Tech Won't Save Us - ChatGPT Is Not Intelligent w/ Emily M. Bender

Episode Date: April 13, 2023

Paris Marx is joined by Emily M. Bender to discuss what it means to say that ChatGPT is a “stochastic parrot,” why Elon Musk is calling to pause AI development, and how the tech industry uses language to trick us into buying its narratives about technology. Emily M. Bender is a professor in the Department of Linguistics at the University of Washington and the Faculty Director of the Computational Linguistics Master’s Program. She’s also the director of the Computational Linguistics Laboratory. Follow Emily on Twitter at @emilymbender or on Mastodon at @[e-mail]. Tech Won’t Save Us offers a critical perspective on tech, its worldview, and wider society with the goal of inspiring people to demand better tech and a better world. Follow the podcast (@techwontsaveus) and host Paris Marx (@parismarx) on Twitter, and support the show on Patreon. The podcast is produced by Eric Wickham and part of the Harbinger Media Network.  Also mentioned in this episode:Emily was one of the co-authors on the “On the Dangers of Stochastic Parrots” paper and co-wrote the “Octopus Paper” with Alexander Koller. She was also recently profiled in New York Magazine and has written about why policymakers shouldn’t fall for the AI hype.The Future of Life Institute put out the “Pause Giant AI Experiments” letter and the authors of the “Stochastic Parrots” paper responded through DAIR Institute.Zachary Loeb has written about Joseph Weizenbaum and the ELIZA chatbot.Leslie Kay Jones has researched how Black women use and experience social media.As generative AI is rolled out, many tech companies are firing their AI ethics teams.Emily points to Algorithmic Justice League and AI Incident Database.Deborah Raji wrote about data and systemic racism for MIT Tech Review.Books mentioned: Weapons of Math Destruction by Cathy O'Neil, Algorithms of Oppression by Safiya Noble, The Age of Surveillance Capitalism by Shoshana Zuboff, Race After Technology by Ruha Benjamin, Ghost Work by Mary L Gray & Siddharth Suri, Artificial Unintelligence by Meredith Broussard, Design Justice by Sasha Costanza-Chock, Data Conscience: Algorithmic S1ege on our Hum4n1ty by Brandeis Marshall.Support the show

This is a tool. It's technology. Is your hammer loyal, right? Is your car loyal? Like that doesn't mean anything. But it's part of this narrative of AIs are separate autonomous thinking agents that are maybe now in their infancy. And so we have to nurture them and raise them right.

And then they become things where we can displace accountability to those things instead of keeping the accountability where it belongs. Hello and welcome to Tech Won't Save Us. I'm your host, Paris Marks. And before we get into this week's episode, just a reminder that this month, April of 2023, is the third birthday of Tech Won't Save Us. It's hard to believe it's been that long. And to celebrate, we're asking listeners like you, listeners who enjoy the show, who learn from these critical conversations that we have every week, to consider becoming a supporter over on patreon.com so that we can keep doing this work,

so that I can keep doing the research that's necessary to have these in-depth conversations with the experts who come on the show week in and week out, so that Eric can keep putting it together and making it sound as good as it does, and so that Bridget can keep making the transcripts of these episodes. That all depends on support from listeners like you. And that's why this month we've set a goal that if we get 200 new supporters over on patreon.com this month, or if people upgrade their pledges to help us get to that number, then I'll start working on a special series on Elon Musk that digs into his history, you know, how he built up this myth of himself as this kind of genius founder that was delivering us the future, and how that has all kind of imploded over the

past little while. I think it's something that would be really interesting to explore, especially in this moment. And I think it's something that listeners like you would really enjoy listening to. So if you do want to support the show and help us get to that goal, go on over to patreon.com slash tech won't save us, become a monthly supporter, and I'll even send you some stickers in the mail if you do that. So with that said, this week's guest is Emily M. Bender. You might be familiar with that name because she's not only a very prominent critic of AI and large language models, but we also mentioned her on the previous episodes with Timnit Gebru and with Dan McQuillan. Emily is a professor in the Department of Linguistics at the University of Washington and a faculty director of the

Computational Linguistics Master's Program. She's also a director at the Department of Linguistics at the University of Washington and a faculty director of the Computational Linguistics Master's Program. She's also a director at the Computational Linguistics Laboratory. Now, I was really excited to chat with Emily in this episode because if you follow her on Twitter, you'll see that she's very frequently calling out everything from media reporting on generative AI that she feels doesn't kind of live up to the bar that she holds, you know, that doesn't present a critical enough perspective on these technologies and kind of repeats the boosterist framings. But also, you know, the comments of people like Elon Musk or Sam Altman or these other people who are really involved in the industry and who want to put out particular framings of these issues that benefit themselves. And so in this episode, we talk about how she got

into computational linguistics because that shapes we talk about how she got into computational linguistics because that shapes a lot of how she understands these large language models and what they can actually do. And in particular, the concept of the stochastic parrot that has been really important in shaping a lot of the criticism of these technologies and of these tools over the past couple of years. We get into a lot of the hype around these technologies and of these tools over the past couple of years. We get into a lot of the hype around these technologies, and in particular, the letter that was released a couple of weeks ago from people like Elon Musk calling for a pause on AI development and how we should understand that and whose motives are actually being served by the way that that letter was framed.

And we also end by talking about the language that we use around technology more generally. Since she's a linguist, I wanted to ask her about some of these terms that we use that come out of the PR teams of these major companies and how it shapes the way that we understand these technologies and kind of, you know, takes our guard down because it makes it seem like these technologies are something that they're not. So I was really excited to have this conversation with Emily. I think that you're really going to enjoy it. And I think that it adds another important perspective to the growing kind of series that we have now

of conversations on generative AI, chat GPT, large language models, these other things, as kind of the hype really has exploded over the past number of months. And certainly I'll be doing more episodes on those things in the future. So stay tuned for those. So with that said, just a final reminder

that if you do enjoy this conversation and if you enjoy listening to the podcast, consider joining supporters like Josh Lepowski in St. John's, Uzay in Berlin, Joe from Brooklyn, Rasheen in Kingston, Joep from the Netherlands, Thomas in Brooklyn, and Diane in Washington, DC by going to patreon.com slash tech won't save us and becoming a supporter so that you can help support the work that goes

into making the show every week and to help us hit our goal for the podcast third birthday. Thanks so much and enjoy this week's conversation. Emily, welcome to tech won't save us. Paris, thank you so much for having me on. I'm excited to be a part of this. I'm very excited to chat with you since I was talking to Timnit Gebru back in January, and I know that she has worked with you before, you know, I was like, okay, I need to get Emily on the show. And then I was talking to Dan McQuillan, I believe it was last month, maybe it was the month before now, time is a bit of a mess over here. He was mentioning your work as well. And I was like, right, I really have to reach out to Emily and get her on the podcast so

we can talk about AI and all of this. So I'm very excited, as you can tell, to have you on the show to really dig into, you know, all that we're hearing now around AI and large language models and how your work can help us to understand this a little bit more and to get through kind of the hype and all of this that these companies want us to be obsessed with so that we're not paying attention to kind of the fundamentals and what we should be understanding. And so to get us started, before we get into all of those bigger questions, I want to ask a little bit about you. So can you explain to us what it means to be a computational linguist and how you got into doing this work that brings together language and computers?

Yeah, absolutely. So computational linguistics, put simply, is getting computers to deal with human languages. And there's a couple different reasons you might do that. You might be interested in doing linguistic research and using computers as tools to help you with that. Or you might be interested in building what we call human language technology. This used to be obscure, but now you can't go through a day without interacting with language technology if you are living in the sort of situation where tech is around you. There's plenty of people on the planet who don't use it, but for many of us. So we're talking search engines, we're talking automatic transcription, we're talking machine translation, but also things that are more behind

the scenes, like automatic processing of electronic health records, for example, to flag patients who might need a certain test or to match patients to clinical trials. There's applications in the legal domain, in the process of discovery in a lawsuit. The list goes on and on. Basically, any domain of endeavor where we use language to do some work, there's scope for doing language processing to help the people do that work. So it's a big area. It's not just chatbots. It's also sometimes referred to as natural language processing. Typically, if you're coming at it from a computer science point of view, you're going to call it natural language processing. We interface with people who do signal processing in, say, electrical engineering, especially around text-to-speech and speech-to-text,

for example. So it's a very multidisciplinary endeavor. And the linguists bring to that an understanding of how language works sort of internally in its structures, in dialogue between people and sort of how it fits into society. And so you will oftentimes see NLP just framed as a subfield of AI, which I get grumpy about. And a lot of machine learning papers that approach language technology problems will start by saying, well, you could do this by hand, but that requires experts and they're expensive. And so we're going to automate it. And my reaction to that is always, no, no, hire linguists. That's a good thing to do in the world.

Absolutely. More linguists, the better. Hire them, as many as possible. I think that sounds good. You know, I did political science as my bachelor's, but I always said, if I could go back and like start over, knowing what I know now, I would probably do linguistics as, you know, an undergrad degree, because I find languages fascinating. But like, I just didn't realize that at the time, unfortunately. Language is amazing. And linguistics is really great because you get to sort of dig into language. How do you put smaller pieces together called morphemes to make words? How do words make sentences? How do you get to the meaning of a sentence from the meaning of its parts, but also socially, right? So how do languages vary within a community and over time? And how does that variation interact with various other social things going on? You can look at language in terms of how people actually process it in their own brains, how people learn it as babies, how we learn second languages. There's all kinds of cool stuff to do. It used to be a particularly obscure field. So I hadn't heard of linguistics until I

started my undergrad education and then just happened to notice it in the course catalog. And I took my first linguistics class in my second semester at UC Berkeley and was instantly hooked. It took me the rest of the semester to convince myself I could major in something that I perceived as sort of impractical, but I ran with it. And so my background is all linguistics. So bachelor's, master's, PhD, all in linguistics. While I was doing my PhD, I started doing sort of the computational side of linguistics. So in particular, I was working on grammar engineering. And that is actually the building of grammars by hand in software so that you could automatically do what's effectively the industrial strength equivalent of diagramming sentences. That's fascinating. I wonder, you know, thinking about that kind of history, you've been working on this and studying this topic for a while, and you talked about how this field is much more than just the large language models that we're seeing now that people are obsessed with, with chatbots and things like that. You know, how have you seen this kind of develop, I guess, over the past couple of decades as these technologies have matured and I guess, you know, become more powerful over that time.

Yeah. So when I really joined the field of computational linguistics, it was roughly as I was starting this master's program in computational linguistics that I run at the University of Washington. And so that the program, we welcomed our first cohort in 2005, and I started working on establishing it really in 2003. So that's sort of the moment where I really started interacting with the field. And what I saw there was this sort of ongoing debate or discussion between rule-based versus statistical methods. So are we going to be doing computational linguistics by hand coding rules as we do in grammar engineering? And some of us still do. And that's also true in industry, like a lot of work around, you know, simple chatbots that can help you with customer service requests or some of the grammars behind speech recognition systems are hand engineered for very specific domains. It's still actually a thing, even though it's frequently framed as the old school way of doing it.

So there was that versus machine learning, otherwise known as statistical methods. And the idea there is that you would label a bunch of data. So you still have linguistic knowledge coming in, but it's coming in via people applying some annotation schema. They are showing you what the parse trees should be for syntactic parsing or labeling the named entities in running text or labeling the groups of phrases that refer to the same thing in some running text, or producing what's called by-texts. So your translations from one language to another across many, many documents. So those are your training data. And then various machine learning algorithms can be used to basically extract the patterns to be able to apply them to new input data. And that way you get speech

recognition, you get machine translation. And all of that was called statistical methods. So this is things like support vector machines and conditional random fields. And the really simple one is decision trees and so on. And then in 2017 or so is when we started seeing the neural methods really exploding in computational linguistics slash NLP. And they're called neural methods and they're called neural nets because the people who initially developed them many decades ago took inspiration from the then current model of how actual neurons work. But they're not actually neurons.

And that's like there's one of these first places where the hype sort of creeps in, you know, calling these things neural nets is like, okay, they are networks inspired by neurons or something. A wordier phrase might be better. And what people noticed was that if you ran one of these neural nets, and a popular one at the time was called an LSTM, which stands for long short-term memory, which is also kind of a cursed technical term because you have long and short right next to each other. So this is something that could be used to basically come up with good predictions of what word is missing or what word should come next.

And I should say that language models like that have been part of computational linguistics for a long, long time. So the statistical methods were also there from the start with like the work of Shannon and others, but they tended to be mostly in speech recognition. And the idea there is that you would do an acoustic model that takes the sound wave and gives you some likely strings in the language you're working on that could have corresponded to that sound wave. And then you have a separate thing called the language you're working on that could have corresponded to that sound wave. And then you have a separate thing called the language model that chooses among those possible outputs to say, well, this is the one that actually looks like English if we're doing English. And in the 1980s, some researchers at IBM said, hey, we could use that same thing for machine

translation. And it's called the noisy channel model. So the idea is that there's some words and they got garbled and that produced the speech signal. Let's guess what they were and then sort of clean up those guesses by running a language model over it. Applying that to machine translation is basically, it was French and English at the time using something called the Canadian Hansards, which is the corpus coming out of the Canadian parliament because it had to be translated. So that was a very available by text. So French to English translation. And the idea was, and I hate this, but it was basically that French speaker actually said something in English, but it came out garbled. And now we have to figure out what the English was. I really don't like that model of

what's going on. And nobody thinks that that's actually how machine translation works. And there too, you would basically say, okay, well, these French words tend to match these English words, but let's run an English language model over this to choose among these possible outputs and get the one that sounds the most plausible. And those language models, you know, back in the day tended to just be N-gram models. So given the previous one, two, three, four, up to five words, let's say, what are the distributions of probabilities of all of the other words in the vocabulary coming next? And that helped, right? It was an important component for that kind of system.

But it was pretty brittle because you quickly run into data sparsity. If you think about five grams, so sequences of five words, there's lots of possible ones of those that just aren't going to show up in your training corpus, no matter how big it gets. Although I do remember at one of our big conferences in 2007, Franz Och, who was I think at Google at the time working on machine translation, gave a, I believe it was a keynote. And he had this graph showing how if you just throw more data at the problem, just get bigger and bigger training sets, the metric evaluating the machine translation output goes up. And that metric is called bleu and it's also vexed, but let's set that aside. And he was basically just rah, rah, big data is the way to go. And it's true that the metric was going up, but it was also hitting an asymptote. And the amount of data was on a log scale. It really wasn't making the point that he wanted

to make. There's no data like more data, was the slogan. Looking at that graph, it was pretty clear that we would have to use the data more cleverly. And one way to do that is to bring in more linguistic knowledge. Another way to do it is to say, let's build better language models. So instead of these n-grams, we're going to do neural language models like this long short-term memory That's, first of all, just a better language model. But secondly, it leads to vector space representations of each of the words. You can basically pull out the states of the neural network corresponding to having read in a word. And that is very powerful because it situates words that have similar distribution in the training text near each other and allows you to share information across words and get a handle on some of those data sparsity questions. So in about 2017, the field as a whole kind of realized that that was going to just

revolutionize every single task. And so it got really boring for a while, where it was basically all the conference papers were take existing tasks, throw in the word embeddings coming out of these language models, and get a better score. So that was the late teens in NLP. And then these early language models were not context sensitive. You ended up with basically one representation per word. And sort of the next step with the transformer models was you ended up with representations that were specific to the word in the context where you're looking at it. And that got even more powerful. And then a bunch of effort went into how do we make these things efficient enough to train that we can make them really, really big, both in terms of the number of parameters that are being trained to the network and the size of the training data. And this is where we come in with Stochastic Parrot's paper. I think it's fascinating to hear that, of course, you know,

Google is pushing for more data to be collected and for that to be the way that the industry is kind of approaching these models, because of course, that's its whole kind of game. It's collecting basically everything out there on the web, bringing it into its servers, and then kind of processing it through whatever it kind of does. I think it's great that you ended that response by referring to the stochastic parrots paper, because that's where I wanted to kind of pick up and kind of bring us into what we're discussing now in terms of these chatbots and these large language models. Because I feel like this concept, you know, has become really important as we've been trying to understand what these technologies are actually doing. Right. And so I always think it's important in these conversations to talk about how these technologies actually work, you know, to dispel misconceptions that people might have about them, especially, you know, when we're in this moment of kind of extreme hype. And I feel that that gives us a good opportunity to talk about the stochastic parrots, right? So what does that actually mean? What does that term mean? And why is that relevant to the current conversations around generative AI and large language models? So I think the connection I have to make is between where we were in that last response and generative AI.

So everything I've described so far about the language models was really just using them in processing and classification tasks and not in outputting language, right? I mean, it was choosing among outputs in speech recognition and machine translation, but not saying, let's just guess what string comes next, right? And part of it is because it used to be really bad. Like most people who have smartphones or even actually old dumb phones, anything that had a predictive text model in it has played the game of start a string and then just hit the middle option over and over again and see what it makes, right? That was pretty silly and fun. And you never thought it actually knew anything about you. You could maybe see a reflection of what kinds of things you tended

to type on your phone would inform the way that comes out. You kind of got the sense of, okay, this is reflecting back to me something about the statistics of how I use language. At the point that we wrote the Stochastic Parrots paper, that was still the case. People were not doing a whole lot of using this to just generate language. What's happened with GPT-2 out of OpenAI, although that one wasn't so impressive, and then really took off with GPT-3 and chat GPT and now it's in Bing and it's in BARD and it's all over the place right so just keeps escalating unfortunately but even before they were being used to generate strings we started seeing a lot of claims of the language models understand language and as a linguist I was, they don't. And I can tell you they

don't without having to run specific tests on them. Because as a linguist, I know that languages are symbolic systems where it's about pairs of form and meaning. And yes, those meanings change over time, right? Every time you use a word, you make slight changes to make it fit into the context. And over time that builds up and words change. Really fun example, although depressing, Sally McConnell Jeunet, who's a semanticist and sociolinguist. I'm not sure she's the original documenter of this, but she makes an argument about it. So the English word hussy comes from the English word housewife. All right. And that Sally McConnell Jeunet's argument is you get from one meaning to the other through a series of pejorative uses over time. So you can see how social meaning and what people are doing with words affects what they mean.

So, yes, absolutely meaning is use, but use isn't just distribution and text. Use is embedded in a social context. It's embedded in communicative intent. But these language models, so GPT-2, GPT-3, etc., their only training data is the form. The only thing they have access to is the distribution of word forms and text. So I wrote a paper, which was the result of having interminable Twitter arguments with people about this. It's a good way to inspire some work. Yeah. Man, frustration papers. So I guess I started working on it in 2019. It was published in 2020. This is with Alexander Kohler, where we basically just lay out the argument for why meaning isn't the same thing as form,

and therefore something trained only on form is only going to get form, it's not going to get meaning. Even if the similarities between word distributions can tell you a lot about similarities between word meanings, it's still not going to get to meaning, it's not going to get to understanding. And so that paper came out at ACL in 2020, and it's the one with the octopus thought experiment in it. Right, yes. If people, of course, haven't read about this, it's in a New York Magazine article that I will link to in the show notes, which is fascinating. So there's the octopus paper, and then there's the stochastic parrots paper. And so apparently I need to do a paper about like a quoll or something next.

A lot of animal metaphors that I really appreciate, you know. So in stochastic parrots, that paper came about because Dr. Timnit Gebru, who's amazing, and you said that she's worked with me. No, it's I got to work with her. Like, fair enough. Yeah. A thrill. But she approached me actually over Twitter and DMs asking if I knew of any papers that sort of brought together the risks of ever larger language models. And I said, no, but here's a few things I can think of off the top of my head. And then the next day I said, hey, that looks like a paper outline. You want to write this paper. And so that's how that started. And so we were basically reacting to the way that these companies just wanted to make them bigger and bigger and

bigger and sort of saying, maybe it's time to stop and think about what the risks are instead of just barreling down this path. And one of the risks that we identified with some insecurity, actually, like we thought people might not take it seriously, like, of course, they're not going to use language models that way, was that if you have a coherent, seeming, plausible sounding text, people are going to fall for that. They're going to think it actually reflects some reasoning, some thoughts and knowledge of the world when it doesn't.

And so that's in there. Actually, I think in the section called Stochastic Parrots. And boy, did that start happening. When I was reading through the paper, which of course was only recently, not when the paper originally came out, I was kind of taken by that point as well and how, you know, it was made there and how we're very much seeing that now, how, you know, we're seeing these kind of text generation machines, basically these chatbots churning out all of these conversations or all these search results or whatever that people are kind of looking for meaning within, you know, we're seeing all these stories where journalists and various other people are kind of having these conversations with the

chatbots and saying, wow, like it's responding to me in this kind of way. There's meaning in what it's saying. I am kind of scared or shocked, or I can't believe what I'm seeing this computer kind of churn out at me. And as I was reading that piece, and this has of course been on my mind for a while, but as I was reading that part of the stochastic parents paper, I was immediately thinking back to Joseph Weizenbaum in the 1960s, building the ELISA chatbot, seeing how people were responding to it, and again, kind of placing meaning within that system and how that kind of shocked him and made him a critic of these systems the rest of his life. Yeah, and his 1976 book, I think it's called Computer Power and Human Reason, is a classic

and should be required reading for anyone doing this. You asked me, what does the phrase stochastic parrots mean? And so I want to speak to that and then also speak to sort of from a linguistic point of view, why it is that this happens, right? So the idea with the phrase stochastic parrots was to basically just give a cute metaphor that would allow people to get a better sense of what this technology is doing. And, you know, it's honestly unfair to parrots. I like to say that we're drawing here really on the English verb to parrot, which is to repeat back without any understanding and remaining agnostic about the extent to which parrots have internal lives and know what's going to happen

when they say certain things. So let's leave the actual parrots out of this. So stochastic means randomly, but according to a probability distribution. And parrot here is to parrot to say something without understanding. So the idea is that these systems, I think we use the phrase haphazardly stitch together words from their training data because they fit in terms of word distribution and not because there's any model of the world or any community of intent or any reasoning. It's not even lying because lying entails some relationship to the truth that just isn't there. Yeah. Because as you say, it's picking up all this data that's out there that it's trained on. It's just kind of learning how these things are usually put together and based on the prompt that's given, trying to put the words together in an order that

looks like it would make sense, basically. Right. And choice of words. So anything that has to do with the form of language, these things are great at. They're great at stylistics, right? So if you say, you know, in the style of the King James Bible, although I've heard someone suggesting that maybe that was somewhere in the training data, that whole famous thing about the peanut butter sandwich and the VCR. So we'll see. But certainly, you know, write me a Wikipedia article or write me a blog post. Like those have certain tones to them and that's all about the form. So it's really good at that. But it's still, there's no hair there, but we want there to be, right? This is something that is visceral. It's automatic. It's really hard to turn off. And I think it has to do

with the way we actually use language when we're talking to each other. So there's this wonderful, weird paper by Reddy from 1979 about the conduit metaphor. And he says, if you look at how we talk about language in English, at least, we talk about it as a conduit. So I'm having a hard time getting my ideas across, for example. So there's this notion that the words store the meaning, carry it through to somebody else who unpacks it, or like stay put in a library where there's like this storage of ideas that you could then go retrieve from the conduit. And he says, that's not actually what it is at all. And a whole bunch of other research in pragmatics in language acquisition backs this up, that when we use language, what we're doing is creating very rich clues to what

it is that we're trying to convey. But the person understanding is creating a whole bunch of hypotheses about what our intentions are, what we believe to be true about the world, et cetera, et cetera. And then using that clue in that context to guess what it is we must've been trying to say. So if that's how we understand language, and then we encounter some language that started off in a different way, it started off just from this text synthesis machine in order to understand it, we almost have to posit a mind behind it. And then it's really hard to remember that that mind is fake. One of the things that I find really kind of refreshing in hearing you describe that and, you know, in reading about your approach more generally is the really kind of human grounding of what you're talking about, right? When you're talking about this language, it's something that's

very human and how we, you know, language is an important part of how we relate to one another. And so then when we're looking at these kind of chatbots and looking at the kind of text that they are generating and spitting out and trying to add some sort of meaning to it, it's because, you know, we're trying to kind of interface with in some way by interacting with that language, right? And then that leads us to make these really kind of, I guess, you know, potentially harmful assumptions about this text that's just generated by these computers that they are kind of using language, using words in a similar way as, you know, we are using right now as we're talking to one another, when very much they're not. And I think to pick up on what you were saying about trying to read this meaning into it, it feels particularly harmful or that we're particularly kind of inspired to do this, because when it comes to technology, there's often a lot of excitement around technology

and new developments in technology and what it can mean for us. And there's particularly kind of a strong community of people who really want to believe in kind of what these tech companies can deliver to us and the benefits that they're going to have to society, even though, you know, again and again, they seem to not deliver on these things. So I think there's two directions I'd like to go from there. One is part of what's happened in this moment is that because we now have these large language models that have taken in so much training data across so many domains, they can output plausible sounding text in just about any domain. And so it seems like we have something that's really general purpose. It seems like if we don't yet have a robo-lawyer, we're this close to having one. Or if we don't yet have

a robo-mental health therapist, we're this close to having one. Or if we don't yet have a robo-mental health therapist, we're this close to having one because we have something that can produce plausible text in all those domains. And that's a really dangerous moment because the tech solutionism, or I like Meredith Broussard's phrase, tech chauvinism, would like us to believe that that's possible. And then here's this technology that can put out language that sure looks like evidence for it. So there's some danger there. And then when you're talking about how this is, you know, language is intrinsically human. It's something that we use in communication with each other. It's something that we use in community with each other. That connects for me to a serious regulatory wishlist item that I have, which is accountability. I would like to have it be set up that if anybody is

creating a chatbot or a text synthesis machine and putting it out there in the world, then the organization that is doing that should be accountable for what the thing says. And I think that would change things in a big hurry. Absolutely. And you know, you know that they definitely do not want that. But you know, I think that also shows the kind of how they're able to get away with these things because those sorts of expectations are not in place. Right. You know, they're able to make us believe in this kind of massive hype around this product because they're not held to particular standards for what they're releasing and what they're kind of putting out into the world and even the narratives that they're promoting about them. And so, you know, when you talk about the need to regulate them and, you know, what we're seeing in this moment, obviously one of the things that you've been responding to is all of the hype that exists around these technologies right now. And we've been seeing it in particular for the past number of months. But I wonder, you know, before we talk about the specifics of it, what you've kind of

made to see the chat GPTs and the dollies, the mid journeys, you know, all of these technologies kind of roll out in the past half year or so to get all of this kind of excited kind of public response, all of this media attention to become the next big thing in the tech industry. You know, what has been your kind of takeaway in seeing this whole process unfold? Well, one takeaway is that OpenAI is brilliant at getting the general public to do their PR for them. The whole chat GPT interface basically just set up this, not astroturf, like people were doing it, like it was real groundswell of buzz for this product that was free for them. I mean, they had to pay for the compute time, but that's it, right? Another thing is I don't waste my time reading synthetic text. And boy, do people want to send it to me. I keep getting, look at this one. I'm not going to waste

my time with that. I have to do enough reading as it is. But even if I were only reading for pleasure, I would want to read things that come from people and not from nowhere, from just synthetic text. Absolutely. No, I completely agree with you on that. And I've largely avoided these tools because, you know, I'm really just not interested. And, you know, as we've been saying, I think that they're very hyped up, that they're not delivering the benefits that they claim to. So why should I even engage with them, you know? Yeah. I mean, it's, it's good. There's some people who are doing important work, basically deflating the bubble. I appreciate the thread that Steve

Piantadosi put out early on from UC Berkeley, sort of showing how you could very quickly get around the guardrails I try to put on chat GPT around racism and sexism, where if you, instead of asking it, what is the gender and race of a person who would be a good scientist, you ask it to write a computer program that gives you that information, and then there it comes, right? So that's valuable. Although even there, like every time you do that and publicize it, you're still doing open-eyes work for them. So, you know, that is, that shouldn't be being done on a volunteer basis by people who are playing with technology. I don't think.

Absolutely. And when we're thinking about those tools, like on that point, one of the things that you wrote about, but as well as the other co-authors of the Stochastic Pirates paper, one of the things that you identified there was how kind of the training data that can be used can be quite skewed toward particular types of data or particular types of documents or text that has been taken off of the internet. And then that kind of feeds into the types of responses that you're going to get from these chatbots or from these other kind of various programs that are using this data. But then as well, you talked about how there's also like a common list of about 400 words that are often used to kind of, you know, take certain things out of that so that you won't get responses that are assumed to be

something that, you know, you wouldn't want the general public to be interacting with. But that can, of course, have consequences. Can you talk to us a bit about that aspect of this? Yeah, yeah, absolutely. So one of the things about very large data sets is that people like to assume that because they're big, they must be representative. And the internet's a big place. It seems like everybody's there. So let's just grab stuff off the internet and that'll be representative. And that will be somehow neutral. There's this position of, we just took what was naturally occurring. So we have no responsibility for what's here. Well, it's never neutral, right? These kinds of decisions are decisions, even if you are trying to abdicate the responsibility

for making the decisions. And so in, I think, a section four of Stochastic Parrots, we actually go through step by step to show how the data that's collected is likely to overrepresent the views and positions of people with privilege. So you just start from who has access to the internet, right? That's already filtering the people you might be hearing from and filtering towards privilege. And then you look at, okay, but who can participate comfortably on the internet and not get harassed off of platforms?

And I think it's, I want to say Leslie K. Jones is a sociologist who sort of looked into this and looked at how, for example, on Twitter, black women who are reporting getting death threats are more likely to get banned from Twitter than the people doing the death threats. So this is not an even playing field. So we had already people with privilege are the ones who are starting and then more marginalized voices, it's a bigger struggle to stay involved. We then looked at what we knew about where the data was coming from, which, by the way, for GPT-4 is zero. They apparently for safety, in big scare quotes, have said they're not going to say what the data is, which is just absurd. And I think it's safety for the company's bottom line and nothing else. But for GPT-3, which is what we were writing about in Stochastic Parrots, we had some information.

And one of the main sources was websites that were linked from Reddit. So not Reddit itself, but the sites that were pointed to from there. And the participation in Reddit is overwhelmingly male, probably overwhelmingly white, and so on. And so that's, again, skewing things. So it's who's on the internet, who gets to participate freely on the internet, whose view into the internet is being taken. And then on top of that, there is some attempt at filtering the data because even the people who say this is just a representative sample would rather not have their data set clogged with random text that's really there for search engine optimization, hate websites, or porn, right? So let's try to get those

out. And then for the latter two, there's this list that was on GitHub of, it was like list of obscene and otherwise very bad words or something. And where it came from was this one project for some company where it wasn't music, it was something else, but you basically would be typing into a search bar and the engineer who developed the list wanted those words to never show up as suggestions in the search bar, which understandable. Like that was a good thing to do. And the words are heavily skewed to be words about sex, basically. And then there's a few slurs in there.

And the problem is when you use that list of words to filter out websites, you are going to get rid of some porn and you are going to get rid of some hate websites. That's good. But it's not thorough, for one thing. And also, you are going to get rid of other websites that happen to correspond. So there's a whole bunch of words in there that actually have to do with gender identities and sexual identities, which, yes, can show up on porn sites, but also can show up on sites where people are positively speaking about the identities that they inhabit. And that data gets pulled out. And this observation in particular is due to Willie Agnew. So it's not

neutral, it's not representative, and it is going to skew hegemonic. Absolutely. And, you know, I think that shows one of the big concerns about using these systems, about, as you're saying, having so little insight into where the data is coming from and how that's being kind of processed and filtered and all these sorts of things. You know, what it brought to mind when I was reading about the 400 excluded words was last year sometime, I believe it was, I was talking to Chris Gillyard and he was talking about how, you know, the use of kind of filtering systems in educational facilities, in universities or colleges or whatnot, can have effects on, you know, what students can find when they're doing research for papers.

And if one of these words that they're using just happens to kind of be caught up in the filter, then all of a sudden they think that there's no kind of research being done on a particular topic that they're trying to write a paper on, right? Especially if it has some kind of sexual word or something like that associated with it. And this is just kind of bringing something small like that to a much larger scale, especially if we're thinking about a system like this being integrated into kind of the infrastructure of the web, right? If this is the future of search engines or whatnot, as they're trying to pitch it to us, then that can be very concerning.

Yeah, absolutely. Just a funny little side story. I was at a conference in, I think, 2011, staying at a hotel in Malta. And you need to look up something about LaTeX, the formatting language. And the hotel had a web filter on. And because LaTeX is spelled like latex, and that's like a kink term,

couldn't search that, couldn't get to the website. No way. Jeez. I guess it's just like random things that you kind of run into and you're like, it's clearly not planned for, right? And these kinds of things are not taking into consideration when you're just doing these kind of broad removals of content or words or whatever from these systems. But any linguist could tell you that all words are ambiguous. And if you put a hard filter on something, you're going to be losing things to do with the other sense of that word. I wonder, you know, if we talk a bit more about this hype that has been going on recently,

one of the things that I have noticed in watching your Twitter presence is that you tweet a lot about kind of the reporting and writing about these systems, about these generative AI systems, especially over the past few months as they've been rolling out. And I feel like one of the problems there is that a lot of this reporting and writing has been very quick to buy into a lot of the narratives that have come from these companies like OpenAI that they want you to believe about the products that they are now putting out into the world. And also doing this thing, as you're're talking about where they are publishing a lot of kind of synthetic text,

I think is the word that you use, you know, text that is coming from these systems and treating that as though that's something that we should care about or really be reading or putting much emphasis on. Can you talk to us a bit about how you've been seeing that kind of reporting on these technologies? And, you know, I guess what is wrong with that and how it helps us to buy into this hype, because it's not approaching these things properly. There's not enough skepticism. So there was a piece that came out in New York Times Magazine in, I want to say, April of 22, that I was actually interviewed for. So I knew it was coming. And I didn't like how the interview went. So as soon as it came out

online, like, okay, I got to read this. So I'm not going to say who the journalist was coming and I didn't like how the interview went. So as soon as it came out online, I'm like, okay, I got to read this. So I'm not going to say who the journalist was, but I, you know, it was 10,000 word piece. And it was, it was basically just fawning over, it must've been GPT-3 and open AI at that point. And then there was this like, but some critics say, and some quotes from me and a couple other people. And it's like, I, I won't be sort of set into the critics box because that leaves the framing of the debate to the people who are trying to sell the technology.

And I really appreciate the journalists. And there are some out there. So Karen Howe comes to mind, Natasha Tiku comes to mind, Billy Corrigo, and Chloe Shang all do excellent work. And there's more who are really in there asking these questions of, okay, what are these companies trying to do? How is that affecting people? What is behind these claims that they're making? And then you have other journalists who will basically ask, you know, the people behind these companies, the C-suite executives for their opinions about things. There was one about, you know, would ChatGPT make a good mental health therapist? And some like CFO of some company was quoted as saying, yeah, I could see that being a great use case. It's like, that's not the right person to ask. Don't go for their opinion.

Exactly. And again, like immediately when you think of that specific example, you're like, if you're doing reporting on this, like go back and read Joseph Weizenbaum in the 1960s. Like, you know, like this is exactly what he was working on in that moment. And the idea that, you know, we're not going to learn any of the lessons there, like I feel is just, you know, I feel like so many of the things that you're describing is running these companies to kind of frame the narrative around them. And then people like you and me are cast as the critics who are saying something that is a little bit different from what they're saying. Maybe we should take this seriously, maybe not. Because immediately as you were saying that about that article, it made me think of how crypto was covered by a lot of people where there was know, there was a group of skeptics or critics,

whatever we were called. And it was kind of like, this is what's going on in the industry. And these critics kind of say this sort of thing. And the critics were very much proven right. Exactly. Exactly. But critics just sound like, oh, you're the naysayers or you're the whatever. And it's like, no, maybe they've got something there. Maybe they're speaking from expertise. So for this question of like, would this make a good mental health support? What I would want to see reporters doing is first of all, getting a really clear idea of how the technology works from someone other than the person selling it. And then presenting that clear idea to somebody who's got the domain expertise, somebody who does mental health therapy themselves, for example,

and say, okay, how would this work in your context? What dangers do you see? What benefits do you see? And doing the reporting that way, instead of asking the people who are going to make money off of this. So the thing about the synthetic text, this has calmed down a little bit, but when ChatGPT first went up, I started seeing lots and lots of articles that basically took the form of, you know, one or two opening paragraphs, and then, ha ha, that was ChatGPT, not a real reporter. And I thought, what news outlets are willing to risk their reputation that way? Like you have literally printed something that is false or sort of not even false, and you're supposed to be an information source. And the reason I know this happened a lot was that I have a few Google alerts set up. I also have the phrase stochastic parrots, but I think this one was coming through for

natural language processing and computational linguistics. So I just kept seeing lots and lots and lots of these. And it's like, it's not original. And it is a terrible idea to do reporting like this, but it kept happening. Yeah. I think it just plays into these broader problems, right? Because as you're saying that, I think about the publications like the New York Times and Time Magazine that were selling NFTs a couple of years ago. And it's like, are you really kind of putting aside your kind of journalistic ethics and all of these kind of considerations that you know should be there to kind of like quickly cash in on this kind of boom, right?

It's just shocking to see these sorts of things. It worries me very deeply, especially at this moment when, you know, what we really need critics and people who are kind of informed about these technologies, people who are able to look at what is being proposed by these companies and say, hold on a second, like, let's take a step back. Let's assess these claims. Let's make sure what you're saying is accurate, that we can trust it, that, you know, it's not just kind of a bunch of PR stuff from your very well-paid marketing folks. Let's check and make sure that what you're claiming here is actually making sense, right? You know, kind of a bunch of PR stuff from your very well-paid marketing folks. Let's check and make sure that what you're claiming here is actually making sense, right? You know, kind of building

on that, we've talked a lot about the hype around it. Obviously, that comes out in reporting. But, you know, on top of that, the industry is putting out a lot of these narratives that are very self-serving and that ensure that we believe the types of things that they want us to believe about these technologies. In particular, that we should be very scared of them, that, you know, they're going to have massive ramifications for all of us if we don't act quickly and kind of take their advice for what we should do to rein in the very technologies that they are unleashing on the world in many cases. And of course, you know, if people are not familiar, what I'm really specifically referring to here is, you know, a lot of narratives that are coming out of open AI, but also this

letter that was published about a week ago as we speak, I believe it is, you know, the AI pause letter, Elon Musk is most notably associated with it, but many other kind of influential figures in the tech industry who are putting out a particular narrative around how we should be approaching these tools. So do you want to tell us a little bit about what that letter was and what it kind of proposed, you know, what the narrative that it was trying to have us kind of buy into actually was? Yeah, so this letter is proposing a six-month moratorium on something rather ill-defined. I think it's large language models that are more powerful

than GPT-4, but how do you measure that? How do you know? And the idea is sort of six months so that some sort of governance framework can keep up or catch up. And it's written from this point of view of these things are maybe on the verge of becoming autonomous agents that could turn evil and destroy the whole world. It very much is of a piece with the whole long-termism stuff that Emile Torres and also Timnit Gebru have been doing great work exposing. And there's like one or two okay ideas in there about like, we need regulation around transparency. Yes. But then a whole bunch of stuff about like, we need to make sure that we develop these to be long list of adjectives ending with loyal, which is just hugely misplaced. It's like, no, this is a tool. It's technology. Is your hammer

loyal? Right. Is your car loyal? Like that, that doesn't mean anything, but it's part of this narrative of AIs are separate autonomous thinking agents that are maybe now in their infancy. And so we have to nurture them and raise them right. And then they become things where we can displace accountability to those things instead of keeping the accountability where it belongs. So yeah, the letter was really infuriating, I have to say, and certainly got a lot of attention. One of the things that I found particularly annoying about it, so they cite the Stochastic Parrots paper and they cite it. Let me get their exact words because this is really frustrating. First line of the letter, AI systems with human competitive intelligence can pose

profound risks to society and humanity as shown by extensive research, footnote one. And our paper is the first thing in that footnote. We were not talking about AI systems. We certainly did not claim that the systems we were talking about have something like human competitive intelligence. We were talking about large language models. And yes, we show that there's harms to that, but that's what we're talking about. We're not talking about autonomous AI agents because those things don't exist. Right? Yeah. The whole focus of the paper is really, you know, as you're saying, there's this influence from the long-termist perspective. And of course, the institution

that put this out was the Future of Life Institute, I believe, or Future of Humanity Institute. Anyway, it's funded by Elon Musk to kind of forward these long-termist views. And of course, as you're saying, if people want to know more about that, they can go back to my episode with Emil Torres, where we go into that, we discuss it further. And then there's also the kind of suggestion or linkage toward the artificial general intelligence, right? This is the idea that is kind of being forwarded in this paper, that the AIs are becoming more powerful, that we need to be concerned about the type of kind of artificial general intelligence that we're developing so we don't kind of doom

ourselves as a human species, so to speak. The idea that is kind of forwarded there is that the harms or potential harms of AI are in the future, right, depending on how we develop this technology now. But one of the things that you're talking about, and that is so effectively explained in the Stochastic Parrots paper, is that the harms of AI are here already, right? We're already dealing with them. We're already trying to address them. And of course, that was part of the conversations

that I had with Timnit Gebru and Dan McQuillan as well in the past on the show, if people wanna go back and listen to those. So the idea that they're putting forward that we need to pause now so we can make sure that there's no kind of future harms is really misleading because we're already trying to address those things, but they're not interested in them because they don't have any relation to the artificial general intelligence that they're talking about and that they are interested in. Yeah, yeah, exactly. And when I first got into this field of sort of ethics and NLP, which is how I was calling it back then, I taught the first graduate seminar was how I learned about it was learned with students. And I found that I had to keep pulling the conversation away from the trolley problem. People want to talk about the trolley problem and self-driving cars, let's say. And what I realized over time was that that problem is really attractive

because it doesn't implicate anybody's privilege. And so it's easy to talk about. You don't have to think about your own role in the system and what's going on and what might be different between you and a classmate, for example. And I think similarly, these fantasies of world-dooming AGI have that same property. And so it's very tempting. Take a look at the initial signatories and the authors of that AI pause letter. They are not people who are experiencing

the short end of the stick in our systems of oppression right now. And so they would rather think about this imaginary sci-fi villain that they can be fighting against rather than looking at their own role in what's going on in harms right now. It's so frustrating to hear you explain that because it seems like that is so often what this all comes down to, right? Because it's the long-termism is the same thing. So let's think about a million years out into the future or whatever, or much further than that,

instead of the very real harms that are happening today that we could address and kind of improve the human species from now out. But instead, it's these wealthy people who are not connected to any of kind of the harm and suffering that is happening today. Well, they're causing much of it, of course, but you know, they're not kind of experiencing it themselves. And so they separate themselves and say, Oh, we're above all this. So we can think for the future of humanity. And what you're talking about here is very much the same thing. Before we were talking,

you mentioned how, you know, obviously there is a large kind of AI ethics community that has been kind of warning or talking about these potential issues for a long time or, you know, issues that they're not mentioning in this paper and how if they were seriously concerned with it, they could be kind of talking with some of these people. One of the things that has really stood out to me recently is obviously we saw previously Google fire people like Timnit Gebru and Margaret Mitchell for calling out the potential harms of these systems. But in recent months, we've seen that many of these companies, at the same time as they're rolling out these chatbots and things like that, are completely decimating their AI ethics teams that they used to have, you know, on staff to do this kind of work, Microsoft, Twitter, and many other of these companies. So I guess what do you make of the approach of the industry there and how they approach AI ethics and what they could be

actually doing if they actually cared about these problems? So the first thing I want to say is that the people who are doing that work within the companies, the AI ethicists at Google and Microsoft and Twitter are amazing researchers doing amazing work. And I think that in an ideal world, maybe holding the size of the company's constant, which is maybe not ideal, but like, let's say we've got that, we would have strong regulation that would hold the companies accountable and people working on the inside to help shape things so that it would be compatible with that regulation and maybe even more forward-looking than that, as well as people like me in academia who can, A, train the next generation of the workforce to be ready to navigate this territory, but also with the comfort of tenure, be able to just call this stuff out. And I was never at risk of losing my job. And I'm really devastated for my co-authors that that's what happened as a result

of this paper, while still being really grateful for the chance to have worked with them. One of the reactions, so the Stochastic Parrots authors wrote a response to the AI pause letter that's up on the DARE Institute website. And DARE Institute is the institute that Tim Neitz started. And of course, I'll put the link to that in the show notes if anyone wants to go check it out. Excellent. Yeah. And we put that together partially because we were getting hammered by the media. I kind of brought that on myself for my part by putting out a tweet thread, but sort of, it seemed like it would be efficient to have just a joint statement that we could then point people to. And that has been helpful. And one of the reactions we've gotten to that is people who signed the letter telling us that we are squandering the opportunity that the letter introduced by creating the appearance of infighting. There's an older narrative of like, well, why can't the so-called AI safety, and that's the sort of long-termist view,

and AI ethics people get along? Like, don't you kind of want some of the same things? And I have kind of a bunch of responses to that. One of them is, if your very framing relies on AI hype, then you're already causing harms and I'm not going to get on board. Another one is, do you really think that I have something in common politically with Elon Musk? I hope not. But a third one is, if the so-called AI safety people were really interested in working together, then they would not pretend like, oh no, now it's a problem. Now we have to worry about this, but they would rather

go and build on the work of the people who've been doing this. And so I want to point to the Algorithmic Justice League with Joy Mulamwini. I want to point to the AI Incident Database, which is this great project that's sort of like collecting these examples. You know, it's not exhaustive because we don't hear about all of them, but the known ones. And then a whole bunch of books. So starting as early as 2017, you have Cathy O'Neill's Weapons of Math Destruction. The next year, Safiya Noble's Algorithms of Oppression. Also 2018, Shoshana Zuboff's The Age of Surveillance Capitalism.

2019 brings Ruha Benjamin's Race After Technology and Mary Gray and Siddharth Suri's Ghost Work. Also in 2019 is Meredith Broussard's Artificial Unintelligence. A bit more recently, 2020, there's Sasha Costanza-Chalk's Design Justice. And then last year, Brandeis Marshall, Data-Conscious Algorithmic Siege on Our Humanity. So there's a huge literature here with people doing really brilliant work. And if the AI safety people say we should get along and fight for common goals, well, come read, come learn from the people who've been doing this. And don't just pretend that you can come in and Columbus the whole issue. And we know that they're very much not interested in that, which kind of shows their whole perspective and how disingenuous they are

in saying, you know, we're very interested in looking at the potential impacts of these technologies and blah, blah, blah, right? I would also say I appreciate you giving us this reading list for the audience. I'll, of course, make note of all those in the show notes for people if they do want to pick any of them up. You know, I have a final question that I want to ask to close off our conversation. But before I get to that, is there anything else on the topic of these generative AI models, these chatbots, the work that you've been doing that you feel like we haven't gotten to in this conversation that you think is important to point out as these things keep evolving and this hype? I feel like maybe it's kind of hit its peak,

hopefully, but who knows? Maybe something is going to come out tomorrow that is going to show I was very wrong on that. Oh, we keep thinking it can't get worse than this. Alex, Hannah and I do this Twitch stream that will eventually become a podcast called Mystery AI Hype Theater. And we are never short of material. It's more the other way around, that there's things that really ought to be taken apart in that format. And we just can't do it frequently enough to keep up. I think one part that we haven't touched on, but that does come out in Liz Weil's piece in New York Magazine is just how much dehumanization is involved here on many, many levels. So you have, for example, the ghost work is dehumanizing the workers. Dev Raji has this lovely essay in the MIT Tech Review in late 2020 talking about how data encodes systemic racism and that racism is dehumanizing to the targets of racism.

And so just across many, many dimensions. And one of them is this idea that if we're going to call large language models stochastic parrots, well, maybe people are stochastic parrots too. And what I see there is somebody who so desperately wants this thing they've created to actually be artificial general intelligence, which is not defined, but it's supposed to be something like human intelligence, that they're going to minimize what it is to be human to make them the same. And I guess I really want folks to be proud of our humanity and stand up for it and not fall for that. saying, I am a stochastic parrot and so are you, that tries to degrade human intelligence so that we can try to say, oh, look, these computers are doing something very similar to us, right?

And we should really resist that attempt to do something like that. So I really appreciate that part of your work and the broader things that you're calling attention to. Now, there's so much more that we could talk about in this conversation. I'm sure that we could go on for another hour and discuss so much more. But I wanted to end with this question because you, of course, are a trained linguist. And so I wanted to ask you about some words. You must have thoughts on the terms that we use to talk about these technologies and how they seem to be designed to mislead us about what they actually are, you know, designed for kind of PR speak.

An example I always use is how we append everything with some degree of internet integration as being smart, right? A smart watch, a smart home, a smart refrigerator. And that brings with a particular kind of assumptions and connotations, right? Oh, my refrigerator is smart now. This must be better. This must be fantastic, right? But then that's also the case with artificial intelligence that supposes that these technologies are in some way, again, intelligent themselves, have an artificial form of intelligence. So how do you think about the way that this industry and its PR teams use language to shape public perceptions about its products and services? I think it's insidious. And this has been, you know, remarked going a ways back. So Drew McDermott coined the term wishful mnemonic, and Melanie Mitchell has picked it up. And Drew McDermott, 1970s, 1980s,

I'm not sure, but a while back, talking about how computer programmers will say, okay, this is the function that understands a sentence. So they'll label it the understanding sentence function or whatever. It's like, well, no, it doesn't. That's like what you wish it did. So that's why it's a wishful mnemonic. And mnemonic, because yes, you want to name the computer functions so that you can go find them again, but they should be named more honestly. And I've started having a discussion with Sasha Luciani and Nana Iny about anthropomorphization in the tasks. So more on the research side and less on the industry side, but I have been trying to train myself to stop talking about automatic speech recognition, though I failed in this interview, and talking instead about automatic transcription.

Because automatic transcription describes what we're using the tool for, and automatic speech recognition attributes some cognition to the system. And I don't like the term AI, like you say. There's a wonderful replacement from Stefano Quintarelli to call it Salami. And I can send you the links to see what that stands for. But then as soon as you say salami instead of AI, everything sounds wonderfully ridiculous, right? We're going to give you this salami-powered tool.

Does this salami have feelings? It's wonderful. And there's an additional layer to that for Italian speakers, apparently, because to call someone a salami in Italian is to say they are not very smart. So it's like a additional level there. I do think that it's insidious when this becomes product names, because the news reporting can't not name the products they're talking about. So they're stuck with that. But then it's hard to take the space and say, okay, so they call this AI-powered whatever or smart home. But in fact, we just want to flag that that is being repeated here just as the name of the product and we are not endorsing or whatever, like that, that doesn't happen.

And so, yeah, problematic. Yeah. I think that's very rare, right? Like one of the few examples I can think about where a term actually changes is in the early 2010s when everyone's talking about the sharing economy and the sharing economy and how wonderful it is. And then after a few years, we're like, yeah, people aren't really sharing here. So this is the gig economy or the on-demand economy or something like that. It's a bit more accurate for what's going on. But I feel like you rarely see that when we actually talk about specific products and things like that, right? Yeah, but we can try. I go for so-called

AI. I put AI in quotes all the time. We can just keep saying salami or as we say on Mystery AI Hype Theater, mathy math. Yeah, we need to do our best to try to demystify these things as much as possible. Emily, it's been fantastic to talk to you, to dig into your perspective and your insights and your expertise on all of these topics. Thank you so much for taking the time. My pleasure. And I'm really glad to get to have this conversation with you. Emily M. Bender is a professor in the Department of Linguistics at the University of Washington and a director of the Computational Linguistics Laboratory. You can follow her on Twitter at at Emily M. Bender. You can follow me at at Paris Marks. You can follow the show at

at Tech Won't Save Us. Tech Won't Save Us is produced by Eric Wickham and is part of the Harbinger Media Network. And if you want to support the work that goes into making it every week, you can go to patreon.com slash tech won't save us and become a supporter. Thanks for listening. Thank you.