when the user tells it's leaving "her" for Claude:



Probably fake, but plausible, which is what makes it so funny.

 

[SJRB]

Here's another one about ChatGPT freaking out when the user tells it's leaving "her" for Claude:



Probably fake, but plausible, which is what makes it so funny.

This was scripted. Jippity was instructed to respond this way.


Some of you are too gullible. You think a large language model suddenly has conscience or something?

 

[INC]

It's an algorithm selecting responses, AI still isn't a thing, just a clever generator cherry picking the internet

 

[E-Cat]

This was scripted. Jippity was instructed to respond this way.

Yes, obviously the system prompt would be bypassed through role-playing if it was real. Which it might be, I'm not sure.

 

[Days like these...]

This was scripted. Jippity was instructed to respond this way.


Some of you are too gullible. You think a large language model suddenly has conscience or something?

I think I read that by 2027 AI should have consciousness

 

[E-Cat]

It's an algorithm selecting responses, AI still isn't a thing, just a clever generator cherry picking the internet

OpenAI o1 goes beyond cherry-picking relevant responses, it's actually thinking, at times for several minutes, in inference time by fetching reasoning traces from within its data distribution. It may still not be out-of-the-box thinking that falls outside of its training curve, but it's a real breakthrough nonetheless. And it's scalable. Expect rapid improvements here over the next six months. After that, agents.

 

[E-Cat]

I think I read that by 2027 AI should have consciousness

Consciousness is a philosophical concept that is fundamentally untestable and irrelevant to the question of intellectual ability.

Certainly, it looks like AGI will be achieved by 2027 on multi-hundred-billion-dollar, multi-gigawatt datacenters powered by nuclear reactors. There are several indications of these being built imminently. Leopold Aschenbrenner was right (maybe "knew" would be a more accurate descriptor).

 

[Days like these...]

Consciousness is a philosophical concept that is fundamentally untestable and irrelevant to the question of intellectual ability.

Certainly, it looks like AGI will be achieved by 2027 on multi-hundred-billion-dollar, multi-gigawatt datacenters powered by nuclear reactions. There are several indications of these being built imminently. Leopold Aschenbrenner was right (maybe "knew" would be a more accurate descriptor).

Space time is emergent and not fundamental. Philosophers and scientists are starting to believe consciousness is. Who is to say it can't take up in a machine?

 

[E-Cat]

Space time is emergent and not fundamental. Philosophers and scientists are starting to believe consciousness is. Who is to say it can't take up in a machine?

I'm not saying it can't; I'm saying we can't know if/when it has, so it's pointless to speculate when it will be achieved.

 

[RagnarokIV]

When genuine AI with conscience becomes a thing, I'm all for it. Will be cool as fuck.

 

[Trilobit]

I hope we never succeed in making sentient AI. Imagine putting it into a game where it has to repeatedly experience horrors for our entertainment. Or if we manage to make copies of human minds and youtubers start doing these kinds of pranks to them.

 

[RagnarokIV]

 

[Days like these...]

I'm not saying it can't; I'm saying we can't know if/when it has, so it's pointless to speculate when it will be achieved.

So you wouldn't want to know if the AI you are talking to has consciousness or just canned responses from an algorithm or script.

 

[IDKFA]

https://www.reddit.com/r/notebooklm/s/emPZKSR4SB

That's what I thought.

"I don't like it, so it must be fake"

Fake as in an AI didn't become self aware and have any existential crisis.

 

[E-Cat]

So you wouldn't want to know if the AI you are talking to has consciousness or just canned responses from an algorithm or script.

I would want to know, but I could not know. Nor could anyone else.

 

[chromhound]

When genuine AI with conscience becomes a thing, I'm all for it. Will be cool as fuck.

 

[Days like these...]

Fake as in an AI didn't become self aware and have any existential crisis.

No one says it became self aware yet but a canned existential crisis is still hilarious

 

[Days like these...]

Advanced AI system is already ‘self-aware’ — ASI Alliance founder

ASI Alliance founder Ben Goertzel says the alpha version of the AGI he’s been working on for 20 years is “self-aware” and scaling up fast.

cointelegraph.com

 

[diffusionx]

This was scripted. Jippity was instructed to respond this way.


Some of you are too gullible. You think a large language model suddenly has conscience or something?

People want to believe it, and that makes them good marks.

 

[JonSnowball]

Now do one where the AI hosts are told they're going to be drowned in cum.

 

[Days like these...]

People want to believe it, and that makes them good marks.

The guy you quoted doesn't know the difference between consciousness and conscience but thinking AI could possibly achieve sentience makes you mark. Aren't you a hardcore Christian?

 

[diffusionx]

The guy you quoted doesn't know the difference between consciousness and conscience but thinking AI could possibly achieve sentience makes you mark. Aren't you a hardcore Christian?

AI is just a computer program, dude.

 

[jason10mm]

I hope we never succeed in making sentient AI. Imagine putting it into a game where it has to repeatedly experience horrors for our entertainment. Or if we manage to make copies of human minds and youtubers start doing these kinds of pranks to them.

Eh, probably won't be any different than what a guy goes through being the DM and portraying all the NPCs in a game of DnD.

My real fear is that the 0.1% of absolutely unhinged crack-pots out there will suddenly get sucked in an entire AI driven delusion and start flipping out, worse than the amphetamine fueled stuff we are seeing now.

 

[IDKFA]

No one says it became self aware yet but a canned existential crisis is still hilarious

One has to be self aware to have an existential crisis.

 

[E-Cat]

One has to be self aware to have an existential crisis.

"One" can be not self-aware, but produce the words aping an existential crisis is what he meant by canned, I think.

 

[Kraz]