Loading...
Loading...

You guys ready to do this? Are you guys ready to have the greatest text predictions ever?
Yeah. Okay.
Anyway, I'm sorry. Casey, are you okay? Are you on Twitter arguing right now, Casey?
Uh, no, actually, I was just going to post something that is relevant to my prediction.
Oh, I could reference it on this, this particular show.
Oh, he's playing chess. We're playing checker. Yeah, he's playing chess right now.
Oh, I got as a no pad up.
I'm going to make the most ridiculous predictions for 2026 and tech and not only me,
but also Casey, myratory, trash dev and teach.
We're going to be making the greatest tech predictions of all time at the end of it.
At the end of this, one of you will call us no stodomas because we're going to be
correct. 2026 bingo card is going to be filled in and we are going to be future
tellers. Yes, anyways, that's true. That's what we're going to do today.
And so he was going to say that. Yeah, he did because you were on mute and you were
actually saying the same thing. Uh, but the thing that makes this special is that we're
not just doing any sort of prediction. I like some lame ones like Sam will say AGI in
2026. We want real predictions. Okay, we're talking about ones that are outrageous and
outlandish and that people will find shocking if not disturbing.
So be prepared. All right, who's going to go first is the real question.
Who's going first in this tech prediction cycle?
I assume what we do is we each give one prediction. Yes.
Then we each give another prediction. Not going to be all three, right?
Yes. Yeah. You don't want to do all three at once because I feel like that's a waste
opportunity. We all give one. Yes. Yeah. Yeah.
We're tailored by lists based on everyone else's. I have a cup.
I have a lot of choices here.
I'm not going to go first with my with my most obvious one.
Yeah. Okay, because it's obvious because it's obvious.
You're killing me, merch cop. That little stun of yours turns with six hour post mortems.
He's emerging to prod.
You have the right to remain silent.
Come on, Comish. That code wasn't clean in you. You know, I don't got time for this.
Then our Cs are being vandalized across the city.
A diff. No, no, no. The difflers are mint. The difflers out there.
And I'm going to be the one to deprecate. You need to focus on your Gira tickets.
Not chasing a ghost. No more cowboy coding for you.
I'm assigning you a partner. A partner? You can't do this to me.
I am a lone wolf. I'm an IC.
I do not need some dead weight junior dead. Hold me back.
Oh, no. You're not reverting this one, merch cop.
He might actually teach you a thing or two. He did graduate top of his class with a flawless
C.I. record.
Merch cop, meet your partner, Lieutenant Squash.
Pleasure, then, at your acquaintance.
Merch cop, Merch cop, unit tests, functional tests, and the end tests, acceptance tests, performance
tests, load tests, stress tests, math tests, system tests, internationalizations.
I think it's all units. There has been a reporting of a diffler sighting at a local
cafe. Copy. We're on it. It's our chance to get the
diffler. Get in. Compile to ability test. I just
want to throw more kinds of shots at twish. It's time to get the diffler.
Standard tests, snapshot tests, smell them.
I've been telling them I need new meals.
The commission that we had to hold by the car, I'm about to get forced my fist in the
diffler's face. We have to stick to the process.
What do you think you're doing? Building in public?
Wrong answer, diffler.
I know my rights. This is just a side project.
What is Merch cop even doing? I'm working on my side project. I don't even need him.
I'm using code rabbit. With something like code rabbit, it's like having a
co-founder, always watching my back. I'm not going to leak customer information.
I'm always going to be up to date on coding best practices.
You don't believe me? You can try it too at coderabbit.ai.
Next week on Merch cop.
So my obvious prediction that I think everyone should be predicting now,
not just because it's true, obviously, but because it's funny, obviously.
I think 2026, it will in retrospect, like when the dust has settled in the year 2040 or whatever,
when we look back at computing history, and we're like, when did it happen?
What was the year? What was the year when it all finally happened?
I think we're going to say 2026, and I'm referring, of course, to the year of the Linux desktop.
The actual year of the Linux desktop. I think 20. That's a good prediction.
I think 2026 will be the year that goes down as the tipping point, because 2025,
Linux got a lot of momentum. Some of it from Microsoft, like by just the controlled flight into
terrain that is Windows, the operating system, that's giving Linux a huge boost right now.
I can certainly speak for myself and say that I have so many more Linux machines running now
than I ever did, and it's all just based on fear of what Microsoft is doing to Windows,
and it not being a stable platform, which was the only point of using it in the first place,
right? The only point of paying for it is that it was supposed to be more stable,
easier to maintain, you know, more turnkey and all that stuff.
Soon as that stuff isn't true anymore, you know, if your driver compatibility is suddenly worse
on Windows than it is on Linux, why are you there, right? And so there continue to be more and more
reasons to leave, less and less reasons to stay. And so I think 2025, it's kind of picked up
this momentum. And in 2026, in theory, we're getting like the steam box, for example, like our first
sort of standard consumer Linux distribution pushed by a major, basically monopoly in the space,
like they own game distribution in PC. And they are now saying, here is our official platform,
and it's not Windows, right? So I feel like there's just a lot of momentum going into 2026,
and I feel like this could be the year where adoption starts to really pick up. So I'm saying,
when we look back, 2026, you're at the Winx desktop. Love that. I love it. That is, I feel like
that's a correct prediction. Yeah, I'll go. I got one that I know, I've got one that I know,
Casey will like. I was going to say, let's go to alphabetical order, which is also you next.
Is it? Oh, yeah, because the prime is in teach and trash. And so,
I like that. Okay. There will be a viral vibe coded app that gets hacked, but this time,
it will be unnoticed for at least one month. So people who continue to use it. So this time,
it won't be some security researcher blah, blah, blah. This will be someone is secretly taking over
the vibe coded app for at least one month. That's my measurable prediction. Okay, but is there any
like extra sauce you're going to put on this prediction? Like, if you had any idea how computers work
at all, a quick look at the logs would have told you you would have been being hacked for like months
on end. It would have been like, oh, I'm getting 5,000 requests per second from China. Yes.
Sure. I think the hack will be so obvious in retrospect that we will have to do at least one
episode of the standup about it. So okay. That's okay, but that's good. Okay. So now you actually
put something on it. A hack will be so unnoticed and go on for so long that we will do an episode of
the standup on it. Right. Yes. Well, I have to feel like saying greater than one month. Unnoticed,
that's my measurable thing. I'm trying to do, you know, measurable predictions here. I want it to
be, you know, to be known. So there you go. That's that's that's one of mine. All right. That's
pretty good. That's pretty good. All right. No, no, I'm next. Wait, I'm at the ATT arc mod, dude. I'm
next. Okay. I've been out of order. Okay. So it is challenging. This is true. It's true, especially
in this JavaScript. All right. So I've been thinking a lot about this. And, you know, I kind of
made a realization about the AI industry. Is that there's some big players in it right now,
right? So you got open AI, obviously a big player. You have Google really just, I'm in shocked at
how much progress they have made in the last six months. They're they went from losing on the
poly market to absolutely dominating the poly market right now. And so you see all these kind
of companies cropping up like an anthropic, a company that's solely just doing AI and you're just,
okay, well, you know, there's a lot of people popping up, but I realized something. There's a
there's a name that is mysteriously missing from the list, despite them actually having an AI
product that people just don't say very often or don't bring it up. It's like Ben brought up. I've
seen it said once or twice, like Apple's Apple already gave up. They already said, Uncle,
and they went and said, Hey, we're going to just use Gemini, right? So I see where this is going.
Yeah, I think I'm feeling it. So here's my big one. Here's my big prediction, which I think
is going to have like I genuinely actually think this is going to happen. I think Amazon is going
to purchase anthropic. Oh, I think it's yeah, they had own bond, but I think that Amazon is
purchasing anthropic because they need to make inroads into this whole thing because they already
had whisper right there. He were trying to do some level of this kind of coding interaction world.
They just never really got anywhere with it. I literally know nobody that's using any of their
AI products. And so it's like they have one called Q or something, right? Yeah, they have something.
They have like a coding. But the thing with Q, the craziest thing about that is you can tell
where all of the Amazon headquarters are where they have in office thing because that's the only
place when you look at Google trend search data where people are searching for Amazon Q is literally
in the station. You can find every single one. Yeah. So this is my this is my guess is that
Amazon to make bigger splashes and to get that because they also they haven't been doing like
they haven't been doing the world's greatest on the stock market, right? They've been hovering
around their price for quite some time at this point. Their year-to-date earnings is not good,
not it's not looking like a fantastic thing. It's only up about 4.68% or something like that.
So they haven't really been nailing it. They kind of missed the boat on AI. At least that's
perception I'm getting. And so I think 2026, they're going to try to turn that ship around and
they're going to come back hard. I love that. Thank you. That also solves a bunch of problems for
anthropic because my understanding is they're hosting an oversight. They are not as good at doing
that as like AWS would be. So it acted like there's a I like this one prime. That's a great one.
Thank you. I mean, teach, are you referring to synergies?
So it sounded like you were referring to synergies that could be exploited.
So let me put on my business hat. Yeah, there we go.
Is they see? Oh, yeah. I was like, why could I see the writing for you?
Yeah, he took off the top part of his head. He's putting on the business side.
Wait, why is your head green? Let me put on my business hat here. It's an elf hat.
There you go. I will call it synergy. And we'll circle back to this later. Thank you.
I feel like there's no way anyone would ever want to be acquired by Amazon. There's a lot of people
that have been acquired by Amazon and have made a lot of money. I mean, that's like the main reason.
Is it acquired by Amazon? Twitch is still here. That's true. I'm currently streaming on an
Amazon acquired service. Rather, surprisingly, because it was, I feel like it was not managed
particularly well, which may have also been true prior to the acquisition, but it's still here.
Oh, the management is horrible right now. I would most certainly say I'm not going to bring
anything up, but let's just say that they have the single most controversial unbanning. I think
of all time. Yeah, well, it's always been that way. They're kind of like a,
this is going to be a horrible thing to say. They're kind of like a parent who is
on an addictive substance, right? They do completely like unpredictable behavior in a way that
makes people scared and unhappy and then they do something nice. It's really bad. If you just
analyze that as a human relationship, you'd be like, I have to get all of these streamers out of
the home. You know, because of like the way they behave and play favorites and all these other
things, so it's weird. I don't understand why it is that way. But chatter protective services.
Yes. Yes. I feel like, yeah, I mean, streamer protective services, even, I don't mind changing
the acronym. It really does feel a bit wrong. But anyway, that's the topic for some other
for some other cast. So, all right. So Amazon buys Anthropic. Yep. That's a, that would be
a big announcement, obviously. Yeah. That got said. Yeah. I think what the price is too.
Oh, I never thought about the price. The price, I honestly haven't looked into Anthropic, but I
know it would be at least 10 figures. How many numbers? I had to like process. I mean, I'm sorry,
11 figures, right? It'd be 10 plus billion, right? Yeah. Yeah. Well, didn't they just acquire
bun for like, yeah, multiple billion. It's got to be more than 10. No, it is valuation.
Anthropic is also under a settlement to pay 1.5 billion dollars in damages already. So I'm
pretty sure 10 billion isn't going to cover it. Yeah. Well, I just said at least. Okay. I
say under 100 billion prime, over 100 billion. I would be shocked if it made it to 100 billion. Okay.
That puts a hard cap on it. I'm just, I'm just picturing like, you know, that meme where it's
the guy in the theater from the boys or whatever. You know, when someone's mad or watching it. The
lights just flashing on his face. I'm imagining all the people from bun saying they, they changed
their Twitter profiles of I work at Anthropic. They're now switching to I work at Amazon.
They're just all upset about having the change into Amazon.
Like two months later to just like, all right. And they're just all so sad about it.
That's that's how I would feel to be honest, if that actually maybe, maybe they can buy with rainbow
six siege credits, though. And they have billions of. So I feel like everyone else's predictions
have been way more realistic than mine. So maybe, but I got work. Okay. Here's, I got one that's
kind of weird. Uh, okay. I think we're going to have a reckoning of, I don't know how to say this
without like sounding terrible, but I think we're going to go through. I've been like trying to
be how to say this in my head this whole time. I think we're going to go through another wave
of segregation between AI users and non AI users. I was watching teachers face. He was like,
what are you about to say? So like, uh, so there's going to be specifically human only spaces.
So I think like one of the big things right now is a lot of people are like relying on AI as a
crutch for like personality, knowledge, whatever. And like, nothing feels real. And I think
this is only going to happen like in spaces like in San Francisco or something where like, you
can't come in. It's like basically like a no device policy, right? I think it's going to be
something like that entered in. I don't know how realistic this, but in my mind, I would love to
like, like, when clearly came about and their whole commercial was like, I'm going to wear these
glasses and I can just know all this stuff. It really just turned me off. And I would really
hate a reality like that. Um, and if it comes in if devices or companies like that become even
more bigger and more advanced, this will be more and more of a problem. So in my mind, I was like,
okay, people are like, especially people that are hardcore anti AI, I can see them like popping up
like a human only space, even for like hackathons or something, right? Um, but yeah, that's,
that's one of my predictions. I'm a prediction. I'm a prediction. I just
similar one where people put stickers on like projects or other things like no AI used in this
project like organic, you know, anti clankers. I love it. No clankers allowed in this. Exactly.
Exactly. I think it's. Yeah. Yes. For sure. 100% on your team. That's a great prediction.
Steam already does this though, right? Doesn't steam have some sort of, uh, they had like a no AI
generated art was used in this or no AI something was used in this, right? I think they're having
some limited AI disclosure policies. I don't remember exactly what they are. I don't know if
they're required or voluntary, for example. Yeah. Okay. I know it's something. Okay. But they,
yes, you can put that on your page. Like there is, it has happened. Yeah. AI disclosure. Like
how much AI did you use? You can put it. Interesting. Okay. You guys want to do snake draft?
Can we go back? Yeah. Oh, okay. So it's not me. Yeah. It's not me then. I get to go again.
Okay. I go, uh, I forgot to write this news, name down, but the Palantir CEO, the guy that was
like jabbing the sword. Yep. Yep. Yep. I predict he gets arrested this year for something because
he just freaked me out. Okay. He's getting arrested. That's all I'm saying. I don't know what for what
and why? But that do creeps me out. He's buying the addresses here. 2026 prediction. All right.
What? Okay. Do you think he will be? Don't ask me. Do to use use of Palantir software that is
able to track his like illicit child pornography, money laundering, drug deal, whatever it is,
or do you think it will just be a regular arrest? I think he might stab someone with that sword. He
was just jabbing around. Okay. If I had to guess. Okay. If no complex observation software,
murder like that. No, man. That's right. Just holy cow trash right now. Go check Palantir stock.
It's it's going down. You just crushed their stock. You just, yes. That's as far as my,
if it happens, we have to like do some kind of watching party. I don't know.
A watching party. We watch and purplock. We watch and hit purplock down on like loop on prime
stream or something. I was right. Yeah. There we go. Can't wait. Okay. Prime your turn.
All right. All right. All right. All right. Here's my second one. Nice second one.
Yeah. I feel like this one's a fairly obvious one and I kind of feel like I'm cheating a
little bit saying this one. It's okay. And I am going to say it and I'm going to say it with
my chest. Unlike trash. I'm not going to bid. Kind of feel bad saying if I do love how you said
segregation, then that's why I thought that I wanted to avoid that word. I didn't know.
So you paused on it. I said, you made everybody question. What was about to happen?
All right. Let's see. I think so. This is actually a long case. He's lines where he was talking
about how bad Windows really has been doing and how good other services have been doing.
This is really spawn from cursor and it's a recent acquisition. I think cursor
is going to make a play into the repository hosting space.
And the reason being is that I think that a lot of vibe coders, they don't know let alone care
about repos. They don't know what the heck GitHub is and what the difference between Git and
GitHub is. I think you have no strategic mode in the day and age of vibe coders other than they
need version control to be able to walk things back and to be able to store it so that other people
can access it of some kind. And so I think that we're going to hit this thing where GitHub has lost
a bit of its stickiness in the next gen version two of programmers, so we say, or version three
at this point. And with the V3 programmers, since they don't care. And then I think cursor will
deliver something that is objectively better. And I think when it does, there's so many assumptions
in so many pieces of software about like using GitHub. Like a good example of this is a
NeoVim. In a lot of the package managers, you type in like the Primagen slash Harpoon. And it's
like, ah, I'm going to go check GitHub for that. I think the moment this thing comes out,
that people are going to flood to whatever this alternative service is so much that that service
by the end of the year, there will be package managers that default to cursor.hub.com or whatever it
is. I do think that the Microsoft in much the same way that they've primed people for an
Exodus for Windows, they've also primed people for an Exodus from most of their software.
Like, yeah, in general, their posture towards their users is like openly hostile most of the time.
And so what you basically create, when you do that, is this sort of power keg that a competitor
can come along in light, right? And, you know, I think they've definitely done that for GitHub.
And probably for other services too, I'm imagining things like Microsoft Teams and Office 365,
which we don't probably have a lot of insight into because I don't think any of us here has to use,
the, have to use them. They're probably in a similar situation where like,
they're open to like collateral attack because they're just so abusive, right?
I've got a prediction that's going to run a little bit counter to that one.
Okay. Oh, no more controversial. I'm just going to throw it out there. Okay. Here's what I think.
I think GitHub is going to split from the AI division and release at least one feature this year
that Casey thinks is good. What? But to be clear, last then five, I'm not saying Casey's going to
use GitHub. I'm just saying there will be at least one feature in release where they're going to,
they're, they're seeing writings on the wall. They saw last time, holy cow, we wanted to make some
pricing go down and things are bad. It's so bad that Palmer's going to get in there, not
Palmer lucky, but Jared Palmer, they're, they're going to just be like, we're so dumb,
we're going to throw this all the way time to get this out of the AI division and do something
different. That's, so there we go. I think they're going to try this year. Do I have a lot of hope
for that? No, but I want to stake my claim out early that they're going to try and get out of the
AI division and release. I'm one feature in a year, guys. Come on. They can do it. I know it.
There you go. We'll come back to here. We'll come back. We'll see.
Is it me? Yeah, go ahead, Casey. Hold on. We got to pause.
Is anyone reacting to that? I have reaction to that, but it's definitely anti, it's
definitely kind of going in the opposite direction as primes. Not really, but I mean, it's,
no, very opposite. I'm like, what do you think? I think cursor can still do the same thing.
I still think cursor should try and compete with GitHub. I'm saying, that's why I think
on the Microsoft side they're going to do. I'm just throwing it out there that I think that
your take is so much less likely. But I will want to say that my next take, by the way,
is going to be hilarious based on your take right there. Okay. Nice. Nice. Wow.
So, Prime, you should just say your next take then. No, no, it's okay. Sorry. Turn.
All right, Casey, go ahead. So this take is, wow, sorry, Casey, your reaction to that was,
so this next take is perhaps we'll get disqualified for non-specificity. And I apologize for that,
but I'm not knowledgeable enough about the space to give the like, here is exactly how this will
play out take. But my prediction is that in 2026, there will be a highly controlled implosion
of open AI. And what I mean by that is that like open AI will not like go bankrupt or whatever,
like it will not implode in a way that appears to be an implosion. What will happen is there will
be some kind of, you know, back room thing where they figure out like what's going to happen with
open AI because it's not really sustainable. But a lot of people who are very powerful and
control very large successful corporations need to make sure that it doesn't be seen to just
completely explode. And also it has potentially valuable IP that they want or contracts that they
want or whatever else. So there will be some kind of like transition of open AI from its current
state into some new state that is sustainable. And that will be done by external people with lots
of money or something. So I guess at very vague, but just like it seems like they've hit a price,
a capital outlay point that is not sustainable. And so they will have to kind of just be owned
by someone who actually makes money soon. I like that one. I think that that's fair.
It was actually that was, I'm not going to say this, but I was on mute. I was actually going to
have that as my that was that was runner up to my third prediction was that open AI will have to
capitulate to the fact that they are in over their head. And they have to be some sort of like
transition where Nvidia will effectively adopt them in or some other company will become owners
of open AI, which I and again, I guess like I said, I'm not knowledgeable enough about like the
inside info of like the I don't I don't have the kind of insider knowledge necessary to predict
what this will look like. It could be that it's way worse than I'm thinking and there actually
will be like a severe implosion. That's possible because I don't know like, but my assumption is
that it's not really possible based on how many deals have been done and how many players there are
like open AI is currently losing the race for being the AI that people care about the most.
Like it seems like they are not going to probably win that race. But they're also not last or
anything. So people who have been struggling more like meta, like Amazon, I would see them
you know, having a vested interest in acquiring that IP or something, right? Like to someone who's
trying to play catch up. So it just doesn't seem likely that they actually implode in a real in a
real spectacular sense. So that's why I say controlled, highly controlled implosion like an acquisition
or whatever. Yeah. It is true though. Open AI is like second place on everything.
Like everything they do. They're not the best coding one. They had the most emotional gooner
like capabilities for a long time, but they neutered that with 40 prime. Who do you think has a bigger
like consumer subscription amount of people like subscribing to chat GBT? No, I know that they
they're winning currently by name. But what I mean is like, they're not the best coding one.
Everyone likes. Okay, let's just look up the numbers of how many people visit chat GBT.com versus
like, who are you going to compare it to? Well, I think Google Joe and I is going to start doing a
large. I think it's going to start eating into a lot of stuff. I think all the other ones are
going to eat it away. And I think that like right now in business for business usage and
thropic has succeeded and has now overtaken open AI as the number one used AI for coding
in business businesses. Yeah. Yeah. Yeah. So it's like, I think they're going to just start
losing like an image generation in video. I think they're going to lose the gem. Like they're
just going to keep on losing because they're just they're they're not quite number one.
Yeah, I don't know. But I'm just saying, I think you're discounting how many huge name recognition
though. Yeah. Well, but I'm not saying just like not just that people know the name.
I'm saying, everyone uses shatterers to chat GBT.com is unfathomable. Yeah. Oh, yeah. I know.
That's I just assume that's just all name recognition, right? Like, which is again,
is why he's the yellow is the AIs that they're all talking about, right? Like they don't have this
concept that there's many AIs, which I think again is why it's unlikely that they will just
completely implode because it's like they will have there will be so many people like the only
question is have they created a company structure so complicated that no one can actually come in
and legally salvage a bit. So just feel like lawsuits for like 10 years or something. So there is
that, which I don't know how to factor in. But like, yeah, otherwise, I imagine, you know,
Acrimony aside, somebody or some set of somebody's will come in to clean up the mess because there's
too much value in the mess for them to let it go, right? Chat GBT gets six billion monthly visits.
There you go. And they and they have an AI that's very good compared to a lot of the further
down the list players. And so unless all of those players massively catch up in the next year,
they're going to be a very attractive acquisition target even at a high price point and even in a messy
situation. If you can basically then take that sort of as a as a way to jumpstart your failing AI
program at some other company like meta or somebody, right? Yeah. I that might be harsh to meta.
I'm sorry. I don't know the state of their AI. I just know that nobody seems to care about their AI
right now. So like getting something like a chat GBT seems like a win for them. Would they have
like a model or anything? What's what's up with? They were doing all the llama stuff. They have
opened all my all that stuff. I don't know if they have a closed they have they have a gigantic
investment in it and like a special like division entirely for it and all that stuff, which I've
read about, but like I again have no insider knowledge. So I don't know like where they're at.
Like for all I know 2026 could be the year of meta's brand new amazing AI model that they've
been working on or something, right? But but I just have no insight into where they're at.
They could just be running around like chickens with their headcuts off as well. So to be clear,
I think open AI will I actually like like Casey's prediction. I'm just saying calling them second
and everything is crazy. That's not true. Okay. Yes. Not second in visits. I was saying seconds
and capabilities. I don't think I think that the other models are superior to what they do.
Well, and also just to clear, I mean, since I haven't talked about that part of it, I'd also
just clarify like this is my prediction is not a statement on the quality of opening eyes. AI's
either. It's more just the fact that the capital expenditures required for these things are just
absolutely insane. Yep. And so when you look at the kind of game they're going to have to play if
compute really does continue to be the primary thing that gates your advancement, right? Either
because you need to run larger kinds of inference runs or just because even if you don't,
but you need to be able to try lots of different kinds of inference runs as your engineers try to
figure out algorithmic advancements because straight compute doesn't really cut it anymore.
That still means the more compute you have, the more independent experiments on training
you can do, right? And so when you look at the capital expenditures that you're going to have to
make to compete in this space, the problem is just that all of open AI's competitors, all of them
have way more flexibility. Like Google doesn't even lose money. They make astronomical
amounts of money today and they can do that capex, right? So they can just outspend open AI,
they know they can look from the outside and basically know exactly how painful it will be to open
AI for them to buy this much power for them to build these many data centers for them to have this
much advanced chip orders or RAM supplier, whatever it is that year in terms of what's the hardest
come by power? Electricity is currently the thing, right? RAM. And not to mention lobbying.
Everybody, Google, Microsoft, Meta, they have all developed very advanced lobbying capabilities
in the United States. Open AI to my knowledge has very little, right? So when you're talking about
we need to deregulate this zone so we can build a nuclear power plant here or whatever the long game
is, right? They also can't compete there currently. So I just don't see them being a long-term
independent player. It doesn't make a lot of sense to me. And so maybe, I mean, there is,
there are ways my prediction could be very wrong. Like, like, they just never employed. Like,
not only is it not 2026, but it's never. And that's if they manage to have enough sort of
deal-making and like weird balancing act stuff to just keep people keeping them afloat without ever
actually fully acquiring them. It could happen, but even then, again, like I said, it's like the idea
that they're going to run as like this independent company, I think, is just kind of over.
All right. Casey, you get to go again now. All right. So the last one, and this is by far my most
important prediction. The other two were about fairly trivial matters, such as the world switching
to Linux and OpenAI having a controlled implosion of some kind. This one is actually high stakes
and very important. And if you open up my Twitter feed right now, you will see that I posted
a tweet from January 21st of like this, this past January 21st, right? So almost a year ago.
And what happened was I had attempted to send Jonathan Blow a box of William Sonoma signature
peppermint bark for Christmas. And I ordered this well in advance of Christmas, like several weeks
before Christmas. And to this day, they have still failed to deliver this box of peppermint bark.
And I posted many Twitter updates as they have falsely updated the shipping page for this thing.
I made a prediction last year that Jonathan Blow, who is currently working on order of the sinking
star, would ship that game, which at the time was still quite a ways out from shipping. And
still still is that he will ship the final version of that game out to customers. And you can buy
it on steam and on console, whatever else, before William Sonoma figures out how to put one two
pound box of peppermint bark in the mail. And currently, he's seemed to be on track. John has
stated on this podcast that he has committed to shipping order of the sinking star in 2026.
And even if that's not entirely true, because traditionally, John has often slipped games from
like Christmas to the spring, like Brady anniversary and the witness, I think, both shipped in
like February, right? Just right after in that, because you often don't want to ship at Christmas.
So maybe we'll say this one's for Thecla Fiscal Year, 2026, which technically ends in like March
or something, right? So a year from March. But basically, 2026, at least Thecla Fiscal Year,
will be the time that I improve and correct that William Sonoma will fail to ship one box of
peppermint bark before John finishes the entire freaking game that took 10 years.
The peppermint bark trailer did go pretty hard at the game awards though.
That's a good point. Or order of the missing peppermint bark is the actual game name.
I thought for sure, and I came prepared. I thought you were going to be referencing the Apple
James, okay? I thought you were going to talk about how these are going to not be viral this year,
but I'm sporting one right now. Okay, so what? So teach I felt like I felt like, oh my god,
it's worth it. I thought it's worse than I thought. Wait, that's $300 or $100. How much is that?
I spent $149. That's like a $10 cost these goods. Casey, look, that's a shoelace.
First, the headphones cable taped to my phone. Okay, okay. I thought you bought it. I thought you actually
bought it. I bought that. I bought that. Okay. I hold out my daughter is sneaking in here a
whole moment. I thought you actually bought it. I was like, holy crap. I don't have an iPhone, do you?
I had to tape it. I have a green painter's tape today. It's invisible. So here's what I will
say about that. I definitely thought about that, but I felt like I couldn't use that as one of my
predictions because I'd already made it. Okay. I already said that it wouldn't happen like six
months ago whenever we talked about it. And so like, and we have a date on the calendar that we set
at that time. It was it was March or something. And so like when that comes around, believe me,
we're doing a stand-up on how nobody bought it. Okay, I'll be showing up with 10 of them on.
When I say nobody bought it, I mean besides the hardest core apple shells. Fred, this he's trapped
it. Still one of my all-time favorite episodes when trash realizes he's wearing like three different
grain in the ecosystem. I can't get out of stuck. He's still like as soon as we get off this,
he goes and puts on his vision pro to walk around the house. He's afraid to tell us,
but he just walks around the house with it. That's how I tell if my food's right or not through my
vision pro. Did you know I purchased the idea that that thing cost 150 bucks that you're wearing?
Yeah, I know, I know, I know. I believe that it's literally just that phone cable.
But wait, what the real thing is that? But you're saying that, but what's the different?
How is that actually different than the slate? Oh, that cable is more expensive than what Apple
had to make. It's not made from renewable resources in a carbon neutral way. It's a cable that
mother earth can't approve of. Yeah, exactly. That's the difference. The Apple one will be made from
entirely the finest spider silk woven in a magical factory that somehow has zero emissions.
All right, here's my here's my last one. I have a few rapid fire ones. I'd love to
yeah, you can rapid fire. We can do rapid fire afterwards if you wanted to. I didn't do my last one.
Didn't you just do your last one before? No, that was Casey. We're going back. It's
I think there will be a large, so we'll say greater than $1 million. Probably significantly more,
but at least $1 million pay out to a completely fixed price like polymarket prediction. So someone
will have complete control over the outcome and will just like put $500,000 on no and then not
doing it. And they're the one that's in control. And it will just be like, what can you do? You
predicted it. You're an idiot because they could just not say the word AGI in this press release
and they got a million dollars. Congrats. So I think there will be at least one greater than one
million dollar pay out. That is the person who fixed it is the person in charge of the prediction
who also takes all the money. So you're basically talking about politicians. Probably happening.
Like a boxing match. Yeah. Where they like where they rig the fight. Yeah, exactly. Are we talking about
Snake as again? Yeah. That's what you're talking about. Polymarket becoming like a like a old school
mafia like fixing kind of a market where they just create. They they have a finger in it.
Because it's not gambling. It's not going to be illegal to fix it or something. It'll just be like,
will so and so win a local election, right? And like for mayor. And then they will intentionally
like go cheat on their wife to lose it on purpose and be like, hey, I have to explore or something.
Lose the lose the rate. Yeah. Get five million dollars. Retire to, you know, somewhere far away.
Gotcha. That's actually an incredible idea. And it makes me want to run for office.
Like go out there and get a big party going. Prime greater than one million dollar pay out.
I have a better there's a better payout than that. If you created the show or movie or whatever it is
for Netflix, that is just that plot. A politician who is trying to lose to win on polymarket. But all
of the things they keep doing like cheating on their wife or whatever, keep bolstering their poll
numbers. That is like a surefire like, you know, comedy. Yeah. And it keeps trying to do stuff
like getting caught with drug dealers. But again, that like bolster is super cool and popular
or whatever. Right. And like it just keeps backfiring. And he's like trying so hard to lose.
So there you go. That's mine. Those are those are my top three. That's a good one. You may for all
we know, you may already have been correct. That could have happened. We just didn't hear about it.
So really twice the year we hear about it. Because it's a good idea. You think someone would try it
or whatever it was. The thing where they kept throwing the the lady pleasure device onto the
women's basketball court. I think there was like a polymarket bet of will a till they'll be thrown
over that. That's okay. But someone's like, yes. And it's me. I'll go. Oh, my God. So I know it's been
some amount of money. I just don't know to what extent has the money been. I want to try something
like that. So bad. Wait, what do you want to try so bad, trash? I just want to just a quick
clarification. I'll be above prime quick. Do your prediction. All right. I'm actually still
struggling to this point on which one I want to do because I have two highly competing ones.
And they're both extremely related to each other. They both involve cursor. Okay, they just do both.
I know, but it's really, it's really hard. So the first one here. I'll give the one that I'm
going to throw away, uh, which is after Microsoft loses the battle to GitHub, they acquire cursor
because they lost the editor war and lost GitHub. And so they rebuy it all back to win. I know,
I figured that would that that it feels like the best kind of two part series one right there.
But here's my alternative hypothesis. Is that right now super secretly? I think cursor is developing
a model that will not only rival, but beat cloud code as the premier coding model. And they're
going to go from one of these like, what is it? Level five providers of AI down to a level four
provider, which is they now are an open AI slash anthropic level company as opposed to someone that's
just redistributing these models. He's saying, you're saying opus, not cloud code. You're saying,
yeah, sorry. I said, I said, I, I said, Clod's coding ability. Yes. It's going to rival opus,
whatever their opus is. And they're going to be already have composer. Yeah, composers. They're
they're already trying this. Oh, it's not a secret. They released this already. I haven't been
keeping up. My prediction, mom, Donnie's probably going to win New York.
You should just kept the first one. I keep the first one. I keep the first one. Josh,
take it out. No, no, take it out. The battle can get hub end of VS code. It's own fork.
Found a company. Do you think that they'll finally release the third Avatar movie and you think
it'll do well at the box office? Yeah, that's my prediction. That's my prediction as well.
All right. For me, uh, Shalima is going to do a ping pong movie someday too. No way. No,
no, it's not. That's ridiculous. All right. Josh, give it in.
Zoom in on the right. It's unreal. How much did he represent this guy?
Josh, please zoom in on prime space. I think 2026.
Hi, Lair. Trash, you're right. I tried.
Yeah. So I'm usually asked to say, I was going to say that Microsoft really just guts their
their work and just replaces it with cursors team. What my backup one is,
Hey, Trash, to be clear, your prediction is they, they just like fire everyone from GitHub and just,
well, I mean, they, they, they ultimately acquire cursor, but like slowly, you know, how
which people start in an aqua hire, but not like as a company, they're just going to try and
poach everybody. No, no, they're going to take cursor over, but they're going to end up like
just gutting like the existing employees, placing them with a cursor. My backup one is there's
going to be another major, major outage in tech, like at the scale of Cloudflare, Amazon,
whatever, where the post mortem is going to be that they're going to say it was AI generated
code, and they don't actually understand what it said or what it did. Oh, that is a good one.
That isn't good one.
So there's only one reason I disagree with that. Come on, Casey, give me it. The reason I disagree
with that is because the only people who will have that outage have your 100% correct,
but they will have too much invested in having tried to convince you that AI coding doesn't suck.
Oh, that they will never tell us that that is what actually happened. It's actually one reason I
disagree with you is they have such a down that they have to buy Anthropic to fix it. Trash,
I actually had a similar one. I was going to say that Amazon through so many series, or
technically it was going to be Cloudflare because they've been down so many times this year.
Was that I was going to say one of the major services was going to get it up only a 3-9 rating.
They are going to be down for like 180 hours total.
We're there now. We call it a 9-5 rating in the business.
Oh, what? What? Do you guys, okay, so do you guys know what 5-5 you just said, but I will say,
do you know what 5-9s is? Yes. Yes. So, 5-9-9-9.
So, say 5-9s, I think. People say 5-9s, they mean 99.99% right or what?
Yes. It's the standard 3-9s. How many standard deviations is that, right? So in
traditionally in the industry when the thing sucks and never stays up, it's called 9-5s,
which is 55.555% chance that the thing is running. Got you. No one's ever heard this before.
Come on, guys. No, no, no, no, no, no, no, no, no, we're not trying to throw shade at anyone.
9-5s is great, okay? Everyone's trying their best out here.
Hey, man, it worked half the time. One point is better than a coin flip.
I mean, 55%. The thing I was working on right before I left Netflix was a automation framework,
and it had a 1-9 uptime. At least you had a 9-summer. We had a 9-d or 9-d or 9-d?
No, it's so 1-9 would be 90. So it was like 96% of the time it worked,
but that's a really high number for running tests on when you ran 1,000 tests. That's pretty good.
Trash, you probably remember that that every time you did a TVY PR, it was like 40 tests would
just fail, and you're like, I don't know why. And then you just reroute it again, you're like
four tests are probably running to the beginning, like they all passed. It's quite a problem over there.
Do we have any, do we have any like side quest predictions that are just like fun ones that you
wanted to get out there? Yeah, for another serious one. Oh, go ahead. Oh, me?
Well, yeah, we'll start with TJ, because he was the one that originally did it trash,
if he just rudely interrupted like that. It's about you, Trash. So I need this one,
I'm going to need help because I'm obviously not going to do the work for this, but I know that
there will be a few people in the audience who can help me out this. Trash will have greater
than 20 unique snacks on the standup this year. So please, someone do spreadsheet and keep
track of that for me, I will not be able to. I have a counter one. I have a counter one.
I was going to, so one of them was that trash will give up snacking. What? These are completely
opposite predictions again, exactly. Yeah, just stopping this is going to get some sort of horrendous
medical report that he doesn't want to share with people. And that's he'll be like not seeing salad
these days. Trash looks great, bro. You can't eat twigs all day. He's going to have like,
D he's going to have like a tooth problem or something. He's like, dude, bro, I got to stay off,
I got to lay off the twigs off guard. This is me, this is me laying off the snacks to be honest.
So here's my question about why I'm confused about all the like Microsoft GitHub co-pilot
and then like cursor by whatever things that I'm confused about. So from Microsoft's perspective,
and I don't really know because I don't know anyone in this org to ask, from Microsoft's
perspective, isn't GitHub just kind of unimportant to them? Like, why do they care about? I didn't
really understand why they bought it in the first place. But if they can end up in a situation
where somebody else like cursor has to manage all that stuff and just has to pay Microsoft
for all the cloud compute and whatever, isn't that just a straight win for Microsoft who then
doesn't have to actually spend resources maintaining a code base for an online source code repository
that currently probably just loses them money? Or do you think they're making enough from like fees
or private hosting to make it worth it? I'm assuming the enterprise plan goes crazy for them.
You do? Okay. That's my assumption.
Other thing, Casey, that's just like, I'm curious from hearing this get talked about. I mean,
I used to work at a company that was kind of competing against GitHub in some ways like at source
graph. We did like code search. We didn't do direct hosting, but we did a bunch of other features
that GitHub did. Yeah. Right, but we tried to like, you know, make them fast or like correct.
You know what I'm saying? Yeah. It's a lot of it's about just being able to vertically
integrate all their stuff, right? So it's like, oh, you, you host on this thing with your code here,
here's a one click way to like make this be on Azure with like, here's direct to get into VS code.
Okay. Right. Or like, here's a deep integration between VS code to GitHub to Azure to GitHub
actions to. And then like, you start at GitHub. Everyone's on GitHub. So then you're like,
oh, I click this thing and it's like open in VS code. Okay. Sure. And then not right. So it's not
nest is like, I think a lot of people like the training data. I think there's something there,
probably, although like, everyone can access that. It's been reordered to core AI though.
Right. Okay. So basically what you're saying is because I open in VS code, who cares?
Because again, VS code is free. But if what you're basically saying is, okay, if we own GitHub,
we can make it easier to deploy directly to Azure than to deploy directly to AWS. And that's
a good thing or whatever, because that it means that more people will use our cloud hosting service.
And that's what we actually care about, which is what they actually care about. So is it basically
that? Or do you actually think that they're making money on like, trash was saying the enterprise
plans are actually valuable? Yeah. Is that true? I think, I think they make money on GitHub.
Okay. I don't know. I haven't seen anything like in a while or anything for that. It's just from
what we heard, like competing in some of that same space. A lot of it was just like, hey, when we,
we do the Microsoft way of we vertically integrate all of these, we run any part we need to
add a loss at any time to force out all the competitors. And then we try and lock enterprise people
in for a five year contract, make it impossible to get off and charge them of certain amounts of money.
Right. And it's like, okay, check our balance sheet, what we make quarterly at Microsoft. And it's
like, yeah, we'll just we can we can do the GitHub thing. Like it's important, though, because then
now we can make it easier to do all of this other stuff. Same thing with VS code, like the point
of VS code is so that they can make it so that it's like, you get to GitHub, you get to add. Yeah.
Yeah. And presumably they will also then be able to sell the AI like if they owned something like
like cursor, they could then charge for the AI compute time, which is also what they want. Right. Yeah.
And I don't know how much that's changed. I mean, it's been kind of a while since I like, you know,
maybe it's much more now about the data than it was before. Right. Right. Right.
But it's also about just being like in the spot to be like, they're going to suggest co-pilot
as the thing. They're going to turn on co-pilot by default on your repos. And then that's what
people are going to think, Oh, I need co-pilot. That's AI review or AI completion or whatever.
And then boom, now the company uses it. No one gets fired for picking Microsoft
previously. Maybe soon they will start getting fired for picking Microsoft or whatever. But then
that's like, that's like how teams is like what? How much bigger than Slack? It's huge. It's like
double the size or something. Yeah. It's like similar situation. Right. I think. So that's my
understanding. I mean, I'm totally on board with that. I just wanted to understand it because I'm
just like, why do they care? But yeah, if it's like, if those are actually good fees for them,
and all that, and I could see, I could see it being like, if people, if everyone wants to use
something like cursor as their IDE, it does, I guess, if Microsoft is like, we want to make a lot of
money off of something like co-pilot. If cursor just decides not to provide co-pilot as an option,
then that's kind of the end of it for them, right? So I can see why that would be valuable to make
sure that you kind of own the thing. That's the front end. I did have one side, one more side one.
I'll make one more side one. Is I think that somebody will release such an egregiously poorly
vibe coded application that a one or more founders slash C suite will go to jail.
Go to jail. Yeah, due to like the type of data breach and the type of data leak that ends up
getting caused or the type of harm that gets caused by it, that they are, because at some point,
if you hurt somebody or you make material losses, you can actually find yourself in more than
just a civil suit. And so I think that there will be some sort of damages so big that they will
be thrown into some sort of jail. Now, this may be a 2027 prediction. Maybe I'm a little off on
this one, but I do think we'll see that. And this was kind of birthing my underlying conspiracy
theory, which is that I think we will hit some sort of qualification or push for a software
engineering. And the AIs will be the ones that have this qualified positioning to like alleviate
you so you can't be held responsible for the code you release. That will be over multiple years,
though. But I think the first step is that I think people get arrested and held countable for
the code they wrote. That's that one. The 2026 someone will get arrested. Has that ever happened
before, because this would be a gross negligence standard, I assume, because basically like presumably,
you're not talking about someone who intended to do something. Yeah, because like if you're just
talking about an AI, unless you're talking about, are you talking about intentional, meaning someone
told the AI to do something nefarious, you're talking about saying where they accidentally like,
they asked to do something that they actually were trying to sell and that is not nefarious,
but it was so poorly executed by the AI that, you know, somebody got killed or something like that.
Yeah. Yeah. I'm calling in the second one. They get they get held like manslaughter or some sort of
the delete button does not actually delete data. They go to jail in the EU. Easy.
True. I mean, there are, yeah, I don't know. I mean, there are fraud cases, but they're usually
you have to be intentional. Like Elizabeth Holmes is the person that's coming to mind. I try to think
of like C-suite people, Jeffrey Skilling, Elizabeth Holmes, thinking of people who have actually
gone to jail. It's rarely because of negligence though. So like, I feel like, I feel like that's a
that's a tough one prime. I feel like you're you're on an uphill battle on that one because they
I agree that they should go to jail a lot of times though. So I don't know.
All right. Is it far one? That's why I did it. It's a tough one. That's a tough one. My actual list,
but it kind of is pretty sensational. I got one. All right. Someone, someone actually legally marries
in an AI. Oh, which country? Yeah, where? America. America. So first, one of the 50 stages legalizes,
legalizes AI marriage, one of the 50 states, which which state would it be? California.
Yeah. I know. I'm a let's say real here. California. California. 100% yeah.
Specifically San Francisco. Right. And then Mississippi will lower the legally required age
of the AI to 14. Oh my god. Yeah. I didn't even think about that. I just want the state that
I did not approve. My next one, I don't know if it's going to happen anymore. The stand-up live
in a theater or big venue for 2026. So then I don't know. But then yeah, you're going to have to
cut out the like legally marrying an AI. I apologize to Mississippi. No hard feelings. Dude,
she's 4,000 years old. Okay. My prediction is ruined. TJ is actually upset right now.
He's going to take his apple. I'm taking my toys away from you guys. Okay. I wish you could
be this cool. Well, your prediction was we're in a theater. Is that what you said? Yeah, we'll do
an episode like live, but like add a like cool venue or something. We should do it. We should go
to Nashville and do it in a bar. Like Nashville is super cool and people show up to those kind of
events. And they do like shows in bars. Okay. Is there a lot of coders in Nashville? No. Yeah. They
also have singers. I know, but I want to do something fun. You know, like do something that's
like unusual. I'm not like we're in GitHub's old office before they got acquired by hers or
okay. Yeah. Okay. Whatever. I mean, I see what you're saying. The question though, I think it's
fairly important to pick someplace where you could actually sell tickets to it, right? And if you
pick someplace that doesn't have much of a, you know, presence of people who would watch the show,
that makes it hard to do it in San Francisco. Okay. I just want the idea. We could also do it
in New York. We're gonna extra that us San Francisco. Yeah. Yeah. That's a good point.
Wait, people throw eggs. Oh, yeah. They do have a lot of wealth inequality there. So they've
probably thrown this high cost high cost items at people. I'm catching them, bro. I'm like,
dude, this is free money. Thank you. Only the freest range eggs. They were like, we don't even
have eggs like it's grown from. Yeah. Each chicken gets its own like one quarter acre plot
and and and it all and a little like housing. It's like a way to subsidize housing and a little
like a little place where it can get avocado toast in the morning every day. It pecs at the avocado
toast like to get the, you know what I mean? It's my chicken drinking. I only like Ethiopian roast.
Yes. Absolutely. Absolutely. What is this little timer for? Oh, it's to make sure that the chicken's
tea has steeped the exact right amount of time before it's too hot. Yeah. Someone saying would
be doing the stand-up as a session as a tech cough count as doing it in the theater. Like,
you know, like for React Miami, if React Miami, we said, hey, we're gonna go there and do something
or for some conference. Well, Teage definitely wouldn't, his prediction would be at least half
right. Like he couldn't be accused of being wrong, but he might not be completely right.
Because you're not looking more clear. We have a stand alone.
Just the latter. It's not part of it's not under the auspices of anything else. We just say,
here's a date you guys can come and we do and we record one with these four people. So Casey has
to be there in person too. All right. What about is there going to be an opening act or is it just
going to be a singular thing of standard? Oh, I'm playing. We're having a position. So probably
it's going to start with a juggler. He's going to probably start with three balls. Then he's going
to go four. Then he's going to go five. And he's going to touch me. Six seven. Oh, no. I hate
all right. It's not an episode. That's the end of the episode. Dude, my son's clothes are
size six to seven. He's like, I wear a six seven. I'm like, oh my god.
That was good. That was actually good. Oh, my goodness. I hate. I hate six seven. But that was
that again. Did you see a comment? Did you see a comment at any point or no? No, I just know
why you planned that. There's no way he planned that during the juggling joke. It came to him
while he was counting. It had to. There's no way. I feel like the opening act is teach playing
the marimba. Let's be straight. It's he, like, just plays the arimba for a little while. Like
15 minutes of marimba to like get the room set up. If we do one in person, we're obviously going
to play a song with us and some maybe other friends too. Like we'll do some other fun parts of it.
That would be exciting. Can I do my magic card trick that I wanted to do the last time we didn't
have? Yeah. Yeah. You can do the magic card trick. You want to do a family feud. We'll tell
you about it. It's so funny. It's so funny. I have worked on this magic trick card like for a
decade. I have one magic trick. And that is it. Okay. We can't tell you anymore on air because
it gives you the spoiler of the secrets. We'll tell you after. Wait, what was the trick we did at
Theo's Place Prime? Where everyone was like, how do you do that? He was like David Kramer and
everybody. I can't, okay, we can't talk about it. Let's just be quiet. Oh, is that the trick?
I don't know if I've done it before, but I just don't want to talk about it. I think this has
to be the end. Hey, thank you very much for joining us on the extended edition of this. I hope
that you enjoyed it. And also, thanks for hanging out with us for 2025. All right, this is the last
one. So thank you. I hope hopefully everyone had a no blockers today. Thank you for joining us.
Have a good one, everyone. Bye.
Bye.
All right.



