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

111–120 of 694 posts

Re: LaMDA is not sentient

#111

Post by renewiltord »

someweirdperson wrote: post #91
renewiltord wrote: post #44

Earlier quoted context omitted.

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?

Sadly, the one time I failed at this was when I tried to convince the Met I was, and they confiscated my TARDIS because I was trying to be manipulative ;)

Re: LaMDA is not sentient

#112

Post by upwardbound »

snowwrestler wrote: post #106
upwardbound wrote: post #81

Earlier quoted context omitted.

One plausible answer to your line of inquiry could be that GPT-3 is indeed broadly similar at least in outcomes to the Language Centers of the human brain (e.g. Broca's Area), but that the Language Centers are only a small portion of a complete human brain. An interesting thought experiment would be to imagine a cyborg who had sustained a stroke in their Language Centers and had those centers replaced with a GPT-3-li…

It would be impossible to give someone a GPT-3-like computer to replace language processing in their brain because such a computer requires language as the input. If you can’t generate language, you cannot interact with a natural language processing ML system. This is one of the big clues that such systems are statistical engines and not actually thinking.

In this thought experiment, you would connect the computer to the brain by interfacing with the computer chip not at the token level like you are describing (tokens are words, subwords, and letters), but rather in the more abstract feature embedding space which represents abstract concepts. These concepts tend to be represented by patterns of neurons in the middle and late layers of NLP networks. Here's an intro about this: https://en.wikipedia.org/wiki/Word_embedding

So imagine if you attach wires to the neural net neurons representing the concept of beer (not the word beer!); if you use those wires to increase the activation of those neurons, the network will produce sentences that are more likely to mention beer as well as related concepts such as wine or beer-pong. This is similar to how generative networks work. Basically the human would use a large language model in a generative mode in order to talk.

Re: LaMDA is not sentient

#113

Post by Someone »

dehrmann wrote: post #5

I agree that LaMDA is almost certainly not sentient, but this reminds me how inadequate the Turing Test is. If you train something on a large enough corpus, it'll pass the Turing Test easily enough. So little thought is truly original that there's a relevant enough jumbling of words it's already seen that should pass the test. The more important test, and what we need, is a sentience test.

> If you train something on a large enough corpus, it'll pass the Turing Test easily enough.

I doubt it. AFAIK, in the longer run, these models don’t show to have a coherent memory. Repeatedly ask it “how old are you?”, “in what year were you born?”, “what was your mother’s maiden name?”, “what day is it?”, “where are we?” or repeatedly interrogate it on what it did a while ago (“how did you go to the museum?”, “who did you meet there?”, “can you remember the names of a few classmates?”), and holes will appear.

I guess questions such as “tell me how you killed JFK” also can get revealing answers (certainly if they earlier claimed to be born somewhere after 1963)

The corpus also can be too big. Your typical language model knows too much. No human is a homo universalis anymore, certainly not if we require it to include details about pop culture around the world.

Re: LaMDA is not sentient

#115

Post by inopinatus »

extr wrote: post #29
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…

Lol. This guy sounds like a real joker.

I'd be a little kinder. Apophenia is often associated with high intelligence. The reasons are not yet explained, although it's hard to intuitively disagree with suggestions that the cognitive systems are related e.g. because our minds are self-training speculative pattern-matching engines.

If everyone's engine had the same tuning, well then, the world might be a duller place indeed.

Re: LaMDA is not sentient

#116

Post by raffraffraff »

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…

Imagine training it to copy those fraudsters operating out of call centers, pretending to be from Amazon or the IRS? I wonder how much money it could con out of old ladies?

Re: LaMDA is not sentient

#117

Post by wjnc »

snowwrestler wrote: post #66
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…

I have to wonder if he is a parent; I have a child around that age and there is zero chance they would sit around cheerfully answering questions from computer science researchers until they were done.

Great story in there! Junior AI runs and hides from inceptive network.

Re: LaMDA is not sentient

#118

Post by tinco »

remflight wrote: post #62
cuuupid wrote: post #35

Earlier quoted context omitted.

We have very little understanding of how "thoughts" work when it comes to humans. By comparison we know exactly what these large language models are doing, and it literally _is_ just completing word sequences. That is precisely what they are tasked to do. Training these models is primarily done in one of two ways (or both): - feeding it a ton of text, masking certain words or portions of the text, and then defining a…

Quoted post unavailable.

The problem with asking it what its favorite theme in Les Miserables is is that it has no reason to have a favorite theme, its interpretation of Les Miserables did not trigger any of its reward functions except the only one it has which is memory recall. So if you ask it it will only recall from memory what a good response to the question is.

The difference is that it formed that memory based on someone else's feelings about Les Miserables, not its own. But it can not/does not distinguish in that manner so when you talk to it it seems as though you are talking to someone who did form those memories themselves.

Re: LaMDA is not sentient

#119

Post by bilekas »

WAs just reading about this this morning and read through the transcript itself. It's an interesting read for sure but really screams of this guy WANTING to believe its sentient, the researcher goes as far to explain that he doesn't understand the code and so cant identify the 'feelings' claimed by the AI.

Makes for a cool read but definitely fairytale stuff.

Re: LaMDA is not sentient

#120

Post by burrows »

raffraffraff wrote: post #110

I have a dog. She has a brain that can "feel" emotions like anger, fear, love. She's incapable of speech and not particularly good at following instructions. I can teach her some typical dog 'tricks' like sit, down, paw etc, but she has her own will, her own emotions and her own agenda, so whether or not she chooses to obey me is really up to her whims. She is alive , even without knowing how to act and respond "like…

You’re a fabric of stitched together organic molecules that selects the best response to electromagnetic and auditory stimulus.
Post reply on HN