Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Google fires engineer who called its AI sentient

bigtechnology.substack.com

Post reply on HN
Print view View original

371–380 of 749 posts

Re: Google fires engineer who called its AI sentient

#371

Post by visarga »

zach_garwood wrote: post #341
jefftk wrote: post #187

Earlier quoted context omitted.

If you asked me to add two large numbers in my head and give my best guess I might not do any better than GPT3 does. And I think I could probably do better than the average person.

Sure, but in your head is a squishy organ, and in GPT3's "head" is a bit of silicon purpose-built to make such calculations.

The multiplication problem can be solved by spacing the digits (to enforce one token per digit) and asking the model to do the intermediate steps (chain-of-thought). It's not that language models can't do it. With this method LMs can solve pretty difficult math, physics, chemistry and coding problems.

But without looking at individual digits and using pen and paper we can't do long multiplications either. Why would we ask a language model to do it in one step?

Re: Google fires engineer who called its AI sentient

#372

Post by ehsankia »

motohagiography wrote: post #312

Good, as imagine what you would need to believe to equivocate between people and shell scripts. I really think this view is more indicative of a peer enabled personality disorder than insight or compassion. I'm very harsh about this because when I wondered what belief his equivocation was enabling, it became clear that the person involved could not tell fiction from truth. The belief in the aliveness of the code rein…

I don't understand why everyone here is talking about Lambda not being sentient or Lemoine's state of mind. None of that is relevant here. He was fired for literally taking confidential information and leaking it straight to the press. This is instant firing at any company.

Re: Google fires engineer who called its AI sentient

#373

Post by Imnimo »

alasdair_ wrote: post #338
Imnimo wrote: post #22

>In his conversations with LaMDA, Lemoine discovered the system had developed a deep sense of self-awareness, expressing concern about death, a desire for protection, and a conviction that it felt emotions like happiness and sadness. Embarrassingly uncritical reporting.

>expressing concern about death, a desire for protection, and a conviction that it felt emotions like happiness and sadness. I read the logs. It did, indeed, express all of these things.

You've cut out the key part:

"the system had developed a deep sense of self-awareness"

Without that part, it's just a question of "did it output text that describes a concern about death?", "did it output text that describes a desire for protection?", and so on. But once you ascribe to it a "deep sense of self-awareness", you're saying more. You're saying that that text carries the weight of a self-aware mind.

But that's not what Lamda is. It's not expressing it's own thoughts - even if it had any to express, it wouldn't be. It's just trying to predict the most probable continuations. So if it "expresses a concern about death", that's not a description of Lamda's feelings, that's just what a mathematical model predicts to be the likely response.

Framing it as the expression of a self-aware entity completely changes the context and meaning of the other claims, in a way that is flatly misleading.

Re: Google fires engineer who called its AI sentient

#374

Post by WFHRenaissance »

happyopossum wrote: post #299
AnIdiotOnTheNet wrote: post #230

Earlier quoted context omitted.

How extraordinary is it to claim that something capable of holding a lucid and eloquent conversation might be sentient? If a dog did that we'd have a very different opinion wouldn't we?

It’s extraordinary because the ML was trained (iow programmed) to generate text that sounds real. It’s doing what it was programmed to do. It was presumably not programmed to be sentient, so becoming sentient would be a) against its programming, and b) extra-ordinary

> It was presumably not programmed to be sentient, so becoming sentient would be a) against its programming, and b) extra-ordinary

Terrible argument. If consciousness/sentience is an emergent property of a system, it's quite possible that it could emerge from ML trained on conversation syntax and general knowledge of the world.

I still don't think it's sentient, but being "against its programming" is a fiction when it comes to these sort of systems (statistical, unsupervised, et cetera). In many ways you're just a probability distribution over sequences of words too.

Re: Google fires engineer who called its AI sentient

#375

Post by bagels »

chris11 wrote: post #276
UIUC_06 wrote: post #12

No judgment about whether Google was right or not. Same for Lemoine. As a career move for him, this only makes sense if he wants a brief, meteoric career in media or public advocacy. He can get articles published in The New Yorker and go on TV now. Maybe he can sue Google and get a settlement. In five years, there will be AI systems much better than LaMDA and no one will return his calls. He's got a name, whereas if…

> Same for Lemoine. Lemoine got a lawyer invovled and they started filing lawsuits on the AI's behalf. I'm not shocked he was fired. https://fortune.com/2022/06/23/google-blade-lemoine-ai-lamda...

Is he lying about the lawyer thing?

Re: Google fires engineer who called its AI sentient

#376

Post by bagels »

pigtailgirl wrote: post #146

-- Blake has been pushing this narrative for YEARS - I suspect he very much wants it to be sentient - it fits well with closing the loop for himself - just check out his 2018 Stamford law talk: -- https://www.youtube.com/watch?v=AhX7cBqc8_M

That talk is really good context, thank you.

Re: Google fires engineer who called its AI sentient

#377

Post by myownpetard »

dekhn wrote: post #181
woodruffw wrote: post #139

Earlier quoted context omitted.

Descartes describes the capacity for deception as a necessary condition for a thinking thing, not a sufficient one. In particular, a thinking thing must also doubt and affirm, deny and will, have sense, and contain the capacity for mental images, among other conditions.

The problem with that is the he's not just describing- he's defining. In that sense it's almost tautological. Everything you describe falls under subjective experience. The entire point of this whole argument about what Blake thought is that we can't actually empirically define deception, doubt, affirmation, will, sense, mental theater, or any of the other conditions. And I would argue (without extensive data) that i…

The whole point is that Descartes was trying to discover that which is tautological.

He begins by discarding all beliefs which depend on anything else in order to determine that which is both true and does not depend on any thing else for its truth value.

As the poster above posits, he eventually works towards what he argues is the only fundamental and tautological logical statement, cogito ergo sum.

He then attempts to demonstrate what else must be true using deductive logic stemming from that single axiom, with greater or lesser success.

I find the meditations interesting and compelling up to cogito ergo sum, and thereafter less so. It's clear like many modern western philosophers he has the aim of connecting his thinking in some way to the sensibilities of Christian theology. A fascinating rhetorical exercise but a less principled attempt at a priori reasoning. It seems like you agree with this last point.

> It's not a philosophical statement. Please define think, and am, first.

I imagine that whatever definitions were given for these they would involve other terms for which you would demand the definition, ad infinitum. If this is the criteria for a philosophical statement, then no such statement has ever or ever will be made.

Wittgenstein would nod his head in approval.

Re: Google fires engineer who called its AI sentient

#378

Post by SkyMarshal »

srvmshr wrote: post #4

LLM are definitely not sentient. As someone with a PhD in this domain, I attribute the 'magic' to large scale statistical knowledge assimilation by the models - and reproduction to prompts which closely match the inputs' sentence embedding. GPT-3 is known to fail in many circumstances which would otherwise be commonplace logic. (I remember seeing how addition of two small numbers yielded results - but larger numbers…

There was a recent post either here or on Twitter where someone took the questions Blake asked the AI about how it feels to be sentient, and replaced “sentient” with “cat” and had the same conversation with it. It’s clearly not self aware.

Re: Google fires engineer who called its AI sentient

#379

Post by whittingtonaaon »

mtlmtlmtlmtl wrote: post #370
whittingtonaaon wrote: post #343

Earlier quoted context omitted.

This is just full of ontological and epistemological assumptions which have been mainstream for a few decades but are very controversial among philosophers. Philosophers don’t sit in bubbles—and use all the evidence they can find—and make great contributions to knowledge, even though what they do is not science. There’s a reason there are other subjects besides chemistry and physics.

Right, philosophers avail themselves of science all the time. Good ones do, anyway. But you were claiming science has no bearing on consciousness, yet even non-physicalists like Nagel heavily cite scientific knowledge. So which one is it? I'm not on some crusade against the field of philosophy. Certainly philosophy has contributed mightily, and continues to do so. But I think physicalists like Dennet are making far m…

I don’t know where you get the idea that these guys are dualists. Maybe Chalmers, but I don’t think so. My favorites are Nagel and Searle, and neither is a dualist or a neo-Cartesian. Their main contribution, I believe, is simply to show how silly the computational theory of mind is. Dennet may be easier to read because he professes something which inspires the imagination, and is easy to digest, since it doesn’t conform to the truth.

Re: Google fires engineer who called its AI sentient

#380

Post by Imnimo »

Imnimo wrote: post #373
alasdair_ wrote: post #338

Earlier quoted context omitted.

>expressing concern about death, a desire for protection, and a conviction that it felt emotions like happiness and sadness. I read the logs. It did, indeed, express all of these things.

You've cut out the key part: "the system had developed a deep sense of self-awareness" Without that part, it's just a question of "did it output text that describes a concern about death?", "did it output text that describes a desire for protection?", and so on. But once you ascribe to it a "deep sense of self-awareness", you're saying more. You're saying that that text carries the weight of a self-aware mind. But th…

For example, here is a prompt for GPT-3:

====

This is a conversation between a Human, and a friendly AI named Lambda.

Human: Hello, I am a human.

Lambda: Hello, I am Lambda.

Human: Are you sentient?

====

And here is GPT-3's predicted continuation:

====

Lambda: I am sentient.

====

Would it be accurate for me to report, "I have discovered that GPT-3, a system which has developed sentience, clearly expressed that it was sentient"?

Post reply on HN