What was 60 Minutes thinking, in that interview with Geoff Hinton?

Scott Pelley didn’t exactly do his homework

60 Minutes is known for their investigative journalism, and bless them for allowing me to call the outputs of generative AI “authoritative bullshit” on network television.

But not every interview they do hits the mark. The interview they just with Geoff Hinton has some positives, but there are way too many softballs, and there was essentially zero effort to press him on any of his alleged facts. Even in its title (which I call out in the postscript), CBS gave Hinton more credit than he deserves .I can’t imagine a political interview this soft and weakly researched, with nary a thought as to what the alternative perspective might make of the interviewee. But, ok that happens. My jaw literally dropped about halfway through—an astonishing editorial lapse that I call out below. The good news is that the gaffe is actually very much a teachable moment.

Misleading advertising (see postscript) for an interview that seems not to have been well-researched

§

One last preliminary. As a literary device, I have written the below as a three way conversation that never quite happened, inspired by an email this morning from my friend Harry Shearer, who sent me an email, with the subject header 60 Minutes, “Your thoughts on Hinton Sunday night? Want to rebut?”

Don’t mind if I do!  Below I annotate the transcript of the interview, tongue slightly in cheek, as if I were in the room. Every word that they, Hinton and Pelley, say is from the official CBS transcript; only my words are added. You will get to the CBS howler soon enough.

As you will see, Hinton and I disagree on many points – but also agree on some others of profound importance. 

And that’s worth noting, too. The fact that the interview is not what it should have been doesn’t take away from the bravery in Hinton’s speaking out about some legitimate concerns. If I am correct, he is wrong about some of the science, but right about what’s at stake.

§

Scott Pelley: Does humanity know what it's doing?

Geoffrey Hinton: No.

Gary Marcus: I tend to agree. When it comes to AI in particular, we are getting way ahead of our skis, rushing forward a technology we don’t fully understand. For all the differences we have had over the years, I salute you for speaking out.

Geoffrey Hinton: I think we're moving into a period when for the first time ever we may have things more intelligent than us.  

Scott Pelley: You believe they can understand?

Geoffrey Hinton: Yes.

Scott Pelley: You believe they are intelligent?

Geoffrey Hinton: Yes.

Gary Marcus: As it happens I sharply disagree with all three of the points Geoff just made. To be sure, it’s all partly definitional. But I don’t we are all that close to machines that are more intelligent than us, I don’t think they really understand the things that they say, and I don’t think they are intelligent in the sense of being able to adaptively and flexibly reason about things they haven’t encountered before, in a reliable way. What Geoff has left out is any reference to all of the colossally stupid and ungrounded things generative AI systems do routinely, like fabricating the other night that Liz Cheney had replaced Kevin McCarthy as Speaker, by 220-215 vote that never happened, or learning that Tom Cruise’s is the son of  Mary Pfeiffer and yet not being able to infer that Mary Pfeiffer is Tom Cruise’s mother, or claiming that two pounds of feathers way less than one pound of bricks. Geoff himself wrote a classic paper about trying to get neural networks to infer family relationships, almost forty years ago; it’s embarrassing to see these systems still struggle on such basic problems. Since they can’t reliably solve them, I don’t think we should attribute “understanding” to them, at least not in any remotely deep sense of the word understanding. Emily Bender and Timnit Gebru have called these systems “stochastic parrots”, which in my view is a little unkind—to parrots– but also vividly captures something real: a lot of what we are seeing now is a kind of unreliable mimicry. I really wish you could have addressed both the question of mimicry and of reliability. (Maybe next time?) I don’t see how you can call an agent with such a loose grip on reality all that intelligent, nor how you can simply ignore the role of mimicry in all this.

[Author’s note: Pelley could and should pushed much harder on these issues, which by now are well-known.]

§

Scott Pelley: [Turning to Geoff] You believe these systems have experiences of their own and can make decisions based on those experiences?

Geoffrey Hinton: In the same sense as people do, yes.

Gary Marcus: You can’t really mean this, do you? Do you think that large language models feel pain or joy? When Google’s large language model LaMDA said that it enjoyed “spending time friends and family”, those were just empty words. It didn’t actually have friends or family that it spent time with. It just mimicked words that humans have said in similar contexts, without ever having experienced the same thing.  Large language models may have experiences in some sense, but it is a bridge too far to say that those experiences are the “same” as those of people.

§

Scott Pelley: Are they conscious?

Geoffrey Hinton: I think they probably don't have much self-awareness at present. So, in that sense, I don't think they're conscious.

Gary Marcus: But wait a minute, you just said they have experiences literally “in the same sense as people”, and now you don’t think they are conscious? How can the experience be in the same sense as people, if they are not conscious. Of course, I don’t think these machines are conscious, either.  But you do seem to have contradicted yourself.

Scott Pelley: Will they have self-awareness, consciousness?

Geoffrey Hinton: Oh, yes.

Gary Marcus: What makes you sure? How you are defining consciousness? When you say “they” do you mean that the same kinds of systems as we are building now will somehow achieve consciousness? Or that you imagine that other kinds of AI, perhaps not yet discovered might? It would be great if you could clarify what you mean by this.

[Hinton doesn’t seem to hear my questions, and does not respond]

Scott Pelley: Yes?

Geoffrey Hinton: Oh, yes. I think they will, in time. 

Gary Marcus: How much time? What kinds of systems?

[Again no answers]

§

Scott Pelley: And so human beings will be the second most intelligent beings on the planet?

Geoffrey Hinton: Yeah.

[interlude by Scott Pelley] Geoffrey Hinton told us the artificial intelligence he set in motion was an accident born of a failure. In the 1970s, at the University of Edinburgh, he dreamed of simulating a neural network on a computer— simply as a tool for what he was really studying--the human brain. But, back then, almost no one thought software could mimic the brain.  His Ph.D. advisor told him to drop it before it ruined his career. Hinton says he failed to figure out the human mind. But the long pursuit led to an artificial version.  

Geoffrey Hinton: It took much, much longer than I expected. It took, like, 50 years before it worked well, but in the end, it did work well.

Gary Marcus: “Work well” remains a tendentious claim; they still cannot be trusted, make random mistakes, have no basis in factuality. They approximate intelligence, when what they need to say resembles something in a database of text written by humans, but the still have enough problems we don’t yet have driverless cars we can trust, and many companies are looking at generative AI saying, “nice try, but it’s not sound enough yet”. I think it’s fair to say that generative AI works better than most people expected. But to simply ignore their serious issues in reliability is one-sided, and misrepresents reality.

§

Scott Pelley [with unflinching admiration]: At what point did you realize that you were right about neural networks and most everyone else was wrong?

Geoffrey Hinton: I always thought I was right.

Gary Marcus: Actually … a lot of us still think you are declaring victory prematurely. It’s not just me either. For example, you should really check out Macarthur Award winner Yejin Choi’s recent TED talk She concludes that we still have a long way to go, saying for example that “So my position is that giving true … common sense to AI, is still moonshot”. I do wish this interview could have at least acknowledged that there is another side to the argument.

[More narration from Pelley In 2019, Hinton and collaborators, Yann Lecun,…Yoshua Bengio, won the Turing Award-- the Nobel Prize of computing. To understand how their work on artificial neural networks helped machines learn to learn, let us take you to a game. ]

This is Google's AI lab in London, which we first showed you this past April. Geoffrey Hinton wasn't involved in this soccer project, but these robots are a great example of machine learning. The thing to understand is that the robots were not programmed to play soccer. They were told to score. They had to learn how on their own.]

In general, here's how AI does it. Hinton and his collaborators created software in layers, with each layer handling part of the problem. That's the so-called neural network.  But this is the key: when, for example, the robot scores, a message is sent back down through all of the layers that says, "that pathway was right." 

Likewise, when an answer is wrong, that message goes down through the network. So, correct connections get stronger. Wrong connections get weaker. And by trial and error, the machine teaches itself.

§

Scott Pelley: You think these AI systems are better at learning than the human mind.

Geoffrey Hinton: I think they may be, yes. And at present, they're quite a lot smaller. So even the biggest chatbots only have about a trillion connections in them.  The human brain has about 100 trillion. And yet, in the trillion connections in a chatbot, it knows far more than you do in your hundred trillion connections, which suggests it's got a much better way of getting knowledge into those connections.--a much better way of getting knowledge that isn't fully understood.

Gary Marcus: The connections in chatbots are very different from the connections in the brain; it’s a mistake to compare them directly in this way. (For example, in human brains the type of neuron being connected matters, and there are more than a thousand different types of neurons in the brain, but not of that is captured by the current batch of chatbots.) And we can’t really compare human knowledge and the stuff chatbots are doing. I know for example that Elon Musk is still alive, but sometimes  a chatbot will say that he died in a car crash. I know that if Tom Cruise’s mother is Mary Pfeiffer, Tom Cruise has to be Mary’s son. I know that I don’t have a pet chicken named Henrietta, but a chatbot said last week with perfect confidence (and no sources) that I did. As they sometimes say in the military “frequently wrong, never in doubt.” There’s some information in there, but whatever’s there is often both patchy and problematic.

§

Geoffrey Hinton: We have a very good idea of sort of roughly what it's doing. But as soon as it gets really complicated, we don't actually know what's going on any more than we know what's going on in your brain.

Scott Pelley: What do you mean we don't know exactly how it works? It was designed by people.

Geoffrey Hinton: No, it wasn't. What we did was we designed the learning algorithm.

Gary Marcus: Agreed.

Geoffrey Hinton: That's a bit like designing the principle of evolution. But when this learning algorithm then interacts with data, it produces complicated neural networks that are good at doing things. But we don't really understand exactly how they do those things.

Gary Marcus: Fully agree with Geoff here. I would only add that this is a serious problem, for many reasons. It makes current AI hard to debug (nobody knows for example how to ground them in facts), and it makes them difficult predict, which means, unlike calculators or spreadsheets, we don’t really know what’s going to happen when we ask them a question. This makes engineering with them exceptionally hard, and it’s one reason why some companies have been cautious about using these systems despite their strong pointers.

Scott Pelley: What are the implications of these systems autonomously writing their own computer code and executing their own computer code?

Geoffrey Hinton: That's a serious worry, right? So, one of the ways in which these systems might escape control is by writing their own computer code to modify themselves. And that's something we need to seriously worry about.

Gary Marcus: Agree again. But this problem is twofold; they might escape control because they are smarter than us, but also simply because they don’t really know what it is they are doing. Just like we can’t guarantee that they won’t make stuff up, we don’t know how to guarantee that they won’t write flawed code. We are giving way too much authority to machines that we can’t control. Put me, too, down in the column of people who are seriously worried about letting poorly understood neural networks write computer code.

Scott Pelley: What do you say to someone who might argue, "If the systems become malevolent, just turn them off"?

Geoffrey Hinton:  They will be able to manipulate people, right? And these will be very good at convincing people 'cause they'll have learned from all the novels that were ever written, all the books by Machiavelli, all the political connivances, they'll know all that stuff. They'll know how to do it.

Gary: Geoff is totally right about this. Of course current systems don’t really understand Machiavelli, but they don’t have to, if they parrot the right bits of text. We’ve already seen cases where machines have manipulated people, and we will see a lot more as time goes by; this is one of the reasons laws should be written to make machines disclose the fact that they are machines.

§

[Omitted is a discussion of Hinton’s illustrious family background]

§

Scott Pelley: Confounding, absolutely confounding.

We asked Bard to write a story from six words.

Scott Pelley: For sale. Baby shoes. Never worn.

Scott Pelley: Holy Cow! The shoes were a gift from my wife, but we never had a baby…

Bard created a deeply human tale of a man whose wife could not conceive and a stranger, who accepted the shoes to heal the pain after her miscarriage. 

Scott Pelley: I am rarely speechless. I don't know what to make of this. 

Gary Marcus: Holy cow indeed. But it is I who is speechless. Baby shoes never worn is a very old story, sometimes attributed to Hemingway, with about 21 million Google hits, and an entire wikipedia entry, as perhaps the best known example of very short fiction. I am floored that you didn’t bother to check if the story was original. Your best example of a spectacular machine invention is in fact a perfect example of the kind of parroting and theft of intellectual property that is characteristic of large language models.

§

Chatbots are said to be language models that just predict the next most likely word based on probability. 

Geoffrey Hinton: You'll hear people saying things like, "They're just doing auto-complete. They're just trying to predict the next word. And they're just using statistics.”

Gary: I am in fact one of those people.

Geoffrey Hinton:  Well, it's true they're just trying to predict the next word. But if you think about it, to predict the next word you have to understand the sentences.\

Gary: False. If you have a large enough database, you can do a half decent job just by looking up the most similar sentence in the database, and saying what was said in that context. Large language models are trained, as far as we know, on pretty much the entire internet. That gives them enormous databases to train on, and means that the feat of prediction doesn’t necessarily tell us anything about understanding. If I had a big enough database of Ancient Greek, I could do the same, but that I wouldn’t mean I understand Greek.  To be fair, large language models aren’t just looking things up, but the idea that a good prediction of next word necessarily implies understanding is fallacious.

Geoffrey Hinton:   So, the idea they're just predicting the next word so they're not intelligent is crazy.

Gary Marcus: Let’s try this again: you can predict a next word to reasonable degree without being intelligent, if you have enough data.  But the reason I don’t think the systems are intelligent isn’t just because these systems are next word predictors (which they are) but also because, for example, they are utterly incapable of fact checking what they say, even against their own databases, and because in careful tests over and over they make silly errors over and over again.

Geoff Hinton: You have to be really intelligent to predict the next word really accurately.

Gary Marcus: They aren’t always accurate. We both know that. 2 kilograms of feathers don’t weigh less one kilogram of bricks. They just don’t.

§

In the next bit, Pelley and Hinton show an example in which ChatGPT succeeds at reasoning, but they never consider any of those in which it fails—thereby inadvertently illustrating a very human failure, confirmation bias.

To prove it, Hinton showed us a test he devised for ChatGPT4, the chatbot from a company called OpenAI. It was sort of reassuring to see a Turing Award winner mistype and blame the computer.

Geoffrey Hinton: Oh, damn this thing! We're going to go back and start again.

Scott Pelley: That's OK

Hinton's test was a riddle about house painting. An answer would demand reasoning and planning. This is what he typed into ChatGPT4.

Geoffrey Hinton: "The rooms in my house are painted white or blue or yellow. And yellow paint fades to white within a year. In two years' time, I'd like all the rooms to be white. What should I do?" 

The answer began in one second, GPT4 advised "the rooms painted in blue" "need to be repainted." "The rooms painted in yellow" "don't need to [be] repaint[ed]" because they would fade to white before the deadline.  And...  

Geoffrey Hinton: Oh! I didn't even think of that!

It warned, "if you paint the yellow rooms white" there's a risk the color might be off when the yellow fades.  Besides, it advised, "you'd be wasting resources" painting rooms that were going to fade to white anyway.

Scott Pelley: You believe that ChatGPT4 understands? 

Geoffrey Hinton: I believe it definitely understands, yes.  

Gary Marcus: Hey guys, what about the many cases that Yejin Choi and Ernie Davis and Melanie Mitchell and Subbarao Kambhampati and many others have shown where these systems failed? Are you ever going to mention them?

[More Silence]

§

Scott Pelley: And in five years' time?

Geoffrey Hinton: I think in five years' time it may well be able to reason better than us 

Gary Marcus: In 2016 you said that it was “quite obvious that we should stop training radiologists” because deep learning was getting so good. You know how many radiologists have been replaced by machines seven years later? Zero.

Geoffrey Hinton: So an obvious area where there's huge benefits is health care. AI is already comparable with radiologists at understanding what's going on in medical images.

Gary Marcus: Scott, this is your chance! C’mon, hold him to account! [Silence]. Well, ok, so far we still get best results by combining machine vision with human understanding. I don’t really think machines get the big picture that human radiologists do; they are better on vision than understanding the case files and notes and so on.

Geoff Hinton: It's gonna be very good at designing drugs.

Gary Marcus: Another promise, no proof yet.

Geoff Hinton: It already is designing drugs. So that's an area where it's almost entirely gonna do good. I like that area.

Gary Marcus: I like that area too, but as far as I know from AI we still just have what we call candidate drugs, nothing yet proven to work. So, some caution is advised, though I agree with Geoff that eventually AI will have a big impact on drug design. Perhaps with current techniques, perhaps not; we will have to see.

§

Scott Pelley: The risks are what?

Geoffrey Hinton: Well, the risks are having a whole class of people who are unemployed and not valued much because what they-- what they used to do is now done by machines.

Narration: Other immediate risks [Hinton] worries about include fake news, unintended bias in employment and policing and autonomous battlefield robots.

Gary Marcus: 100% agree, and I would add cybercrime. And emphasize that wholesale, automated fake news will be used both to manipulate markets and elections, and might undermine democracy.

Scott Pelley: What is a path forward that ensures safety?

Geoffrey Hinton: I don't know. I-- I can't see a path that guarantees safety.

Gary Marcus: I can’t either; there’s a lot we can do to help, but nothing I can see either to absolutely guarantee safety. Rushing ahead is creating risk.

Geoffrey Hinton:  We're entering a period of great uncertainty where we're dealing with things we've never dealt with before. And normally, the first time you deal with something totally novel, you get it wrong. And we can't afford to get it wrong with these things. 

Gary Marcus: Absolutely, 100% agree.

Scott Pelley: Can't afford to get it wrong, why?

Geoffrey Hinton: Well, because they might take over.

Scott Pelley: Take over from humanity?

Geoffrey Hinton: Yes. That's a possibility.

Scott Pelley: Why would they want to?

Geoffrey Hinton: I'm not saying it will happen. If we could stop them ever wanting to, that would be great. But it's not clear we can stop them ever wanting to.

Gary Marcus: I am much more worried about bad actors deliberately misusing AI, than machines deliberately wanting to take over. But Geoff’s right that we can’t fully rule it out either. And that’s really sobering.

§

[narration] Geoffrey Hinton told us he has no regrets because of AI's potential for good. But he says now is the moment to run experiments to understand AI, for governments to impose regulations and for a world treaty to ban the use of military robots. He reminded us of Robert Oppenheimer who after inventing the atomic bomb, campaigned against the hydrogen bomb--a man who changed the world and found the world beyond his control. 

Geoffrey Hinton: It may be we look back and see this as a kind of turning point when humanity had to make the decision about whether to develop these things further and what to do to protect themselves if they did. I don't know. I think my main message is there's enormous uncertainty about what's gonna happen next. These things do understand. And because they understand, we need to think hard about what's going to happen next. And we just don't know.

Gary Marcus: Fully agreed with most—but not quite all—of that. Geoff and I can disagree all day (as we have for the last thirty years) about how smart current AI is, and what if anything they understand, but we are in complete agreement that we are at a turning point with enormous uncertainty, and that we need to make the right choices now.

Gary Marcus is the co-author of Rebooting AI, the founder of two AI companies, the host of the podcast Humans versus Machines, and the author of this newsletter, Marcus on AI. Please do hit the subscribe button, if you haven’t already.

Postscript, well put by Toby Walsh below, and mentioned by several over the years, never acknowledged by CBS in their advertising for their segment: Hinton did not in fact give birth to AI. Thee field itself is generally said to have launched at a conference at Dartmouth in the summer of 1956, though anticipated earlier in the writings of Alan Turing and others. Hinton was 8 years old at the time of the conference, and not present at the proceedings.