ChatGPT’s Sad Second Birthday
53:01 Share

This is system crash, a weekly technical show willing to tell the hard fruits about how tech and the tightens who owned, Operated IT is really transforming in the world. I'm paris Marks and i'm brian merchant. This week on the show, ChatGPT turns two years old.

OpenAI is looking into ads, you a mask, suing them yet again. And the must open rivalry continues to escalate. Plus, brian has a new report out on A G.

I. And paris. That's me. Coauthored a White paper on digital severity. All that more. But first, brian, how did you celebrate chat V T S big birthday on saturday?

Well, really is an occasion to celebrate is at first we had the whole neighborhood over. You did some barbecue. E sort of printed out our funniest, that ChatGPT stance and just, it's man IT was the best.

I'm so sad.

I missed that. Do you remember when like ChatGPT like first came out, and whenever you went to like any sort of like tech related public event, like everybody had to you like they like the stunt was that they would ask chat pt to like right their intro for them.

oh yes, I remember this really crazy.

yeah. And like you can have them like not along like a very good as IT. Well, I thought I would commemorate this to your milestone by asking ChatGPT.

how do you dare? We're not reading chat V T.

On I how brian and celebrate OK second birthday and ah I think it's very funny ah this says brian merchant, author of the one device and a thoughtful commentator on technology and culture, would likely celebrate ChatGPT second birthday with a mix of reflection and critique. Number one, IT would begin by merging, might start by acknowledging the remarkable strides ChatGPT has made a natural in age processing, especially in terms of its usefulness to abroad audience, and IT knows you so well. I feel like, I feel like either there's been like an intervention here or I have failed.

If this is like literally what I mean, it's funny because we're one of the things we're going to talk about today is like the new sort of drive to put ads into a this is literally just an ad for ChatGPT like baked into the question that I asked IT. So it's also basically just making set up. I have never celebrated a ChatGPT .

makes IT up. What are you talking about?

Is something we take for granted, right? IT has been around for two years. And like the hycy, ation were something that we're gonna get fixed like right away according to all of the like. This was like part of the key promise yeah, still makes stuff up, but it's gonna so useful and so powerful that we still need.

I feel like for most tasks, you know, people who are not skilled with, like getting the systems to sort of work around their own parameters and like really trying to it's still constantly making that up. Even now, IT really isn't improved in a material way for most consumers. I'm looking at ChatGPT four o mini.

This is supposed to be one of its best consumer facing models, and it's just not much Better than IT was. I dip in, I try IT, right? I mean, I I know you know, I know you have a hard, which is good.

good for you. I feel I know the listeners get both sides, right? I just say no and you shot and see what's .

happen and I you have to test I feel like IT as a especially I when .

I see the test that everyone else are doing. And so I feel I don't need to engage in that myself.

Yeah I mean, I you know, for any functionality you like, if i'm actually going to do journalism or do like that is not about ChatGPT specifically or about like N A I model that I never I never touched. I never use that and .

you use google gami stead, right?

Because right? Yeah, I use A I overview. I just cop in pace directly in the my articles yes, that's that's all IT is. But did you even know that I was tragic ties second birthday?

Like, no, no. I do remember that he came out on november thirty when you said IT but yeah wasn't something that was like actively on my writer. R I had other things going on over the weekend that we're not thanksgiving.

I think it's interesting, right, that IT is like I feel like if I had been a smashing success and like things were going really well, I think we would have seen a lot more fan fair and attempts to sort of like boost up the profile.

It's kind of now that we sort of we're pausing on that for a second, it's kind of funny that we're not because I think open an eye and anthropogenic, I think they really want you looking too hard at how well these models are actually doing, how functional they are for businesses that have paid a lot of money for enterprise contracts. I think we're a weird place where like they're still trying to sort of pump up the hypo cycle. They still have like their sort of advocates, people like even molech who are writing their newsletters going right rock.

But it's not really widespread, right? IT hasn't permeated like the public t like this. It's not like google or something that everybody is, is using IT all the time where it's foundation. It's widely agreed upon that IT is like spit, necessary and functional. And we should be building the future of all technological production around at which silicon valley still is, sort of insisting that we do right.

What do you talking about? My A I overviews are completely essential um but no, like to your point though, I feel like you know, if you're really thinking about this though, the story that we've had about chat ch eb t about generate A I recently hasn't been the celebratory story, hasn't been a look at the smashing success we've had and look at how well everything is going.

It's like preferred sort of like how these companies are losing a lot of money OpenAI in particular. I feel like that is a story from like last month of the month before you know that one has waned a little bit. But like the story that has been happening over the past few weeks is the story of how these companies are really struggling to make the models any Better because they assume that as they kept making the models bigger and bigger and bigger, they would get Better and Better and Better.

And, you know, continue to fool us and even Better and know more useful ways to make us feel that these chat bots are getting closer to the agi and understanding things like a human and being intelligent and all like bushes, that we further the companies talking about for so long. But actually, the story has been because of research and a number of things that were published over the past number of weeks that actually these companies are struggling with this. Their models are not getting Better and they're being pushed at this moment to show that they can actually make some money off of these things because the big promises are really not coming true in the way, as you are suggesting, that we were LED to believe in those early months after november of twenty twenty two.

And because these models are so resource in labor, in computer intensive to run like this isn't just another piece of software. They're so consumptive on all three of those fronts that they demand more investment. So they demand more by and they demand more story being told about them to maintain this positive and sort of a Loring light where they yet, this is something that we need.

This is, this is the future. They have to keep beating us over the head with IT as best they can. I occasionally .

see the story that, like the AI agents are just around the corner, OpenAI is going to start rolling the modes like give IT up. Like what are you doing the P. R work for this company? Like this is not what this technology is.

You, yes, IT was sold to us that way. Yes, IT was going to be your doctor in your teacher and all this kind of stuff. That same old saying in the beginning, bit like that is not the story of these companies right now. And if that is like the narrative that is being pushed, then really like not paying attention to what is really going on here.

Yeah and I mean, it's you're right. It's just that it's a story and they have to keep sort of changing the condors and like the specific sort of subject and arc of that story a little bit mean, it's worth noting here. I was going to a wait a little bit to dig into the but I have this report that i've been working on for a while .

for the AI now inside you for a while.

you for a while.

Brian was on the other four gas. I won't save us like what a few weeks ago now almost a months ago were like the reports just of I to come out and I had to like record a little piece of me like ah it's not going to be .

up by the time this interview goes like the nonprofit and the academic word move a little slower than our our podcasting worlds and but yeah I will be out soon.

We move fast and break things speed you know they .

are the old is ah yeah but the justice is that I went back and I looked at sort of the founding of OpenAI and the context into which IT was founded. And there is a number of interesting things about its genesis. But I think the one that I pulled out is this healing to the narrative of agi and artificial general intelligence that from the beginning was central to open a mission even when I was a non profit.

And that's how IT was pulling researchers away from google. That's how IT was sort of getting sort of the tactic interest of investors who were in the room when I was founded, even as a non profit, even though IT wasn't explicit for a profit, but IT was, I attempted to show very much, obviously coded as a potential profit generator down the line. And IT relied on sort of this engine, this narrative engine of agi sort of being a step up from regular AI, which, you know, a has been around forever fifty years ago. A I just meant I the way that they're using agi now, but they have to kind of use new marketing terms. And its worth noting that A I was born as a marketing term by academics who are just using the coinage to try to get some funding for a summer program, and that's how I was born, like this very explicit sort of coinage to use.

I I do love that part of the story right where the term actually comes from, seeing like way back when that they were still like playing this game with us just in the research network at that point, rather than .

out broader, going out to VC and trying to get, you know, billions of dollars. He was just like, well, we Better make these things sound good, so people give us money. And IT wasn't anything that they were shy about at the time.

H I, I think that was mccarthy, who was one of the sort of early A I. researchers. And he just like, would say that in a vies like, no, I go in the term, I did IT to get some money. That was, there was really, we just wanted to pursue to the research. But anyway.

there was so much more honest with us back in little days. yes.

So the report talks about how, you know, we built, or they built this narrative around agi in particular, that you know, it's going to be this center or ci center and all powerful. And notably, per open a eyes explicit coin age like IT can do the economic task of any human like. That's what they consider agi on paper.

And so when they drop ChatGPT two years ago, it's gone through a lot of evolution, but it's alty highly prime to be a commercial endeavor. But they also know that they don't really need a business model or anything like that when they launch because the last ten years, up to that point of silicon valley with uber, you know, sustaining losses for a decade before I ever even sort of gets close. So making any money and the zero percent interest rate period where cash was just slashing into into startups, IT had become clear that you don't need a business model to serve, generate interest, to generate investment.

You need a story. And A G. I is the ultimate story for a silicon valley.

It's going to create an agent or a technology that can replace any human worker. And what business doesn't want that they all do. So this is the story into this day. I'd say it's kind of wind a little the my wage starts to I get a little bit dubs after you've been saying IT for two years and IT doesn't show up. But I have also started in the report at some like key points where like the fortunes, or maybe sagging a little bit.

don't lose faith, brian. Faith guy, oh, we are. You know.

j GPT has shown Sparks of A G I. That's something like researchers literally said at like an opportune time where IT could help sort of get up interest, press attention or investment. And that's kind of where we are now. You know, I roll over the summer. I was arguing that sort of the AI boom had already kind of peaked.

IT hadn't burst, but that starting to look pretty accurate to me like I feel like we have come down a few levels to where you know businesses are, you know kind of openly saying, I don't know this is really all that accurate. And business insider had a good story last month or in october, maybe it's just last, but business insider had a great story about copilot, which is running on open as technology. And microsoft insiders are saying, like the this really can't do seventy five percent of the stuff that we're promising, not looking so good from the inside. They were the .

ability building IT into like absolutely everything we can think.

That's what so crazy about all this. And that's where you, I think, really begin to understand the power of this A G I construct that literally everyone in silicon valley has just kind of, you know, bought, okay, not literally, but almost everybody in silicon valley has bought into and is propelling forward. It's an apple laptops.

It's in apple intelligence. It's in you know, microsoft productivity software. It's on your phone, it's on your google. So it's everywhere.

And it's just remarkable that it's not passing any like sort of quality assurance test by or even once handed the consumers. There is so much backlash when we google rolled this out on the most public stage imaginable. It's not like people said, yes, we like that.

It's just like, well, I guess we have to deal with this now. And now it's everywhere. It's not even certain that it's going to make money.

And our live story today is how because of all this, because the enterprise sales maybe aren't what they wanted to be and because it's so expensive to run all these systems in the frontier models, it's putting opening eye specifically in the whole right now, like five billion dollars a year, and they have to plug that gap. So what do they do? What do solon valley do when they have a whole that .

big jero advertizing .

advertising world mats on there? I love IT.

They need to collect more data on us, target more ads at us and have more servers in order to power all the attack it's I love in it's a wild to like think back to the early days of the internet and the early days of you go particular, right? And I feel like there's like different past that we could have taken with like know how the internet works, how the internet was commercialized by, I think in part because IT comes out of the united states in particular and because the united states has the experience of like for relying on advertising to fund media and all this in the stuff that like this was the real driving force of what the internet was going to be and how IT was commercialized.

And that like when everywhere else, but we have known from the very beginning that this was going to like corrupt aspects of what happened online, going back to that early paper that sergey brin and Larry page wrote when they were still students at stanford working on the google search engine as an academic project and not as a major company, where they admitted that adding advertising to this thing was going to know kind of corrupt what they had built, right? Because they were going to end up having to serve to push in the desire of advertising, the need for profits, all these sorts of things. And then, of course, when the company was spun out, when the company was made private, when you had eric, mitt come in to start running IT like a proper business and they we're gona make money, of course, ads were introduced and of course, IT was corrupted.

And you we have not just seen that very clearly with google, but with so much of the rest of the internet that has become so dependent on digital advertising that has required building out this vast surveilLance apparatus, corporate private surveilLance apparatus, that needs to collect all this data on us in order to power. These things like this is all connected in the direction that we've gone, is so tied up in the business model that is at the base of this. And so then to see OpenAI now pursuing this too, now saying that, okay, we're losing all this money.

Our chatbot that was going to change the world is not actually changing the world and is just another tech product that you has a lot of influence for our microsoft into IT. But we are just trying to become a big tech company of our own to the degree that we're throwing off the non profit status bit by bit. Um you know trying to make sure we become a four profit company and now reaching the state where, as you're saying, we're talking about you know the second birthday of ChatGPT.

OpenAI is not really talking about that very much. But right is that happens as we get to that milestone. We're starting to get these stories that oh yeah, we're talking about bringing ads into chat y now and into our general eye product so that we can actually make any money like that doesn't seem like a celebratory. We one kind of moment IT seems like, oh, you know, we're compromising on this.

Yeah IT feels like IT feels like a defeat .

kind of I wouldn't call the defeat though because like they do have power, they do have this huge valuation. They are throwing off the nonprofit status and they're about to bring in this business model that they hope is going to, you know, start pouring money into their coffers in a way that we've seen from facebook and google in these other companies. Whether that works out, I think, is another question.

Whether they have enough user base to actually produce those kind of profit, whether they can make an add model work, whether the high cost of what they're doing can actually be supported by in a bottle. I think those are open questions that i'm sure we'll see more reporting on and will be more. But that is, I guess, how I see IT.

Yeah I don't think it's yeah defeat is the wrong term because as you said, like it's built into the DNA from the beginning of every sort of you know online best to a company that they're selling often are, and it's pretty used for a lot of people. So IT, IT doesn't take a genius to guess that it's going to advertise. I do think that they were hoping that you know the enterprise sales would go Better, that more people would be paying for this, that would find IT useful enough to pay for IT. But the APP store that like literally nobody talks about the GPT store that like I cannot remember the last time I saw an organic mention of that in the wild, nobody ever mentions that they're going to make .

their own APP store. We're not onna have any anti trust lawsuit over the gpp today.

Ah I well google might argue that it's I like the other GPT still .

we're not a yeah right where not a lot .

but but it's funny. They are speed running like the the live cycle of every tech giant out there. And IT is so much dinner you know you have folks like ez itron out there who've been like pretty boliden on the fact that it's going to collapse, might not even make IT to an IPO.

And this is why you know IT doesn't have sort of monopolization ation that google has to rely and IT can't just extract rents. So opening I really risks, I feel like losing its users if IT switches to ads and there are you know, its service is now fairly and distinguish for most people to like three or four other services. And I don't think that IT has that degree of loyalty where if ads really you know fuck up the experience that, that people might just bail.

And I think you know it's really interesting because like OpenAI may be like the sort of flimsiest company to get one hundred fifty billion dollars worth evaluation. And IT is, you know, as folks like you as a drug, are you at risk of kind of drawing up, especially now that it's like facing sort of a for fall on attack from elon mask and yet another lawsuit? And i'm sure what we will want to talk about that because now that we have this new sort of suda wallets, cac tech and state sort of simbo is emerging in the trump mosque administration that's forthcoming.

There's going to be, you know, power wheels ded in a new way in the taxi and elan must kit, same old minutes. It's well known and it'll be, you know open question to what extent he decides to punish alt men for hit the slides as he see IT. I going back to the point where mask wanted to be CEO OpenAI years ago and alt men sort of. One, this tiny non profit companies allegiance and sort of they voted for all men over mosque. I think that's where all disagree event really stems from ultimately.

Yeah, it's so funny to think about that too because then you know so much of the company has turned on altman as well as he has like pursue this commercialization part like all those A I safety people as they call themselves, like have totally been of pushed out of the ordinary date or or left under their own will because you know they are attempt at pushing him out, fail, do all the beevers exactly right? Who, you know believe in this agi stuff and and all that.

But I I wanted to have one point to what you are saying. And then we can pick up on the most stuff, which is just that when we think about that revenue model, the one piece of IT as well, you mention the enterprise descriptions. The other aspect of that is that if chat B T is not working out, if you're not seeing the kind of benefits of the results that open a sold to these companies as what this product was going to be, I think inevitably you're going to start seeing, you know, the cancellation of those enterprise policies. Are the revenue from that enterprise stream decrease? And so that naturally makes an add model necessary, right, in order to try to bring in the revenue in order to try to fund this thing.

Yeah, I mean and that is where they're going. I mean, in its altman has been saying that theyd like desperately need to open up more data centers. They need to get more power to keep training the models to make them Better.

So they have all these other sun costs still. And you do have some enterprise customers, sort of there's been a few public ones, you know, I think they're probably at the phase where they can still sell enough new ones. And I think that sounds like reading between the lines that their a business, which I think they were counting on being more selling, you know, this technology to other companies so they can build their own bots has been disappointing.

Some of the comments from Sarah friar, who is there sort of leading this drive to sort of get get into more profitable roms are really they're really pushing. They hired somebody from instagram to do ads. And I think IT is fitting.

I mean, everybody knows the word in shatila y now like there's a reason in citified is like the word of the year from from like marum webster isn't really they declared that the word of the year in certification because it's something that at least describes is something that everybody recognizes. There's more ads. Digital systems don't work as well as they used to. And ChatGPT is only been around for for two years. And now it's basically going to initiate fy itself like right off the bed even if if you such a things possible even you know, you could argue .

it's IT started in citified. How can I get worse?

You know, IT is it's again, it's really, really wild. Look back two years of ChatGPT. What's the biggest impact that it's made? I think you know if you're being generous, you can say some codes use IT to like generate route code.

I'm sure like email marketing, people use IT to write their their emails, which is not something most of us wants to see more of in the world. But the biggest impact remains is used in like classrooms and teachers are going to deal with plagiarized homework assignments. And it's being used as like as an automatic homework generator.

I think that still to this day, it's most notorious use case. Again, you have these like fringe business things and maybe some customer service applications where IT was infused on to already extent, chatbot, like customer service chat pots were around away before ChatGPT. So maybe that is like being used specifically. But but IT hadn't transformative in that sector by any stretch of the imagination.

It's just been in addition. And they're still terrible even with ChatGPT .

and they're still terrible. So to sort of wrap up our whole section on chat G P S birthday, you can see why they don't really want to have a big triumphant, you know, uh, a stabbed having a celebration, why they'd rather just kind of put their heads down and you keep making these promises that I will emerge eventually. Just trust them, let them build more data centers.

And I think that is a good segway to talking about elan mus and sam altman and because you know some altman has wanted to build more data centers and get more even they want more nuclear power for this step. The three small and nuclear reactor famously was uh commissioned by microsoft in the service of more a eye stuff. But there's all these like points opening up now where as we as we just mentioned, like elon mosque has much more power and his cronies have much more power over sort of making any sort of determinations about sort of state level matters when IT comes to technological development. And you know, I think elan mosques already get pretty comfortable on .

the pedestal .

first body first to he is not gonna shy about wielding power again, copy out as long as this weird relationship lasts. But so the news over that, that just recently broke, I think, over the weekend or on monday was that moss has filed his fourth lawsuit against OpenAI. Again, the primary point of contention is that it's departing from the non profit status that IT was found IT as elon mosque did give the non profit company money and was a cofounder. And sam altman sort of has eclipsed him and a long ago and and he has grievances.

Ces, with that, but he's expanded the last suit now, I think telling ly to include not just OpenAI people, but also read half men who was also, I think with the first thing I noticed with with that name being included in the new lawsuit was that he is a vocal opponent of trump. So elon mask is like bringing in to sort of his project to punish, open an eye. A very distinct political flavor, I think, from where i'm sitting. And I think it's just hearkening the beginning of this era where he's wheeling power against political and incorporate rivals and and all man .

is someone who has positioned himself as more democrat, if I remember correctly.

Yeah, all man is a registered democrats you know, he came out the gates super abseil. We used to trump a lot of the other silicon valley CEO, but he's a register democrat. He didn't endorse trump.

He was very quiet, very neutral. And musk would love nothing more than to see OpenAI fall apart. And then, you know have clock or his X I company sort of pick up the slack.

It's all, see. So take you is so stupid, but it's also reality. Where where is such a thing could happen?

I can't wait for the future where rock is our A G I. overlord. A making terrible jokes as IT tries to replica most sense of humor because the codes, have you kind of forced .

that in there, have been forced train IT. Unlike elan musk tweet, like the a dear leader, A I but like being used to try to replace all the government workers that musk wants .

to get rid of with his dog agency. I going .

to be much different really. And the other notable name that lept out to me in the new law, IT, is the new plaintiff, which is shavon zilli.

right? Yeah, part of his know of the called down in Austin where he's bought these merchants to try to get all of his kids and baby boomers to live together.

Yet she's the mother of at least two of his kids or one .

of his kids .

i've lost count who can .

keep is a new one in every other week.

IT seems yes, but IT is just like to me it's a reflective of everything we talk about last time about this sort of like emerging nexus of object, sort of state power getting grafted onto the you know more right wing elements in silicon valley who are just willing to sort of step into this new frontier.

And in just shows how subtitle, I think this model of technological development has been to this very all along right? Like musk is able to buy one of the biggest communication networks that exist on the entire planet, and he was just able to pump IT full of conservative propaganda. IT helped get president elected.

Now he just owned you could do IT ever he wants with IT. It's just the worst cesspool. It's worse than it's ever been, and it's going to continue to be that way.

Then there's nothing we can do about IT. He owns IT like our the political avenue would be to try to impose some kind of standards or regulation on IT. That's a non starter because we have this style conservative governments.

So we have this situation where all of our digital infrastructure is owned by a handful of tech monopoly and tech titans. And you know, again, we've we argued last week that blue sky is like a respite from all that because IT reacts a democratic input in a way that no other platform does. But all this is to say, I think this really feeds into the subject matter of your paper that you are just right.

Do you want to talk about that? And I the thing I agree with very much all of that is i'm just a bit more skeptical maybe and I think you are two of this narrative that twitter actually play to see significant role in the election and kind of shifting IT pheromone d trump. I think IT was just beneficial for us to frame things that way. And yes, IT has become this hot bet of conservatism and you know, even more kind of extreme right wing views but I would be skeptic to the role that had actually played in getting trump l elected at end of the day.

I don't you know, I go back and forth, I go back and forth on IT just because IT is such a useful place for for these narratives to be hammer out and then sort of globed onto I can see that .

part of IT yeah.

it's not that it's like reaching more people and they going to go I saw on twitter and i'm .

going to go vote. It's like it's the place where the narrative around the hatin migrants eating cat dogs can be formed and then pushed out across the White right media network that's right.

And those things were elevated and sort of you boosted uh because they had here to mosque some viewpoint and and the algorithm clearly favors a lot of that stuff. They're not let anybody really see what what's actually going on under the hood. I think most people would agree that it's just like the stuff that's being surfaced is just more violent rhetoric.

C, more massage, generous stuff, more masculine coded stuff, more span, more blue check. And so people IT became this intensely sort of sort of moscow space. And I I think that I think I did have an effect, and I think i'd like, is that I go back and forth, but to sort of.

Just say like, well, I was just another platform and platforms only have limited a sort of ability to sort of reach voters and take like catalist, you know any sort of action when where the other I don't know. I feel like that also helped us kind of underestimate the extent to which x had been transformed into this place that was very friendly to people who have overtly national alist and racist tendencies. And they could find each other, they could phone messages, they could just blast shit out there. So yeah, I don't know. I'm pretty in a bivalent .

that that makes sense though I I feel like i'm just always hesitant of the narrative that is this tech platform that made this real significant material difference because I feel like too often I see like centrist parties that are losing relevance and losing support say, oh, that was the tech platform that did this rather than eto interrogating the problems and the mistakes with their own campaign s and their politics of y people are turning away from them. And I feel like we ve seen that time and again where the influence of tech platforms, whether it's facebook or you know, now we see a lot of this directed around tiktok. I feel like I often that is exaggerated because it's beneficial for certain interest, too exaggerated because IT kind of shifts the lens on our focus from maybe their own mistakes are are the things that they did wrong to be able to say, oh, you know, the tech platforms are causing these negative harms in society, which I agree with like I think that there are, you know a lot of harms that comes with these platforms that me know people will be very familiar with if they are familiar with either of our work, right? But yeah, I feel like i'm just a bit hesitant allowing that narrative to take hold.

no. And I I think that stems from a lot of due criticism from the last time trump was elected in a lot of people wanted to blame facebook for spreading misinformation. I would argue that it's different this time. Instead of being at a scape goat, I think you can look at x and say, IT is a channel for a very sort of well define sort of mode of politics.

And I think the problem that you know that was, and I I agree with the election post morum a lot that sort the democrats intend to become so corporate zed and so wide that there was barely any politics to speak up, is just like we're not the other guy. So you can't really inject any politics into any sort of platform, digital or otherwise, if your message is ultimately so vacuous, if you're not actually standing for anything. But that said, conversely, you see mosque on air tweet, you know, nasty memes about immigrants and like that does fire people up.

It's horrible. But IT does fire up a number of people that goes viral. People can interact with IT. And I do think that that can sort of catalyze its own form of politics, because you can say that these things don't matter at all. You know, we've spent so much of our time online, we are being influenced by this stuff.

I would love to see more study about the more specific ways that messages sort of were transmitted from candidates or news events into the platforms, back out to the elector. And i'm sure will get some good work on that stuff. But all that is to say that was like it's a huge communication platform and IT was colonized by a very specific set of ideas.

And now it's going to continue to be used for that as long as musk is at the home, are still interested in IT in this way and IT has an audience, I think it's interesting. It's still shein audience, right, like the audience is flying. So IT could be that this was the point where I had any kind of utility at this scale. I feel like when I go on twitter now, IT is just a shell of what I used to be even a year ago.

Yeah no, you. I think we agree much more than we disagree on this, even though I might sound otherwise. But if but I think as we've talked this out, like I think it's pretty clear that were largely on the same page.

Yeah no for sure. That is interesting to think about. Don't think like anybody again, like they used to make the firehose sort of available to academics that you could study what was actually happening.

Twitter researchers and I could then make just love twitter because you got so much and maybe to a fault where you could know yeah certain and lies up to IT. But IT was much more open, impossible. And now now it's a more us and we can sort of extrapolate as though what's going on from what we see.

Well, I I think that also kind of fees into part of what this reporter, this White paper that I wrote with a number of colleagues of mine.

yeah.

tell us about your report. Yeah, I coauthor this White paper called recalled iming digital severely a with systemic k APP src don po boto and emilion. You know who are that american and european and academics at universities in europe? And this really came out of people might remember a few months ago, twitter or x got banned in brazil because IT wouldn't suspend these accounts who were engaged in pushing like far right politics and the attempt cut down there, and ella mosques, like refusing to comply with these supreme court orders.

And so the platform itself was like just completely blocked in the country for a while, right? And so initially we had like a shorter letter that we put together that was basically just supporting brazil in defending its sovereign against someone like you on moss, you know, this american billionaire american tech company that would not abide by the rules of this democratic country and this democratic government, and was saying, like, no, we won't follow the things that you expect for us and to us, that was completely unacceptable, right? So then we kind of built that into this broader report by bringing together a lot of the people who initially signed that initial open letter, asking for their feedback as to the types of things that should be in this kind of broader policy agenda for what digital sovereignty should look like.

And then the five of us kind of brought a red all together and put IT into something a bit more coherent, right, rather than just having a bunch of little points. And so, you know, listeners of the show will probably know i'm not an american, i'm a canadian. So I often commit these issues from a slightly different perspective.

Maybe then, you know, is often the case in the united states because i'm not just watching what's happening in the united states, but i'm seeing how these american teacher companies are often affecting my country, and how you know comparatively little power that we have to actually force them to follow the rules here or to make sure that they abide by kind of our social norms and our values, and how the types of things that used to be common in other forms of media. You know, before the internet came along, if you think about public broadcasting or or certain regulations that we had, unlike the types of content that could be produced or the type of expectations we had around local context production and all these things are always thought to the nail when IT comes to the internet. Um and so what we are really saying with this White paper is to say, listen, what we see right now is the united states really trying to push out this notion that you're either with us or against us, right?

The united states in silicon balin particular feel that they are technologically threatened by china and IT feels like we are increasingly being pulled in two directions. Either you're in the chinese pole or you're in the american pole, right? You're allied with one, but you can't work with both. And so the americans, this is completely unacceptable.

And so at the same time, silicon valley is aligning itself with these really far right government or are increasingly far governments are pushing out these really rightwing politics that are very apparent to a lot of us, you know, not just in the united states, but beyond as well. We are often outside the united tes. There's far less support for this kind of a politics, and there's a lot more kind of public anger influx ration, I think it's fair to say with a lot of these tech companies and what they're doing.

And so we are trying to say is what needs to happen is these countries outside of china in the united states need to start working together on a common agenda to think about a different way of approaching digital technology and the politics around digital technology to make sure that those things actually serve us right and are not just serving the bottom lines of these major tech companies. And that requires not just being critical of the big tech companies, but actually making sure that our governments are working for us right, not to reject this notion that the government should have nothing to do with the internet, but to say, listen, the only kind of source of power that we really have that can chAllenge these major tech companies, often coming from the united states, is our governments. So what we need to do is make sure that our governments are making the investments to develop different kinds of technologies to make sure that we control the digital infrastructure, you know, within our boundaries, within our borders and to make sure that the platforms were creating are aligned with the norms and the expectations that we have within our countries and are not just trying to like, become the next unicorn or whatever.

And so basically, what we do with this report is put out a whole set of, you know, policy proposals that are aligned with this to make sure technology service people in the planet, not major american tech companies, and that we actually have some power over our digital destinies in a way that IT feels like we don't right now, and that silicon bali and the trump administration want to make sure that we definitely don't have into the future. And so know when you talk about in the power that he wheels in this administration and the way that he can wheel IT against same outcome, but also like the broader vision that these people have, it's about saying that vision is not one that a lot of american support, but it's also not one that a lot of people outside the united states support. And we should have the power to say, we do not want this. We have a different vision for what technology should do for us in our societies, in our world. And our government should be working together to realize that instead of just cowering to the trauma administration, to elon mosque and do with the united states once of us yeah.

it's really, really important project. I mean, all around I think one of the major projects I think that we all have going forward or finding ways to uh, sort of blunt this. increasing. I mean, it's not clear now if you can see how top down and technological development is in the way that IT IT is built and then deployed around the world, then I I don't think you ever will, right?

I think this alliance of musk in trump just makes IT so clear that we have this very anti democratic mode of technological development that really is built to serve some very narrow interests into fund profits upstream. So yeah, I really applied this effort to try, you know, form coalitions to sort of to fight back with with government because you can I think we forget that there are means of Operating these companies in. And with the united front, you can get a lot done.

I mean, there's there's a reason that it's been canada in australia that have been able to sort of push back on google from canada, zing the news and forcing them to pay up. Yes, sure. The solutions are deeply imperfect.

Any democratic solution is. But instead of saying, oh, well, I guess we're going to let big tech like you devour our journalism industry, here's an idea that is aligned with the civic good that is in the interest of the public. And we're going to, you know, find a policy solution. Imagine that a policy solution.

Let's do that. So wild. Yeah, so wild. I think you're completely right, right.

It's about saying that brazil shouldn't be doing this on its own. Canada shouldn't be doing this on its own. Australia shouldn't be doing this on its own. Why can't we all come together and see what we have in common and try to work to some common goal here? I feel like that's another thing that we forget that we can take control IT of over technology and make sure that I actually serves us and the people in our societies and the planet as a whole and is not, as we're seeing with generated by eye in these data centres, increasingly like know causing emissions to soar and depositing our environments and staff, but that we can also come together and work on things together, right?

I think part of this report is also trying to bring back this notion that, like through the united nations, which is a really imperfect institution, countries used to work together a lot more to create these kind of systems that were beneficial to a lot of different people in the planet, instead of just having a dominic by the major powers that exists right now, right? If you think about the universal postal union or the international telemachus union, you know, all these countries came together to make sure that these international systems worked, and that IT wasn't just the united states setting the rules, but that we were all kind of collectively working on these things together, or even thinking about the nona line movement that existed during the cold war, where you had all these states who were like, we're not going in the soviet block, are not going in the american block. We are going to work together on these things on our own and kind of say we are not, you know, engaging in this.

I think that is some of the inspirations we should be looking to, to say we don't need to accept domination by silicon valley or domination by shen zen. We can do something different and we can think about technology in a very different way that has very different goals with IT. And I think that especially in this moment, seeing what Donald trump is doing, saying how silicon valley has completely blown onto him and made their project this project, I think we need that more than ever. And so that's kind of what we're trying to do.

Yeah I think it's great. Um and I just have one question and that's why do you hate america?

Well, that's a long answer. Everybody should .

should check out this paper. When is IT dropping? It's out now.

People can go check IT out um you know they can go to my blog, disconnect the blog and I have a post up there where you can find the link. I I would just have one more thing to what you're saying. They are right.

Like I feel like thinking from the united states, we don't actually often recognize the degree to which the united states really pushes silicon valley internationally, right? I mean, canada. And so we deal with the united states a lot.

You might soon be our fifty first state yeah if .

doll truckee this way yeah yeah but it's like we passed the news code thing and you know the american government was actually like kind of mad about that, but didn't go too far with IT. Then we pass this digital services tax. So this is trying to make these multination often us.

Tech companies actually pay a fair or amount of tax up here because of the know kind of schemes that they use to move their tax around and reduce their tax burdens. In the united states, the american diplomats, the american embassy of threatened retaliation against canada, know through the international courts as a result of that, even though many other countries have already done this right. But because we are right on the door step, because we are very visible, because so many the companies are very dominant empowerment up here.

It's like you can't properly regulate these tech companies. Even when we renegotiated that new nfa last time under trump, there were explicit clauses put in there to make sure that we were limited in the types of tech regulation that we could pass because, you know, the tech companies got in there or lobby is got in there and mature, that the united states included those sorts of things. These are the types of things that really make me angry, that and really feel like our sovereign, our ability to pass rules in our own jurisdictions that benefits canadians, is really being restricted by not just the american ted companies, but also what the american government is doing. And so like outside the united states, those things are much more like in our focus in in the discussions that we're having that I feel like often get recognized in the united states itself and the policy discussions .

that happen there. If you don't like our internet, you can build your own.

Maybe we will no cana to try to do that. Like before the internet, there was this whole, I think IT was called .

talent on much more polite internet. I could use that. I could use a polite canadian internet.

Everyone who will be wanting to join, yeah and and then all the american right wingers will .

get mad and say that we're .

restricting .

free speech because no exit exit well before our system fully crashes. Today we have a few final headlines to keep you all updated about the state of our collective this topic here and a abroad you want to kick software. Paris.

sure thing. Ah, so read alberg. I I I hope i'm pronouncing that right in saf has reported that there's a new lawsuit from employees that IT uses apple of spying on them via personal eyes, cloud accounts and non work devices.

I'm sure this is something we will definitely be paying attention to. But think of all those narratives about apple in respecting all our privacy. IT doesn't seem like they're doing that to their employees.

Yeah, apple has really enjoyed like a reputational boost from sort of touting its privacy stuff and this is no an interesting little dent in its armor there or IT should be and yet another sort of example of of potential workplace wrong doing. So we'll definitely be watching to see how that place out.

We also have a couple new reports about google project nimbus as contract with israel that has been so much in the focus over the past year. So the new york times reported that google had concerns about potential human rights violations connected to project in bus at the one point two billion hour cloud contract before the deal was signed with zero. In one related to that, the intercept has also reported that the contract is subject to an adjust the terms of service, not the use visual google terms of service, and that google didn't actually take recommendations. IT thought from the business for social responsibility, a firm that I went to as I was negotiating the sort of thing to make sure that IT wasn't going to be in violation of the types of things that IT wants to avoid.

And then we have our friend mark zuker berg, who perhaps seeing the success that even when musk s enjoyed by becoming uh overtly political in favor of trump, he had dinner with the president elect at moral logo and a new ft. Report suggests that zucker seek some more active role in tech policy with the administration hoping to join this new tech. All the darky that is brewing when trump comes to take office.

Well, you know, we both know that he thinks Donald trump is a big bad as he is so bad now.

I mean, this is the way to get in Donald trumps good Graces. You say something that stands out a little big, call him a bad as you go to moral logo. All the tech cees are trying to kiss the ring.

But I think zuker burg is is doing so most successfully, perhaps well after he obviously has his own sort of past with with elon mosque. So it'll be interesting to see whether that that plays a role in his pathway to power there. absolutely.

I about to close on outside with our final story. Brine.

yeah. And sometimes you just find a story that just Crystalized everything about your emergent dsp. pia. So a potent tly that you can't leave IT off. And we have, in this perfect such example, jail bots are on the way.

There's a prison in georgia that is demoting AI robots that use facial recognition to monitor inmates to ensure they are ware they are supposed to be at all times, according to npr, which has a peace, uh, in which the reporter goes down and watches the robots. And this is just a trial thing, but you've got a few robots literally patrolling the prison. It's yelling at the inmates to stand back. It's conducting around the clock surveilLance and sending data back to the tech company that that is leasing these things that has not been approved yet. It's like a trial thing, but we have literal, A, I infused panopticon in motion here.

Yeah, I thought the canopy con itself was bad enough, but the A I version, of course, is gonna take a to a whole new level.

I think if we have any solis here is that it's almost certainly going to be shipper. Then the feared panoptic on which, you know, of course, was everybody looking out at each other, policing each other. Now we have a tech company just beaming ing data and after its IT looks like r 2， b two or something.

Apparently this is just the kind of thing that's only possible if you have a imani, right, like you have contracts like this. Obviously, surveilLance and technologies are no stranger to the prison industrial complex, but you can only imagine, like the pitch meeting here, where you know some the morning his heard day I was and some. Grifter A, I started up and capitalize by helping up a jail bot here.

yeah. How else can they profit off of the private prison industry? And all these people that know the states keeps locked up.

right? The per minute fees that they charge when inmates call their loved ones were not enough. Now we have a actual jail bot running the grounds. It's scrim.

Well, this is, you know I guess, another insight into our textiles ohia. I'm i'm going to have to organize a ChatGPT belated a birthday party for me, my friends, this weekend because I don't know how we missed IT.

Ah yeah maybe we can celebrate the next birthday after there's an uprising and the jail bottles thrown over the side and .

spread that you going to say the ais, we're going to have their uprising and so we would just be forced into having mand birthday .

body for ah we else in around clapping, reading our very own print out of ChatGPT erra.

你 爱 辣。 了 well that's another episode from us system crash is hosted with by paris Marks and brian merchant production is by area wick of our theme music is destroy a via good friends in yet about the show on Peter on outcomes slash system crash pod to get a monthly bonus episode。 Thanks so much for listening.

I.