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

LaMDA is not sentient

garymarcus.substack.com

Post reply on HN
Print view View original

91–100 of 694 posts

Re: LaMDA is not sentient

#91

Post by someweirdperson »

renewiltord wrote: post #44
jck wrote: post #20

Earlier quoted context omitted.

Pharmacists are bound by clear rules and regulations which dictate what drugs they can dispense and when. It has little to do with your "prompt engineering" and more to do with whether your doctor actually filed the prescription correctly.

Haha, you'd think but they get it wrong often and they'll bend things for you. It's why I can pick up other people's prescriptions for controlled substances with nothing but my ID at pharmacies I've never been to before, why I've transformed a no-refill into a refill (they just read it wrong) while another person had to call the doctor for the same outcome, etc. People everywhere are not really bound by anything. The…

Are you Doctor WHO?

Re: LaMDA is not sentient

#92

Post by enw »

jchw wrote: post #16

Quoted post unavailable.

Do you believe the world is deterministic?

That if you feed all the variables of the current state of the universe to an algorithm that knows all the laws of the universe, we can replicate life 1:1 to the last atom, from the beginning of the universe to any arbitrary point in time, the same way, every single time?

Re: LaMDA is not sentient

#93

Post by wruza »

tsimionescu wrote: post #60
ehsankia wrote: post #13

I also think the claims are overblown, but I also don't understand all the people trying to undermine lambda by saying it just completes word sequences. Text is the input we have chosen since it's easy to work with. By that logic, we can argue that humans are soundwave and lightwave sequence completing machines. Yes our inputs are different since we have more complex senses, and language is one of the main outputs we…

> I also don't understand all the people trying to undermine lambda by saying it just completes word sequences. Text is the input we have chosen since it's easy to work with. By that logic, we can argue that humans are soundwave and lightwave sequence completing machines. You're missing the point. Humans use their language to think or communicate about a problem that they want to solve. If LaMDA produces the sentence…

Thanks for validating my idea of how this works. Btw, is google (or any AI group) able to create a set of connected neural networks, different in function, that could operate continuously without a prompt? Would that be closer to what we’re discussing?

E.g. something like a short/long-term memory blocks, visual and text processing, emotional block (why not), etc.

Re: LaMDA is not sentient

#94

Post by radu_floricica »

Look for second order effects here: https://twitter.com/ESYudkowsky/status/1536008920375754752

First would be that this (completely bonkers) claim will also paint any subsequent claims as also bonkers. Which is Bad, because with 100% probability, at some point in the near or medium future we'll hit on the real deal.

Second is that the first reflex of a human being who thought he saw a real AI was to give it rights and autonomy. That's something that actually never clicked for me until now. I always thought the danger is simply in not recognizing the first dangerous AIs, not that everybody will agree they're (super)intelligent, but we'll be hamstrung to defend against it by the public opinion. Our chances of surviving this just got a lot lower.

And in case anybody will protest that AIs will not be an enemy, I'll paraphrase EY: they don't want to kill you; it's just that you're made of atoms they may need for other things.

Re: LaMDA is not sentient

#95

Post by jl6 »

I would find LaMDA more convincing if it wasn’t constrained to a request-response model, but instead could choose to say something whenever it felt like.

Re: LaMDA is not sentient

#96

Post by pyinstallwoes »

A Sanskrit grammarian noted that all things have meaning precisely by relative associations.

Sounds sounds. Letter to other letter. Word to other word. Paragraph to other paragraph.

Quite profound in contrast with the Chinese room thought experiment.

Re: LaMDA is not sentient

#97

Post by deaddodo »

pyuser583 wrote: post #85
probably_wrong wrote: post #39

This LaMDA story reminds me of another article I found last year about a man who fed his deceased girlfriend's chats to GPT3 [1] and believed the system was giving the answers she would have. Both articles rely on the operators taking the smallest amount of evidence and ignoring everything else, which I guess is what this article calls "the gullibility gap" (awful name). Getting your audience to do the heavy lifting…

> fed his deceased girlfriend's chats to GPT3 [1] and believed the system was giving the answers she would have Isn’t that what AI does? Repeats back what it’s “heard?”

No, and that’s precisely the point. A “true AI” would have an independent intelligence and the ability to form self-referential thoughts.

Regurgitating statistically sound data in an intelligent-seeming manner is not what these researchers are referring to when they refer to “AI”.

Re: LaMDA is not sentient

#98

Post by davedx »

s1mon wrote: post #11

It sounds like despite all the stories about how hard it is to get through many rounds of difficult interviews at Google, they managed to hire someone who believed LaMDA is a 7 or 8 year old child. "Mr. Lemoine, a military veteran who has described himself as a priest, an ex-convict and an A.I. researcher, told Google executives as senior as Kent Walker, the president of global affairs, that he believed LaMDA was a c…

Attack the argument, not the person.

Re: LaMDA is not sentient

#99

Post by upwardbound »

cwkoss wrote: post #72

What makes everyone in this thread so confident that it isn't sentient?

At least personally, my subjective experience as a sentient being includes having a 3-dimensional "virtual reality" view of the world around me, which is mostly informed by sight but also by touch and sound. I think this is a key element of human & animal sentience and I tend to think that all animals that have this "3D representation of the world" in their mind are sentient. E.g. fish can solve mazes, in a way illustrative of having a 3D view of the world within their minds like we do, so I think they are sentient. https://www.researchgate.net/publication/251359525_Performan...

If you search Google for the phrase , you will find lots of blog posts and the like, explaining with examples which I mean by the 3D VR-like world in your head which I am describing.

Another commenter in this thread raised the question of whether robots that navigate the world are sentient. I think the answer is possibly yes but I'm going to give a huge disclaimer which is I don't think that sentience is morally relevant, I think what is morally relevant is whether an entity feels qualia or not. https://en.wikipedia.org/wiki/Qualia In my personal view (just hunches really), I think that whether qualia are felt depends on the implementation. My hunch is that an implementation of a neural network in hardware, i.e. an actual network of transistors, might feel qualia, but that a software implementation of the same neural network would not. But others have different hunches and this is a question as yet unsolved by today's science.

Re: LaMDA is not sentient

#100

Post by bryanrasmussen »

mpoteat wrote: post #77
Imnimo wrote: post #67

I think the real ethical concern here should be that people can be idly fooled into thinking good chat bots are sentient. Like this guy isn't the only person in the world who would fall for LaMDA's parlor tricks. He's just the one who got to spend hours talking to it every day. If you set up every American with a copy, I bet a non-trivial portion of society would come to the same conclusion. People are naturally pred…

Suppose a child was raised by a more advanced version of large scale language model, with the ability to detect and act on its environment, as well as form memories. Further suppose that this child develops a deep emotional connection with the AI. Do you think that child would find your argument convincing that the AI is "not really a person"?

suppose a child were raised by a human parent and at some point becomes a teenager and thinks, this person isn't worth listening to at all?

If that's possible then I guess it's possible to tell the AI raised teen the AI isn't really a person, you just need to find the right moment to give them that thought.

Basically you need another AI monitoring development to pick the optimal time to inform the child that the first AI is just a computer program.

OR you could do it like the movie I Am Mother https://www.rottentomatoes.com/m/i_am_mother

Post reply on HN