Are We Overrating A.I.? Which Movie A.I. Gets It Right? 09.26.23

your cultural disposition. It's also National Johnny Applesey Day. And only bring that up because I remember in Christian school we were made to sing that as like a song before we went to lunch.

about peacock pie. For some reason, he was lobbing that meatball up to you. Yeah, I saw in the discord. I was tagging that, but you took that. Oh so you just do other people's a case. Okay, introting? All right, Well I'm throws to be joined as always by my co host mister Miles Gras.

gender studies, critical race theory, and Asian diaspora studies. He is also a research fellow at the AI Now Institute, the co editor of the upcoming volume The Good Robot, Feminist Voices on the Future of Technology, and the co host of the Wonderful Good Robot podcast. Please welcome the brilliant The talented doctor Kerry mcinnern doctor Belly.

from childhood. But it does mean I didn't have the great apple seed song that you.

all this like fancy shit. I just want to beat people up or just mash the keypad over and over and win like that, and you can you can't do that because I think you know it is like it is about being able to play at different skill levels rather being like, oh, you don't know how to multi use the force while you know you're doing your melee attacks. Come on, some of us just want to be the kids.

Unraveled is good for that too, of like a fun video game that is not like really requiring you to have like these sort of highly developed video game skills and it's an easy play.

live actions. But you know, maybe I'll be proven like totally wrong and the next ones will be amazing, But I'm ready for something else.

We know what is possible here, right. I can't I can't imagine a version like what they're gonna pull out where we're like, oh, no, that is not did not see that one coming?

it to be a more nuanced stories. But sometimes it's just really satisfying and you want this clean cut narrative of this like good guy to beating all these bad guys. When I want to relax, I really enjoy that. In my day job, I think about lots of like complexity and nuance and like what we do with the future of society, and so you know, right, sometimes I find it very relaxing just to be like sonless.

with open AI and or chat GPT. So I had this kind of disconnect in my mind where it was like from an outsider's perspective. We have this C plus level like copywriter thing with like in chat GPT GPT for and then like the Godfather of AI, who I'm just trusting people is the Godfather of AI. But that's what everyone uses that same phrase though, like the Godfather of AI just quit Google and says we're all fucked

in the next couple of years. And I think it's confusing to me because I don't know exactly, like I can't even like picture the way how he thinks we're fucked. And there was this letter that was like, we need to pause development on AI for in the near future, and I guess I'm just curious to hear your perspective on that pause letter and what the kind of dangers of AI are in the near future.

not the only ideas that we have about AI. We've got thousands of years of thinking about our relationship with intelligent machines, and there's a lot of different cultural traditions that have really different ways of thinking about our relationships with AI and intelligent machines that could be much more positive, much more harmonious. And so I do think our immediacy of jumping to this idea of sky net is reflective of very much where we are right now. Right I'm

in the UK, you're in the US. These are countries that have a really long history of thinking about AI and very binary terms. So yeah, I think it's important that we think about these long term risks of AI. And you mentioned the pause letter calling for a wholt to generating more large language models like chat GPT until it had a bit more of a moment to think

about to the long term consequences of these models. But I think it's really important not just to think of the long term risks, but to think about which long term risks we prioritize, because I think the sky net terminator fantasy eats up a lot of oxygen about how we talk about AI's risks. But there's a lot of different risks to AI poses. So another long term risk that we don't talk about very much at all is the climate cost of AI. Right because AI it's hugely

energy intensive. Data centers require a huge amount of water to function, and we have this massive problem of e waste produced by a lot of electronic systems. But that long term problem of climate crisis is much less exciting. It's really scary, it's really grounded, and a lot of our experiences and so it just doesn't seem to get as much airtime. So that's something that I think is

really important. It's changing the conversation a bit to say, Okay, it's sometimes interesting, sometimes scary to think about the terminator option, but what are some of the other long term options that could really shape our lives.

to pass the CAPSHA test. And like in both cases those are like the AI that killed a person in the military exercise, like that was somebody claiming that, and then when they went back they were like, oh, I was just saying it could hypothetically do that.

using to keep it hemmed in. So it does feel like there is an incentive structure set up for the people in charge of some of these major AI companies to get us to believe that shit, like to think to only focus on the AI v. Humanity, Like AI gets loose of its control of our controls for it and takes over and starts like killing people version of it.

I'm just curious, like what are your thoughts on, Like why why why are they incentivized to do that when it would seem like you, well, you don't want to make it. You don't want it to seem like this self driving car will take over and.

left behind. So you see more and more firms saying, well, now I've got to buy the latest generative AI tool so that I look really tech savvy and I look pick forward and I look futuristic. And so it's part of the bigger hype cycle, I think, to a lot of tension towards their products, but also to make them

seem like this really edgy, desirable thing. But I think what's also interesting about both the stories that you raise is when you looked under the hood there was human labor involved, right, There were people who were really central to something that was attributed to an AI. And I think that's a really common story we see with a

lot of the type around AI. Is often the way we tell those stories erases the very real human labor that drives those products, whether it's the artists who originally made the images that train the generative AI through today, the labelers all sorts of people who are really central to those processes.

that's easy. Like I can see how a company immediately goes, yeah, it's a tool, and then suddenly it's like and now you're on your ass because we'll just use the tool now and we don't even need a person to prompt it, or we need many we just need fewer people to operate it. So to me, I'm like Okay, that's an obvious sort of thing I can see like on the horizon.

And you did talk about, well, there was a lot of talk of these sort of long term existential or quote unquote existential threats, that there were a lot of things in the short term that we're actually ignoring. What are those sort of things that we need to bring a little bit more awareness to, Like I know you mentioned the climate, and I look at it from my perspective, I see like the just massive job loss that could happen.

But what are sort of like the more short term things that kind of maybe are less sexy or interesting to the people who just want to write about killer terminators and things.

already seen, and that's super boring. Like you don't want to hear about how the wheel somehow brought about some kind of inequality. We all know the big heart take from today, But yeah, I mean something that I look at, for example, are like very mundane but important technologies, like

technologies used in recruitment and hiring. So I look at AI powered video interview tools and look at how that affects people's particular you know, likelihood of being employed and how they go through the workforce, and yep, it's less exciting seeming than the terminator. But again, when you look under the hood and dig into them, you're like, oh wow, this could actually really really compound inequalities that we see in the workforce under the guise of the idea that

these tools are going to make hiring more fair. And that's a massive problem.

the really big problems that are flaggs. But sometimes the underlying idea behind their product is just fogus, Like it's just a really bad concept and yet somehow it gets brought to market again because of all this hyper AI.

So with the video interview tools that we look at, for example, they basically claim that they can discern a candidate's personality from their face and from their words, so how they talk, how they move, They can decide how extroverted you are, or how you know open you are, how neurotic you are, how conscientious, you are all these different markets of personality, to which I would say, firstly, no,

there's absolutely no when AI can do that. This is just a very old kind of racial pseudoscience making its way back into the mainstream, saying Okay, we can totally guess your personality from your face, Like it's like your friends looking at someone's profile picture on like Pinder or whatever and being like they look like they'd be really fun at a party. Like it's about that level of.

starts talking about what he does. But he was saying, like he pays attention to the market, and he was saying, there's like a big thing propping up the stock market right now is AI. And it really is like that, that's where so much of people's wealth is tied up, is in like the stock market, and it's just tied to like what you can get people excited about in

a lot of cases. Really like that. From that perspective, the incentive structure makes sense, Like you want people talking about how your AI technology can do all these amazing things because that literally makes you have more money than you would have if they knew the truth about your your product.

people money more frictionless. And that's something that you guys

have talked about on recent episodes of Good Robot. Like there's this one example that really jumped out to me that I think was from your most recent episode, or at least the one that's up most recently right now as we're recording this, where you guys were talking about a company that asked a regulating body to make an exception to a law around like a high risk use of AI, and the law said that people had to supervise the use of AI like just because it seemed dangerous,

and the company appealed to the regulating body by saying, well, we just like that would cost too much and we would never be able to like scale this and make a profit. And it feels to me like our answer as a civilization to that complaint needs to be that's not our problem, like that then you shouldn't be doing it.

But instead it seems like the answer too often, not just in AI, but just across the board, especially in the US, is like, Okay, well we have to make an exception so that they can make a profit around this technology, or else the technology won't get developed because the only thing that drives technological progress is like the

profit motive. But that's, you know, as I think you guys talked about in that episode, that's never been the best way to develop technology, Like it's it's been a good way sometimes to democratize existing technology, but like that's I don't know, I feel like that idea of you have to make it profitable, you have to make it easy on these companies to keep trying different things for AI to become profitable is baked in at a like cellular level at this point in how a lot of

you know, Western colonial civilizations operate.

X by Musk. I remember waking up and finding my little Twitter bird replaced with this huge X and just being like, oh, converirstly because it was part of you know, Twitter's low decline. But secondly it maybe you feel pretty disappointed or really aware of the fact that one guy could have such a huge impact on how literally millions of people use a social networking platform that's actually super important, you know, to their daily lives and has played a

huge role in activist movements and fostering different communities. And I think that's a story we see time and time again with some of these big tech companies, which is not only do they have their own profit motive at heart, they also they're not beholden in any way to the public, and they're not being compelled by regulation to make good

decisions that necessarily benefit the public. So I think a really important question going forward is how do we support kinds of technology development that are very much based in

the communities that the technology is for. I think one really big part of that is recognizing that so many AI models, as you mentioned right, they're designed to be scalable and that's how to make money, this idea that you can apply them globally and universally, and I think that's a big problem, partly because it often is really homogenizing. It can involve like explitting a model from the US usually out to the rest of the world, it's probably

not actually appropriate to using those contexts. But also a lot of really exciting and good uses of technology I think come from these really localized, specific, community based levels. So sometimes I think it can be about thinking smaller rather than bigger.

what am I huh like? And I think that's what's an really interesting thing about this as like a sort of technological advancement is before people really understand what it is, there is like from the higher the powers that be sort of going into it being like, well, this is it, Like everyone's using it, but I'm still not sure how.

And I guess that probably feeds into this whole model of generating as much you know, excitement market excitement about AI is by taking the angle of like everything's different because everyone is going to be using AI mostly y'all don't know what that is, but get ready. And I think that's what also makes it very confusing for me, as just like a lay person outside of the text sphere, to just be like, wait, so are we all using it? And even now, I really I still can't see what

that is and how that benefits me. And I think that's a big part of I'm sure your work too, or even like any ethicist, is to understand like, well, who does it? Ben Like, first, we're making this because it benefits who and how? Yeah, And I think yeah, I mean right now, it benefits the companies that are

making it. And it sort of feels like that's the way it's being preserved or slightly being like, yeah, you guys are gonna love this, but really we're going to benefit from the adoption of this technology.

So if you look at say like China and the UK and the US and these countries that are now thinking about what their national long term AI strategy is going to be, they also very much frame it around the idea that AI is completely inevitable, that this is going to be the transformative technology for imaging the future, for geopolitical dominance, for economic supremacy. And again, I think

as an ethesis. What I really want people to do is step back and say, I think we're actually at a crossroad, so we can decide whether or not we think these technologies are good for us, and whether they are sustainable, whether they are a useful long term thing for our societies, or actually whether the benefits of these technologies are going to be experienced by very few people and the costs are going to be borne by many.

to the floor. You know, let's just get out of the way of the big companies. You know, it feels like, but it is such a complicated technology that I think there's going to be inherent cloudiness around how people understand it and also manufactured cloudiness because it is in the overall systems benefit the overall system, being like capitalism in their benefit to generate like market excitement where there shouldn't be any basically.

just a decision tree, you know. Again, it's kind of part of that branding. But also we have these again thousands of years of stories and thinking about what an intelligent machine is, and that means we can get super invested in super cloudy very very quickly. And yeah, I don't want people to feel bad for being scared of being cloudy about these technologies, like it is dense and confusing, But at the same time, I do think that it is really important to come back to that question of

what does this do for us. So, you know, the question of like Luddism or being a lude I think is really interesting because I you know, personally, I do use AI applications. There's certain things about these technologies that really excite me. But I'm really sympathetic to some of the kind of old school ludyet who weren't necessarily anti technology, but we're really against the kind of impacts that technology

we're having on their society. So the way that new technology is like I think it would be things like spinning and weaving, we're causing mass unemployment and the kind of broader rammifications that was having for people in the UK socially, and that kind of has quite a scary parallel to today in terms of thinking about maybe what AI will bring about for the rest of us who maybe aren't researchers in a lab, but who maybe might be replaced by some of these algorithms in terms of

our work and our output.

the friction, Like friction can be really good. Sometimes sometimes your system needs friction to stop and correct itself and recognize when bad shit, when things are going wrong. But you know, there's also a history in even in the US, where corporations are racing to get to a development and ultimately are beat by open source models of technological organizing around like getting a specific solution. Do you have any hope for like open source or in the future of AI.

really hard work. We had doctor David Adlani on our podcast, who's very important figure in Mascana, which is a grassroots organization that aims to bring some of the over four thousand African languages into natural language processing or NLP systems, And he talks a lot about how the work he does with Marskana is so valuable and so important, but it's also really really hard because when you're working in that kind of collective, decentralized environment, it can be much slower,

and as you said, there can be a lot more friction in that process. But counter to this move fast, break things kind of culture, sometimes that friction can be really productive and it can help us slow down and think about, you know, the decisions that we're making very intentionally, rather than just kind of racing as fast as we can to make the next newest, shiniest product.

can't read Chinese. Therefore I don't know if this research is being done, and therefore it just doesn't exist because the larger community is like we all just think in English, like, so how do you like, you know, specifically because you know, in hearing the description of your work, help me understand, like and the listeners too, like, of how we should be looking at these things from that also from that perspective too, because I think right now we're all caught

up and like it's talking sky at and it's not you know, hold on, like there are other subtleties that actually we should really think deeply about because to your point, I feel like those are the dimensions of an emerging technology or trend or something that gets ignored because to your point, it's like the thing, well, of course it's it's unequal, of course it's racist or whatever, But what are those like what are those ways that people need to really be thinking about this technology?

discriminated against female candidates. These are really I think, avert depictions of the kinds of harms a I can do. But I think things like English language hegemony are also incredibly important for showing how existing kinds of exclusions and patterns of power get replicated in these tools. Because do an English language speaker, very crucially, they might use chat GPT and think this is great, this is what my

whole world looks like if they only speak English. Obvious to anyone who is not a native English speaker who doesn't only speak English, it's going to be an incredibly different experience. And that's where I think we see the benefits of these tools being really unequally distributed. I think it's also important because there's such exclusions in which kinds of languages and forms of communication can get translated into

these systems. So, for example, I work with a linguists at the University of Newcastle, and she talks about the fact that there's so many languages like signed languages and languages that don't have a written language, but they're never going to be translated into these tools and never are going to benefit from them. You might think, Okay, well, do these communities want those languages translate into an AI tool? Maybe maybe not, I'd argua, of course that's up to them.

But those communities are still going to experience the negative effects of AI, like the climate cost of these tools. And so I think it's just really important, like you say, to think about what kinds of hegemony are getting further entrenched by AI powered technologies.

And we've we've already talked about like some ways that AI is being overrated as like a autonomous killbot, like terminator style kilbot. You know, there's there's some other examples, like on one of the episodes of your podcast, an AI expert talks about getting an AI enhanced cancer scan and assuming the scan was like taking three D video and doing or doing like three D modeling, and it was just putting a box on a two D image.

And I believe the guests like admits that they were influenced by like Hollywood movies, and it seems like that's what people who are trying to make money off of this one. So like that we're going to have this steady push to make us overrate misunderstand what the actual promise of AI is going to look like, how what tools are going to be given to us in the near future, and like what are the tools actually able

to do? So I'd just be curious to hear from you, like what do you think ways that AI might intervene in our lives in the near future might already be intervening in our lives? Well, what are those things that we're underrating, Like what and what are the things that you think are probably taking up too much of our bandwidth in terms of how we're picturing AI.

expand beyond a very narrow set of controlled circumstances. I think this idea that you know, soon every hospital will have this tool, or say AI and education, that every classroom is going to have access to this tool, I think unfortually just really grossly over its demanding the kinds of resources only in this country that schools and the

hospitals have. I was talking to a friend who's a doctor here and she was saying, oh, well, what do you think about AI in medicine, And then she kind of stopped for a moment and just laughed and said, well, I don't know why I'm asking you that, because my hospital uses paper notes. And I said what, and she said, yeah,

our whole hospital is not computerized. And that was a good reality check for me because it made me realize, like, wow, actually some of the stuff that I'm thinking about is so far away from the reality of what people like my friend the doctor is having to deal with on

the ground. And same with schools. You know, here a whole hundreds of schools I think have had to be closed because there's some kind of issues with the concrete that I might collapse, which is a horrific But b is such a different level of infrastructural problem that thinking about AI in the costume is really not on the radar in that scenarios exactly. Yeah, autonomous, Yeah, it's like a low Yeah, people like what if students cheat? And

you're like, look, yeah, you know. So those are things that I think are maybe a little bit on the overrated side. I think a way things are underrated, Like I'm really interested in how AI can change how we see the world and how we see ourselves, like I think, you know, take for example, something like TikTok. You know, I have to confess I'm a big TikTok user. I love grolling a lot of mind list diarbage. It brings

me a lot of joy and peace. But even things like AI powered beauty filters, right, like, I think that has such a profound effect on how you just understand and see yourself and your own faith. They're often really imbued with gendered and racist kinds of assumptions as well. I often look a lot whiter when I use a

beauty filter, even though I'm multiracial. Like all of those assumptions, I think they get baked in that seep into our lives and like really subtle ways, but I think collectively can have quite a big impact.

this is just how you take pictures now, and like pictures are way better than they used to be. I think I saw somebody speculate that like some of the software like Photoshop and other things that have traditionally had sort of a difficult learning curve will suddenly get much easier for people to use. So yeah, I like those are like having actual tools that are like kind of easy to use and like very simple, straightforward. We know what the goal is here, and we're able to use

these enhanced kind of programs to achieve that goal. Like that that feels more possible to me and like something that we might see in the not too distant future.

will fuck you up. And you're like huh, and I guess that's such a different proposition up front and earlier, doctor Carry was saying, like there are things that genuinely excite you about like this sort of like these emerging technologies from I, and I would love to hear from your vantage point because you are looking at this and a like like what is ethical and what is going to bring meaningful value to people? Like what are those things?

That I can feel like, oh, yeah, yeah, I can get down with that future.

my life so much better and easier. I love being able to transcribe, you know, what I'm saying and have it appear on the screen and not have to type out my emails and have them a pair. I recognize you they're like, all very boring. I should have said something like I DJ not at all I write email

of using voice. It's very you know, but I think kind of, you know, extrapolating out from that, you know, there's really amazing applications and accessibility, and when it comes to the kinds of access people can now have online because of advancements and aipowered tools, and so I think, though, you know, what's really central, there is kind of that broader vision of you know, what is the kind of benefit this is meaningfully bringing to society? What problem is

being solved? And are the people who are like most affected by that problem the ones leading the conversation and saying what they need? Because too often I think we see tech developers creating stuff that actually no one really asked for. I'm sure you have like all seen that thing on Amazon when you're like, you're asked for that banana holder or avocado peeler or something. And too often I think tech add on to be a little bit

like that. Whereas I think when we have like really interesting conscious development of say, for example, feminist tools that are designed to encourage good conversations things like that, that really makes me excited about the kinds of futures we could have with technology.

because it reminds me of Crypto. It reminds me of the metaverse, and it reminds you of Zuckerberg being like every worker will have this fucking headset on no no, but nice fucking tri asshole. And I feel like this is kind of like it has a similar tone of like, get ready, folks for this thing. And granted they have some a shiny toy in the form of these large language models that are fun to do, like do and really they're just skimming the internet and just giving it

to you in a nice, tidy sentence. But yeah, like I feel like that is just kind of like I'm seeing that dimension when you see the hype around it, which feels a little more like y'all are talking about this to make money, whereas the other things that you're talking about, like accessibility and trying to democratize certain applications. With things like that, less there's less less scale involved with something like that, So me, yeah, it hasn't talked.

means we focus on the benefits. You know, Oh, if you bring this to every home, everyone's life will be easier, if everyone will have a more seamless home experience, when actually that's not the case. There's real costs to bringing

those tools into a lot of people's homes. Everything from, for example, the way that certain kinds of elder care is getting replaced by technological tools in the UK care system, through to the fact that those tools can really easily be used for the purposes of domestic abuse, into a partner violence, It can be used to control or trap someone in their home. And these are really ugly truths about those technologies, and they're often not the ones that

are put at the heart of technology development. And so that's a lot of my job, I guess, is just to say, how do we put those vulnerabilities in those costs at the forefront, and then judge holistically whether or not we really want this product to exist.

you know what works you recommend. You know, you talk on an episode earlier, I think it was in the summer about how a lot of the ideas and fears and hopes that we have currently are very similar to what we've seen in movies and like those like when you look at Elon Musk's ten favorite books, they're all like fed by this same like Isaac Asimov, like I

robot thing I'd had coot. A lot of Kubrick's ideas are around AI are go back to this very specific verse version of technology where it inevitably turns into a mean psychopath who is set on dominating all of humanity. So as an alternative, because we do like to talk about, you know, the importance of expanding our imagination, Like you know, with regards to climate, a lot of it is like

either it's either apocalypse or business as usual capitalism. There's not like people have a hard time imagining the alternative on AI. On the subject of AI, like, what are is there? First of all, is there any like mainstream kind of popular movie that you feel like actually like

that example of AI like got it right? And then are there other kind of more obscure works that you would send people to in order to sort of feed their imagination of like what what a world with AI technology could look like?

Big Hero six. So the Disney Pixar film where you have an AI powered kind of we're about healthcare buddy called Bamax, And I think that's a really interesting example of an AI that you know, at one point in the film, he gets dressed up in armor and you know, the kid here to starts to try and use him

as a kind of weapon. But like, this is an AI that so has it been designed to be a weapon that kind of resists being weaponized a lot of ways, And in that sense, it's really countercultural to a lot of the AI we see in Western film and cinema, which is often very weaponized. It's often either this like kind of sexy cyborg figure or it's this hyper masculine terminator figure. Yeah, so I find be Max is kind of gender lid and like just.

And the whole thing is these two thousand word essays, so they're really short and really punchy by lots of different guests with had on the podcast, and they all respond to this idea of good technology dot dot dot, So it might be good technology is free, or good technology, you know, challenges power, and the idea is you can just dip in read one. They're also illustrated, and then just dip out when you need a break or a moment.

But yeah, I really would encourage you to pick that up because it just contains the most incredible feminist philosophers, technologists and vendors, activists who are really pushing forward different ideas of the kinds of technological futures we could have.

I mean, apart from that, I read a lot of like Chinese, diasporic Asian American sci fi as well, particularly sci fi that's like thinking a lot about the climate crisis and AI and insects of that, and so I think a lot of those stories have really interesting and different perspectives on what AI is it can be from like Elliott de Badad's work, which explores, you know, the relationships between like humans and sentient mind shifts in this

base Opera universe. Through the like lyrical lized Faltfish Girl, which is like a very dystopian imagining of like a future society based on labor from clones. Like all of these novels, I think just if they're like a lot more love and a lot more could ask, they're just absolutely interesting and imagine have been gorgeous ideas about the future.

it was an unknown number. And this is just exactly me, like my husband like please, fuck you know what I'm like, I will not I will not answer it under a number. And so that brought me a little bit of joy today.

So yeah there shortly. I'll also check Jack and Eye out on our basketball podcast, Miles and jackout man mas pease check me out on my ninety day Fiance podcast for twenty Day Fiance and also the true crime show The Good Thief, which has all episode eight episodes out, So please stream those binge those talking about the Greek robin Hood, the man who legitimately was kidnapping millionaires and doing some old fashioned wealth redistribution in the in a

positive way, never hurting people either, never hurting people, like again, an ethical criminal if they're is such a thing. Some tweets I like, first one is from you know, shout out to the WGA and all the negotiators, because it looks like the WGA strike is they're close to getting

something ratified. So we like that. This tweet is from Jeff Yang at Original Spin, tweeted just underscoring how WGA negotiators told the studios the union wouldn't go back to work until SAG also has a deal, because if the last five months proved anything, it's in together, win together.

And there's this snippet from it might be a deadline, but it says quote the studios also inquired if once a tentative agreement is ratified by the scribes, if the writers would pick up their pens and hit their keyboards again very soon afterwards, the guild, from what we understand, made it clear that they would not be going back to work until SAG AFTRA also had a new agreement with the AMPTP, reflective of the wj WGA's feeling of

solidarity between the two unions that has characterized their first mutual strikes since nineteen sixty love the solidarity, and I know AYATSI is also going to have a contract that is going to need to be renegotiated next year, and it looks like guess what, folks they did. It only listen to one thing, Yeah, putting your putting the tools down. Put the tools down and pick them other tools up, which is getting in your solidarity back. So we'd love

to see that. And another tweet I like is from t Pain at t Pain just put bartender just doing her job. Me just this prudo of Kevin James.

well as a song that we think you might enjoy. Miles, what's the song people might enjoy?