From Kaleidoscope and iHeart podcasts. This is tech stuff.
I'm as Voloshian and I'm care Price.
Today we're going to get into the thing no One asked for a social media app dedicated to AI video, and then a study asking, somewhat surprisingly whether social media is actually on the decline, Then some advice for parents about how to raise your kids in the age of AI. All of that on the Weekend Tech. It's Friday, October tenth.
Hello Cara, Hi Azie, how are you good.
I thought we'd start with Show and Tell today, my favorite. So here we go.
Show me he's taking out YouTube guys. It's loading and I'm seeing Andrew Ku.
I'm Andrew Cuomo and I could pertain to do a lot of jobs. But I know what I know, and I know what I don't know, and I do know how to make government work. There are a lot of jobs I can't do, but I'm ready to be your mayor on day one.
There are a lot of things I'm not good at, including being a mayor, being a mayor, and using AI to make campaign ads unbelievable. Just for the benefit of our listeners, the premise here is that Andrew Cuomo can't do a lot of jobs. His AI version is looking like a struggling subway conductor, stressed out as a trader on the floor of the New York Stock Exchange, and a window washer. However, there are two Empire State buildings behind him, which that was quite missing.
I really love that video because one, it really makes him look like a boomer.
It really really really does.
And obviously the person that he's going up against has swept sillennials in drove Mamdani.
I know it's interesting. In the YouTube comments, somebody actually commented boomers love their AI slop, which I thought was funny. There are some other great YouTube comments I like, like bro is getting funded by billionaires, pecant afford actions with three skull emojic doll emojis. Mamdani also dinged Quomo obviously, I mean when you when you present someone with an open goal, they tend to strike the ball into it.
He made fun of Quomo Becausemamber earlier in the campaign got called out for using AIS generate his housing plan. Yes that was back in April.
Yes, he uses AI for his campaign videos and for the housing exactly.
But Mamdani also said whether it was very clever in a city of world class artists and production crews hunting for their next gig. Andrew Cuomo made a TV ad the same way he wrote his housing policy with AI.
That was a sweet dig. That was a sweet dig.
And you know you mentioned boomers, and with all due respect to our you know boomers out there, it is interesting how boomerish this use of AI is.
I'm sure that Cuomo jumped at the opportunity to put AI in a campaign, of.
Course, as it feels fresh, and that's what I'm going to do. We're gonna use AI. Do you remember the original Mamdani campaign from the Democratic primary? Yeah, of course,
it was very visually iconic. Extremely Scott Galloway made this point in his newsletter, which is that very surprisingly, despite all the headlines about how AI is going to run designers out of business, the ratio of designers to engineers in Silicon Valley firms is actually increasing because the deluge of AI slop is putting a new premium on really really good human designers, right, and.
So design has to matter more like matter.
You know the old axiom new tools don't always equal better results, right.
Well, you know, speaking of AI generated videos. I want to talk a little bit about something that I can't stop seeing online, and that is Sora to videos, Sora Too videos.
I'm glad these are from open Ai, right, yes, you know, I want to hear what you have telling me about Sora. But before we get there, I do want to just talk about what a big week it's been for open Ai. There's also a multi hundred million dollar chip deal and the release of something that people are calling the everything app. So we're going to get to those other two big open Ai headlines, but first time me about Sora.
So had you you've heard of Sora before I had. It refers to a few different things. One of them is what Sora one was, which was open AI's text to video generator that was the original Sora. Last week, Sra gets an update. I don't know why I want to call her. Her name is Sora Too, and Sora to is what powers this new app called Sora, which is soa up.
Yeah, very good, very very good. I where as in the app charts it was like top three, I actually know the answer one, number.
One one in the total app.
Lap charts, even though it's in beta or beta.
As Beta beta orour Yeah.
I think one thing worth mentioning here is I think this is the first time open ai has ever released an app other than chatchipt. That's correct, Which if you're a big company, that must mean you're treating us pretty dumb seriously to release only the second ever consumer facing app.
Yeah, and you know chat rept. We now know has Google kind of running scared in the search category, but now open ai seems to be coming for social media too. And what's really interesting is that Meta also launched an AI video feed in their meta app just five days before the Sora app dropped. The name of it is it makes you think that Zuckerberg must name these things.
Vibes, oh Man, vibe coding, Meta.
Vibes, It's Pineapple Express.
I saw this. You couldn't miss the Sora and news unless you're living under a rock. I didn't see a peep or hear a peep about vibes.
I think why you didn't hear a peep about vibes is because vibes sucks. You know, there's like a huge difference between what you can do on vibes in terms of Meta AI and what you can do on Sora and what you can do on Sora is that you can be a character in your own videos and you can also put other care I saw Pikachu have a DNA test on Maury.
You're serious?
It was like Pikachu, you are not. So for those who don't know, here's how the app works. You can upload a short video of yourself to the app, and then the model can seamlessly insert your likeness into this hyper realistic video that includes sound.
Do you remember when the text to image feature on chatchypt was released and the whole Internet caught fire and there was this moment where the Internet was completely flooded by studio ghibli characters and you know pixel characters of like you and your friends going about your life. When was the last time you saw one of those in circulation?
Yeah, it's done dead. I like we have it from jet whenever it was Yeah. So what do you think in terms of Sora?
Yeah?
If like it will add.
It's obviously so thrilling to put you on your your friends in Spider Man with Pikachu and Maury, Like, but is it durable?
I think this will be more durable than the kind of quick high jinks of studios Jubilee, Like, I think that this is really fun to do over and over again. Do I think Sora as an app is going to Sora in the app store, you know, in perpetuity the way that Chatgypt is just for sure. No, I don't think it will. I definitely don't think it will.
So telling you mentioned more in Pikachu, which is undeniable, will you confront him with just a flood of this stuff I.
Was, and the way that you know it when you watch a SRA video if you're on TikTok or Instagram, there's a Sora watermark.
Wow. But those are kind of a flex right because it's still in beta.
Right, And what it means is that you got invited to Sora, which is very cool, which we you know, one day.
Why did say in our production METI on Monday? Way, why aren't we going to invite it to?
So I don't know.
The ones that I actually like the most were of Sam Altman himself, who is very smart about centering himself as the face of open Ai.
Now. I just wanted to pause on this because one of the concerns has been can I take your face and sorrow you? And the answer technically is no right. So Sam, that's right, Sam's my dear friend. He decided to basically make his face available to anyone who wanted to use him as a character. That's right.
And there's two really disturbing videos. One is him shoplifting at Target.
Oh my god, that he made himself was made of him.
I think someone made it, put it on X and was bragging that their video went viral. The other one is really disturbing. Again, another user made it of Sam Altman being a sort of crust punk who had seemingly a crystal math problem.
My god. So these are not always in the best taste.
Not always in the best taste. But I think again, show that Sam Altman is the Zuckerberg or the bezos that open AI needs more. He's more more.
Yeah, I mean he is the greatest hype beast the world's ever seen.
Oh absolutely. I just want to go back and talk a little bit about some of the darker questions that Soa raises. To any normal person, you would imagine this is like a huge content moderation issue, because now there's just the Internet is flooded with fake videos that are
fake on purpose, and everyone knows they're fake. The company does state that it quote bands impersonation scams and fraud, and that there are quote extra guardrails to the app when real people are featured in videos meant to block nunity and graphic violence. Okay, so technically you can only create videos of yourself, your friends, or people who choose to let their likeness be public. But as one would expect, there has already been at least one video of a
woman being covered in a mysterious substance. So that sort of tells you how well this is going.
I'm curious. I mean, is there is there really a belief or des I hear that Sora the app becomes a destination to consume, to just gorge on ai slop. Is that the vision is here?
I think a lot of people would say, yes, it's like a it's a new social media destination for crap.
It's a place to hang out rather than just to create.
That's right, Okay, that's right. And there are two crucial things that make this app different than say Instagram or TikTok. The first is that the algorithm is not reigning supreme in the Sora feed. It's in as you said, beta, but apparently the app allows you to customize your feed a little bit, so you can pick a mood, which means that basically you tell the app what you want to see more of and it feeds it to you. The second thing is, we all know how much we
divulge to chat GPT. Sora quote may consider your chat GPT history while creating video, which is a bit strange, Like, to what extent is it going to impact? I guess my Sora videos.
You said something to me yesterday when we were talking about this that really stayed with me. You said, this may be the day the Internet changed.
I did say that, and you know I talk to our producer about it too, And I was thinking, was I being a little bit Pollyanna ish? I just I think we are already like treading such thin ice in terms of like what is true that we see online. There's now an app that is dedicated intentionally to propagating fake video, and it's being run by probably the most important company of this new generation. And so I do think there is a pre Sora and postsra internet. Do
I think Sora is going to change the world. Probably not. Do I think that having more realistic fake video in a time when our president posts Sora like videos on truth social is like incredibly damaging. Yes, I think you know.
Yeah, I mean, I'm I come back to the thought I had at the beginning, is this like a flash in the pan?
Like?
So, Scott Gallaway, who I mentioned up top, he's the podcast host of Pivot and a marketing professor, had interesting take here. He's very much the opinion that this is a flash in the pan that surprise, based on his thesis that most people don't want to be creators, they want to be consumers. And here are the stats that he shared to support that four percent of YouTube videos account for ninety four percent of views on the platform, five percent of TikTok videos generate eighty nine percent of
the views. On Instagram, three percent of videos and eighty four percent of all views. Scott's hypothesis, or his take was that actually Sora is really a B to B product. This is designed to put Hollywood and advertising agencies effectively to make many many people who work there redundant. Yeah, but that's not very culturally favorable cell And so the idea was to release as a consumer app, even though that's not really what it's for. Interesting, it's good take,
but I also disagree. I disagree because of what you said about Samilman catapulting himself into being the most iconic text CEO of our generation. And I think that the more that most people in America and the world experience firsthand the absolute mind bending magic of generative AI, the richer and more powerful Samilman gets.
Yeah. I think that's right. And I think that the last thing I'll say about SORA is like, I don't think it really matters. Is not that many people are using it versus consuming it. Isn't the problem that people are consuming it in such.
Hah ha, We're onto our next story. Okay. There was a piece in the Financial Times with the headline have we passed peak social Media? Have we per This article on the data that underlie is it? Social media use peaked three years ago in twenty twenty two. What So the Ft partner with this digital audience insight company called GWI to analyze the online habits of two hundred and
fifty thousand adults in fifty countries. What they found was that by the end of twenty twenty four, people on average spent two hours and twenty minutes per day on social media, which is down almost ten percent from twenty twenty two. This is a trend.
This is a trend.
You know, you know, per this data, which generation is unsubscribing fastest now?
Gen Z fascinating.
So the columnist posits that the reason this may be happening is in part because of the degradation of content on social media platforms overall. This is also known as incertification, the term coined by Cory doctor O. But the columnists used this great analogy to describe this flood of new content. He calls it quote ultra process content dopamine with at best negligible informational value and at worst corrosively negative, And he basically posits people are waking up to this.
You know, I in my youth because I am an Instagram super user and I post a lot on Instagram Stories. And what I have noticed is the way that the Instagram algorithm works. I can't find things as easily as I used to because everything is algorithmically driven. So I'm seeing the same shit my friends are seeing. And what it used to be is that like there was actually, say two years ago.
Before your feed would surface true value.
Yeah, because my feed actually had things in it that I wanted to see. Now my feed just has things in it that my friends are watching.
Well, that's again part of what this article is about, which is the thrill of social media was really about connecting with your friends, right, But now these apps basically promote just time spent with the never ending feed of sensational content, and so it's become more and more passive and therefore, listen, let's engaging. I mean, the snake is eating its own tail, at least according to this article. On the data, there is a twist here. Guess which continent social media is not declining in?
Not declining North America.
Yeah, the trend line is down everywhere apart from here. Daily usage is still rising year after year in this part of the world, and last year North American social media consumpation was fifteen percent higher than Europe.
It just scares me because we now have an entirely new app that's number one on the App Store that is literally I mean, you want to talk about in schitification. You want to have an ais I mean a Pikachu DJing at boiler Room, which is a real sour video that came out.
Are you looking for Pikachu or you just I.
Mean, I'm always looking for Pikachu, isn't that the but no, I just I think that the Galloway argument gets me a little frustrated because it's like keeping candy in the house all year as opposed to on Halloween, because it's like, if we only have it on Halloween, at least there is a day that's designated to eating it. If we have it in the house all the time, kids are just gonna be like, I want to eat candy every night.
And I think that with sore uh, it might not impact the American social media landscape in the way that you know, TikTok or Instagram did, but it does now put more shit into the world that obviously, based on what you're saying, Americans can't handle.
And I think but by the way, it's engaging. I mean, I like seeing this stuff.
I enjoy watching a giraffe and a zebra walk through suburban America absolutely.
As giraffe and zebra blended together. I promise more about open Ai. We've talked about the first, you know, triumph of the week from open AI's perspective, which was breaking the Internet with a new social video generation app. But there are also two other things. Announcing a deal with the chip maker AMD, which added one hundred billion dollars
to the market cap of AMD almost immediately. I mean, get this, just announcing it down a deal with the company makes that company one hundred billion dollars more valuable. I mean talk about power.
Can you talk a little more about AMD because I'm not so familiar.
Yeah, so AMD is, you know, in video, is the world's number one chip maker.
Yes, I know this islder and.
Open air also did a huge deal with Nvidia recently, and also did a huge deal with Oracle for data centers. But this was an agreement for open ai to start building data centers using AMD chips. And this was a huge thing for AMD because it basically was open Ai laying down a marker saying no, no, we believe that these guys can manufacture chips good enough to power the data centers we need to do the next generation of AI. There was a sweetener. Open ai had got ten percent
of AMD for free to do this deal. Well for one center share options by shares at one center share that is crazy. So they're basically, I mean they're out there saying we know that working with us will increase the value of your business.
So here you go, here's ten percent of it, right, And that's what happened.
So the leverage of this company right now is absolutely extraordinary. Bloomberg pointed out that the announced between AMD, Oracle and Video that open Ai has annouced in the last few weeks could top a trillion dollars. A trillion dollars. I mean even even even the president rarely says that word.
He doesn't say trillion.
One stock analyst told Bloomberg that Sam Mortman quote has the power to crash the global economy for a decade or take us all to the promised land.
What is taking us to the promised land?
Look like? That's what I'm wondering that they didn't have very good question.
They didn't go there not taking rich. Yeah that's rich if they have starvation pays off right right right. I mean, if there's anybody who is going to make the AI revolution pay off, it's the founder of open Air.
It has to be. In the point article makes is now all these companies are so much in each other's business that Open Air with on the software side, AMD and a Video on the hardware side, that this is why some people are worried they may be a bubble emerging because just by doing business with each other and swishing the cash around, they can inflate each other stock prices, borrow more money, do more deals. But of course that's why if we don't get to the promised Land, we'll go somewhere else.
There was also another one little piece of information about open ai this week, really big open air.
It's a huge open Yeah. So we've got Sora, we've got a md chips. But there's also develop a conference on this idea of the everything app. So there was all this concern about how generative AI was basically rendering the Internet useless because all of the Internet's information was
served up through a chatbot rather than you know, in search. Now, interestingly, companies like Zillow, Booking dot Com, Spotify have voluntarily struck partnerships whereby you can use their products through chat GPT, So you can say find me a house in X, book book me a flight to Y, and without leaving the chat GPT app, that app will interact with these other companies. Now, of course you get the exciting bump of we're working with open ai.
With the first exciting Etsy was very excited about it might have been one of the first.
On the other hand, you want kind of alizes people using your app or your website. I mean, it's interesting, it's interesting trade off.
It is an interesting tradeoff, but I think it goes to show. It's kind of like when people started selling products on Amazon, like, are are you going to destroy your storefront or are you're going to go with where people are buying the product? Which is very interesting. So I was thinking about this idea of the everything app, and our producer Eliza reminded me that China has a few everything apps, Ali pay and we Chat. Yeah.
I've actually I've been to China and so you open we Chat and essentially in it you have the equivalent of Uber. You have the equivalent of Amazon, like you can do it banking, thank you to everything through one app and it's been three days. The Promised Land is
what's obvious. Why it's the promised land, right, because if you owned everything up, you're pretty powerful and exactly Elon of course the reason he renamed Twitter X. One of the reasons was he wanted X to become the everything app, which.
Does not look promise but chatchipt Yeah, I mean in terms of a contender, not meta, not TikTok, chatchipt seems to be the closest thing we have for a contender in the everything app. And I'm cure. I mean, I certainly think the phone is the everything app.
The phone is the everything app. And what will it be displaced by building a new physical interface to use their everything app? That I doubt more.
Yeah, I think that's probably unlikely. But I did have this little thought that I can reflect on for the next year, which is, like, to what extent will the apps on my phone dwindle and dwindle and dwindle until there's like three.
Left after the break? Mini brains might be the future of computing. Robots are helping to create babies. Then I look at raising kids in the age of AI, and we're back, Cara. I have a story that I think might blow your mind. What do you think of when I say wet weear? I remember this is a family friendly show.
Oh well, then I think of a wetsuit, but you surf.
So wetwear is not hardware and it's not software. It's organic computing.
And I literally don't understand what you're talking about.
Okay, So I don't really it, to be honest, but I did my best. Following a BBC headline scientists grow mini human brains to power computers.
What what do you mean by mini brains?
Scientists in Switzerland have found a way to create stem cells out of human skin cells, which are developed into clusters of neurons that function as mini brains, also known as organoids.
This sounds like science fiction. This is like Revenge of the Organoids or something.
Well funny, should mention one of the scientists leading this initiative is a huge science fiction fan. He said how thrilling it is for him to make science fiction become science fact. But this particular field of study is called biocomputing, and the goal is to make these organoids powerful enough to function as quote, living servers. Essentially, with the ganoids attached to electrodes, the ideas that they will be mini computers that can also learn, much like AI does.
So essentially they're creating organic AI. Is there any fear of sentience like we have with jen Ai?
Perhaps even more so? So? This quote from the BBC really haunted me. Sometimes scientists observe a flurry of activity from the organoids before they die, similar to the increased heart rate and brain activity, which has been observed in some humans at the end of life. Now you could say, well, this is like a computer overheating before it finally collapses.
Or you could say this is the symphonic moment where you know, all the cells come together for one final burst of energy before they pass into the next world. But it's to me, it's fascinating, It's unbelievable.
I guess my question is like, what is the upside of doing this?
Well, I mean, the big story about the AI revolution is how much electrical power it requires to continue versus us our brains, which require no electoral moath to function. So the idea is basically that, you know, we can do computing with much better energy efficiency.
You know, I've many many questions about this, you know, mostly around the ethics of having living cells as computers. But my first question is like, how the hell do these things work.
Well, it's still early days for this type of work, and it's honestly pretty hard to understand. But the Swiss scientists are testing the organoids to see if they can obey simple key commands. They're not really up to snuff yet, and they only live for about four months, but interest is spreading. Back In twenty twenty two, an Australian firm called Cortical Labs was able to get quote artificial neurons
to play the early computer game Pong. But some of the most promising stuff around this research is, in my view, the scientists that JOHNS. Hopkins, who are studying how these mini brains process information and how that in turn may be applied to treating diseases like Alzheimer's.
I actually think my next story could be similarly unsettling for some people, maybe even more people, because this is more relevant to people's lives. It's from the Washington Post and the headline reads, robots are learning to make human babies. Twenty have already been BORNA where are you going with this? It's actually, you know, it's not as sci fi as it sounds, unlike your story. There are a few companies that are running trials and using AI and robots to help with the IVF process.
Okay, so this is basically automating the IVF process exactly.
One of the companies highlighted in this article is called Conceivable Life Sciences. As a pun, yeah, I you're right, it is fun that's what are we gonna call it conceivable, So conceivable life sciences use AI to identify the healthiest eggs and sperm, which is often a hard thing to do, and use robots to extract the eggs and fertilize them. They started testing their system on mice and are currently conducting a one hundred person clinical trial in Mexico City.
There's also another company that I want to mention called Overture, which ran a three person clinical trial in twenty twenty three, and The Washington Post interviewed one of the participants, a woman who is the mother of a completely healthy and normal two and a half year old.
So, and we talk a lot about replacing doctors, but it tends to be more in the realm of diagnosis. But what this story is about is inserting robots into the medical process. Presumably, if it worked, it would be much cheaper than the human a lot cheaper.
That's a very good point because there are a few advantages to robots and AI. Robots don't get tired, they can see better than a human being can, meaning in theory, they can select better sperm and insert more reliably into eggs. So the hope is that the IVF process can get faster, as you said, cheaper and better with the help of this technology.
So normally this is where we do our chatamme segment where we hit from our listeners about how they're really using chatbots. But this week I had the chance to talk to the hosts of a new podcast called Raising Kids in the Age of AI.
Every single day I wake up and I go, oh, yeah, I don't have kids.
Cool, you don't have to raise kids.
I don't have to raise kids in the age of a phone or a IPEP maybe soon.
No surviving it yourself. So whether we have kids or not, we are living through this sea change about the future of how we work, the future, how we interact, and how we move through the world in general. And this podcast provides actional advice about how to help young people use and harness AI. Kaleidoscope is actually one of the producers on the show, so I'm biased by ConfL I think I do think it's very interesting.
Who are the hosts of the show.
So. Alex Catran is the co founder and CEO of the AI Education Project and the AI Education Project partner with Google to create this podcast with us at Kaleidoscope. Alex got his start in politics and policy and found himself attending all these flashy AI conferences, and Ax became consumed by this idea.
How do you make sure that every kid is having the same conversations with the future that the futurists are having.
And who's the other host?
That would be doctor Eliza Pressman, who's a developmental psychologist and a podcast host and also a parent who's constantly talking to other parents about raising kids in the Age of AI, and said she felt like she needed better answers.
I'm in a position where parents are asking me about this constantly, but I don't, you know, sometimes you.
Don't know what you don't know.
So there have been things that I didn't even know I needed to be either worried or excited about.
I think worried and excited is the balance that we strike on this podcast, especially today's episode.
We'll wired and excited about no unknowns, unknown knowns mini brains. Yeah, I mean, so that's the tone they strike on raising kids in the Age of AI. Alex and Eliza have the goal of reminding parents and teachers and caregi as anyone who interacts with children, there's no time like the present.
If you think about this like learning journey, you really will it will essentially never end. And the good news is, you know, pretty much all of us are very early on in that learning journey. There's almost like there's very very few people who have any expertise in AI prior
to you know, November thirty, twenty twenty two. And so really my message to all teachers, to all parents is like, if you feel out of your element, if you feel like this is that you're in over your head, It's like you're in good company.
The episode of the show that dropped this week is actually all about AI in the classroom, and that's something you and I have been very interested in. Carara, Alex and Alisa talked to a few teachers who are using AI and facing its implications with the kids they teach, and we're resting with the question of how do we motivate students to learn when there's a tool that can just give you the answer.
Yeah, So, how do you make sure students are learning to use AI without giving into cognitive offloading, which we've talked about quite a bit.
Well. Doctor Lisa mentioned one program in particular that gave me some hope for how education might be powered by rather than corroded by AI.
There is a course being taught at I believe Tulane University which was leaning into the fact that the students, of course are using AI. So let's show them how they can use AI and then question it and question themselves and question each other to show that they understand the material.
So what I think she's saying is it's better to not be punitive. It's better to actually engage in a real conversation with students who are sort of at the forefront of all this technology.
Absolutely, and I think you know, it's interesting. The education system, certainly in Britain, I'm not sure about the US, was built around oral exams, and so you know, using AIS a discussion prompt and discussion partner rather than a work
replacer obviously, you know holds real promise. If you want to hear more from Alex and Elisa and learn about how AI is shaping the future of learning and education, make sure to check out Raising Kids in the Age of A, available on YouTube, Apple, Podcasts, Spotify, wherever you listen or watch.
And even though you didn't get one today, we still want to hear those Chat and me submissions. You can tell us your crazy chat GBT stories by emailing tech Stuff podcast at gmail dot com. That's it for this week for tech Stuff.
I'm Cara Price and I'm as Volochin. This episode was produced by Eliza Dennis Tyler Hill and Melissa Slaughter. It was the executive produced by me Cara Price, Julian Nutter, and Kate Osborne for Kaleidoscope and Katria Novel for iHeart Podcasts. The engineer is Beheth Fraser and Jack Insley mix this episode. Kyle Murdoch wrote at themself.
Join us next Wednesday for a conversation on the ins and outs of AI video generation and what's Sam Altman gains from plastering his face across his new platform.
Please do rate and review the show wherever you listen to podcasts, and reach out to us at tech stuff Podcasts at gmail dot com. We really do love to hear from you.