2026-03-14 20:43:11
Last July I wrote a post entitled "We're Seemingly Still in the 'Throw Money At It' AI Era..." The crux was that given Elon Musk's ability to get xAI's Grok models up to technical speed and the fact that Mark Zuckerberg was completely rebooting Meta's AI efforts seemingly showed that all it took to catch up in AI maybe the ability to spend massive amounts of money, fast.
Well, as it turns out, that may not be all it takes...
Fast-forward to this week and we have the same two companies back in the news around AI. This time, it's Musk rebooting the xAI team while Meta seemingly struggles to get their new cutting-edge models up to speed. Naturally, that makes it seem like all that money spent was actually not worth it – at least not yet. And that there's clearly something else that's crucial when it comes to competing in AI.
As it happens, this was also my thesis in a post I wrote just before the one above. Trying to think through the ramifications of Meta's wild "Godfather" offers to the top talent around the AI industry (making them offers they couldn't refuse – though actually, many did), my question was simple: "Can Meta Buy the Future of AI?" And my subhead gave my conclusion quite simply: "No."
My read on the situation was that this strategy wouldn't be enough for Meta to truly get back into the race. And worse, it could backfire.1
Both OpenAI and Anthropic had seemingly built more cohesive teams centered around a mission from day one. Meta was effectively hiring mercenaries, trying to pick off the best of the best to jump ahead. Selection bias aside (perhaps the "best" people also were the most loyal to the actual mission, not just the money), it felt like the problem that many sports teams run into when they try to put together "Super Teams". That is, they build a team that is really just a collection of individual talent. Yes, that talent looks great on paper – perhaps even superb – but there are intangibles that make for a great team.
A few times, those superstars have gelled into that great team – think: the LeBron/Wade/Bosh Miami Heat – but it took some time. Even for the best players in the world. And most of the time, those trying to execute such a strategy don't give the team enough time to cohere. They're paying all that money to win now.
Anyway, it's still a bit early to write-off Meta's new efforts as we've yet to see the actual fruits of the labor. Fruits which happened to be named after fruit. But the fact that we've yet to see those fruits itself suggests that things are not going as planned – or at least hoped. Here's Eli Tan for The New York Times:
Mark Zuckerberg, the chief executive of Meta, said in July that his company’s new artificial intelligence models would “push the frontier in the next year or so.”
Now Mr. Zuckerberg — who has invested billions in the A.I. race — appears increasingly unlikely to hit that deadline, three people with knowledge of the matter said.
Meta’s new foundational A.I. model, which the company has been working on for months, has fallen short of the performance of leading A.I. models from rivals like Google, OpenAI and Anthropic on internal tests for reasoning, coding and writing, said the people, who were not authorized to speak publicly about confidential matters.
The "fallen short of the performance of leading A.I. models" is exactly what doomed Meta's first attempts in AI. Despite billions spent, their initial Llama models could not match up to the competition. This may have been a strategic blunder in focusing on "open". Or it may have been more granularly a mixture-of-experts (MoE) mistake. Or a people problem. Or, most likely, all of the above.
And that's why Zuckerberg blew up the effort, spending a mere $15B on a "hackquisition" – and billions more on other deals and comp (not the mention the tens of billions spent on CapEx) to reboot the effort. The people may be different. The models may be closed. But the song seemingly remains the same.
Worse than that:
The model, code-named Avocado, outperformed Meta’s previous A.I. model and did better than Google’s Gemini 2.5 model from March, two of the people said. But it has not performed as strongly as Gemini 3.0 from November, they said.
As a result, Meta has delayed Avocado’s release to at least May from this month, the people said. They added that the leaders of Meta’s A.I. division had instead discussed temporarily licensing Gemini to power the company’s A.I. products, though no decisions have been reached.
I mean if Meta just spent billions to try to beat Gemini only to have to spend billions on Gemini... that's not a good look. At least Apple didn't have to suffer through the billions spent competing! They simply whistled while they worked on the wrong things while convincing themselves that everyone was wrong to spend the billions in AI.
Well, maybe some of them were wrong. Though seemingly not Google! As Apple is now spending billions for their tech.
This was clearly the right call for Apple. For Meta, it would be far more embarrassing, even if it's still the right call. Like Apple, it would buy them time to get their own AI house in order. The difference, again, is that Meta is spending hundreds of billions on those efforts. Apple is spending something closer to zero (which to be clear, may be more insane – time will tell).
Anyway, this isn't about Apple (for once), this is about Meta. And xAI.
I had quickly second-guessed my post about Meta's AI rebooting strategy not working because it seemed like it might have worked for xAI last year. Musk was able to throw money at the problem – specifically at the data center problem, getting his 'Colossus' cluster up and running in record time. This allowed his team to build Grok models which looked competitive with the leading models in a relative instant.
But looks may have also been deceiving there. While Grok may have looked good on paper, as a product, it simply wasn't getting the traction of ChatGPT or Gemini. And that was despite merging with Xitter and baking it directly into a product that millions use. Worse, as a business, Grok clearly couldn't hold a candle to the leaders. That's undoubtedly a big part of why the just-merged xAI was further merged with um, SpaceX. xAI was burning an insane amount of money with little to show for it. But as a small part of a bigger (and better) picture, they could suddenly afford (figuratively, not literally) to show far less while still being able to stay afloat.
Which is to say, "throwing money at it" seemingly didn't work for Musk either. While it may have gotten the infrastructure in place, clearly xAI was lacking something else. To hear Musk tell it now, it was the team: "xAI was not built right first time around, so is being rebuilt from the foundations up."
I would suggest that this is perhaps most damning of the investors who poured billions into xAI, but come on, clearly they were just betting on Musk. And it paid off, quite literally, in the form of SpaceX stock.
Anyway, Musk's brutal assessment of his team was apparently triggered by the fact that Grok cannot compete in coding with the likes of Anthropic and OpenAI (not to mention the myriad "smaller" AI startups building in this space). And so he's down to just two of the people he founded xAI alongside, with eight – eight! – having left in just a few short years. It's enough to make OpenAI – where Musk himself was famously one of the many co-founders who left – look downright stable.
The fact that this is going down so soon after the SpaceX merger certainly paints that maneuver in a different light. At the same time, the high-level assessment is the same: xAI wasn't going to be able to compete without some sort of major shake up.
And so here we are. With Musk out there now trying to shake up other companies, seemingly running a playbook similar to the one Zuckerberg just ran! Poaching people from Cursor and Thinking Machines Lab (which, of course, is in the midst of their own turmoil). And undoubtedly OpenAI and Google. And just as Zuck and others have learned, probably not Anthropic.
So will this work? Musk is trying to calm the turbulent waters by noting that this is the same thing that happened with Tesla. A company which famously nearly died. But Musk pulled them back from the brink. And now he's going to try to do so with xAI, but this time he has the warm blanket of SpaceX for protection. How SpaceX investors feel about that may be another matter. But the public markets will probably soothe that over soon enough too...
Oh and what's that in the rearview mirror? It's Tesla! Zooming in to save "Macrohard" with "Digital Optimus" or something.
Meanwhile, Zuckerberg is now trying the Musk playbook! Forget the relatively small people spend, he's betting his entire company on the infrastructure build out. Oh, a new data center project just opened up with others worried about cost (and/or debt)? Where do we sign?
At the end of the day, to compete you do have to actually compete. With the best models, products, and business. xAI and Meta remain in the same camp there: on the outside looking in, despite billions upon billions spent. What actually changes such equations? "Superintelligence"? Maybe? "Personal Superintelligence"? Probably not? Bot social networks? Nah. Agentic orchestration? Could be, but good luck there. "World Models"? Well, Zuckerberg just shoved those aside. War? Sure! Robots? I mean, maybe in a decade. Data centers in space? I mean, maybe in a couple decades. But this race is happening right now. And time may be running out...
What's the strategy to get back into the race? Money doesn't seem to be enough...
1 There has seemingly been some indications of how the strategy has backfired internally. And other, more damning reads on the current situation – which, to be clear, Meta has shot down – in totally normal fashion. There are may be new mass layoffs on the verge of happening. Again... ↩
2026-03-13 18:41:08
The biggest question I had about the forthcoming 'iPhone Fold' is which OS it would run. Given that it's still an iPhone, you'd assume iOS. But that would presumably require some new flavor of iOS given the whole foldable screen element. That mixed with the fact that it sure looked like the interior screen, when unfolded, would be similar to an iPad mini led me to wonder if we might get a true hybrid: iOS on the outside, iPadOS on the inside. A sort of mullet device.
Not so, says Mark Gurman in his latest report about the device for Bloomberg:
Apple is developing new iOS app layouts and revamping its core iPhone programs to add sidebars along the left edge of the screen, similar to many of its iPad apps. Developers will also be able to adapt their iPhone software for the new interface, which will use proportions similar to an iPad in landscape mode.
Despite offering an iPad-like app experience, the foldable iPhone will run the standard iOS — not iPadOS, the company’s tablet operating system. This means it will retain a simpler multitasking system, rather than adopting the more desktop-like interface introduced in iPadOS 26. It also won’t run existing iPad apps out of the box.
While the foldable iPhone won’t run several windows at once like an iPad mini, it will be able to show two apps side by side. That matches a key feature of the foldable phones offered by Samsung, Alphabet Inc.’s Google and others.
So the experience will be "iPad-like" but not actually iPadOS. Instead, we get a version of iOS that looks more like iPadOS, but without the true windowed multitasking (though we can argue about how "true" that system is, even now). This sounds potentially confusing for iPad owners – in particular if the inner screen truly is iPad mini-like in size – if it looks like an iPad, but can't do iPad things. But it undoubtedly would have been more confusing to run two different OSes with different functionalities, simply depending on if the device was folded or not. And there's no real point in running standard iOS (just scaled to a new screen size) without taking advantage of the new screen real estate in some way. So this sounds like the compromise: iPadOS-looking iOS.
I'm just as curious about the actual form-factor here, because it continues to sound – and look, via CAD leaks! – like it will be quite different from other foldables to date. I've had a few versions of the Pixel Fold now, which I like and has gotten better with each iteration, but it's really just exactly what it sounds like: a fairly standard smartphone that can unfold into a bigger-screen smartphone. Yes, it's interesting that it's a more square-like experience inside – and yes, it can do the app side-by-side thing – but it still just feels like a smartphone. I think there's a pretty good chance that the 'iPhone Fold' (if Apple even calls it that – all bets are off after the wildcard 'MacBook Neo' naming scheme) will actually feel like something a bit different than just a "foldable iPhone" given the form factor.
In my mind at least, it almost seems more like a 'iPhone mini' on the outside, while yes, an 'iPad mini' on the inside. But this new iOS will be key to the experience, of course. Does Apple tailor it to leverage those new screens in ways that differentiate it from a "regular" iPhone? I suspect they might! Maybe that front screen becomes your sort of "vital" apps/widgets screen. Things you always open and/or apps for fast actions. While inside, that's where you do your real "work" (or play, I suppose!).
I obviously won't even pretend that I'm torn on getting one. I'll certainly have to to see what it's like. I am a bit torn on leaving the old, trusty "standard" iPhone form-factor behind, but would I really "dual-wield" like I'm Ahsoka or something? That's a pretty expensive proposition!
Speaking of, there's also some new leaks around potential pricing for the device. Are you ready? Take a deep breath...
According to the Weibo-based leaker Instant Digital – who had a strong track-record on such things – the range will be from $2,320 to $2,900, depending on storage options (from 256GB up to 1TB). That's, um, quite a bit above my early (and obvious) $1,999 starting point guess I made a year ago.
Of course, those numbers are extrapolated from the more rounded prices in Chinese Yuan by the leaker. When I asked AI to extrapolate how those implied numbers might translate into USD, sure enough, the $1,999 starting price point seems quite likely. Then again, given everything going on with chip shortages and pricing issues around memory... who knows. Perhaps not even Apple just yet.
Would anyone buy a $1,999 iPhone? I mean, yes, it's Apple.
And some already do, as Apple hit that mark with the 2TB iPhone 17 Pro Max option. What about a fully-loaded 'iPhone Fold' at, say, $2,999? It wasn't that long ago that we were debating the viability of a $1,000 iPhone. That seems quaint now. The iPhone Pro models are huge hits for Apple. There's undoubtedly still room to grow – and importantly for Apple, grow that ASP.
One more thing: As I wondered in my prediction post for the year, what if any sort of newfangled experience for the 'iPhone Fold' also included a new type of accessory? Maybe an 'Apple Pencil Mini' – heaven forbid, the actual 'Apple Stylus'?!
M.G. Siegler
M.G. Siegler
M.G. Siegler
2026-03-13 07:21:38
One obvious side effect of Netflix's (ultimately failed) bid to buy Warner Bros? A revisiting of the discussion about theatrical windows. Given that Netflix was promising 45-days for Warner's movie slate going forward, well, if you give a mouse a cookie... they're going to want a cow. As Brooks Barnes writes in The New York Times:
Universal said on Thursday that it would immediately begin guaranteeing theaters a minimum of five weekends of exclusive play for new movies, ending a pandemic-era policy that guaranteed only three. Starting in January, Universal will move to a minimum of seven weekends of guaranteed exclusivity.
Yes, that's a move from 3 weekends to 5 weekends immediately. With a move all the way to 7 weekends – roughly the 45 days guarantee Netflix was making – early next year. We are so back... to where we started.
Hollywood will say – and is saying – this is just a correction for the mistake made during the pandemic. Almost exactly six years ago, with the world in lock-down mode, it was none other than Universal that took a hammer to the theatrical window. They announced that Trolls World Tour would be released for rent at home the same day it launched in theaters – if theaters were even open by then. Other studios followed suit, perhaps most famously Warner Bros, which went Full Monty, much to the chagrin of many filmmakers. This seemingly destroyed Christopher Nolan's relationship with the studio, over Tenet – a movie which was just as audibly indecipherable at home as it would have been in a movie theater. Even Disney was forced to follow suit and as such, faced a suit from Scarlett Johansson, which helped end the brief Bob Chapek-era at the studio. It was a rocky time, to say the least.
This also put Universal at war with AMC. Once the head of the largest theater chain, Adam Aron, read what then-NBCUniversal CEO Jeff Shell – now, incidentally the president at Paramount Skydance (at least for now) – said about the success of Trolls on on PVOD after their decision, he went nuclear:
"It is disappointing to us, but Jeff’s comments as to Universal’s unilateral actions and intentions have left us with no choice. Therefore, effective immediately AMC will no longer play any Universal movies in any of our theaters in the United States, Europe or the Middle East."
People may have been a bit on edge... But eventually, even AMC had to recognize the reality of the situation, and suddenly we had a de-facto 17-day window.
Anyway, reading AMC's comments today, and you wouldn't know any of that happened:
AMC Entertainment, the country’s largest theater operator, called Universal’s shift in strategy “extraordinarily beneficial” in a statement, adding that it “strengthens the entire theatrical ecosystem.”
Here's the thing, it did happen. And while it may have been a bit of an overreaction in hindsight, it also should have served as a good wake up call for Hollywood. The theatrical model had been slowly dying for years, but it was masked by inflation. The industry kept touting ever-growing box office numbers without any actual context. Had they simply reported numbers of butts in seats, the numbers would have looked bad (and worse if you cut it by per-capita moviegoing). COVID, as it did in so many industries, simply accelerated a trend that was already in process.
And now we're back. Having learned seemingly nothing!
The main problem for Hollywood is that the box office still isn't nearly back to those pre-pandemic numbers, even when masked by inflation. And again, they'd argue the windows are the issue. To be clear, it will undoubtedly help a bit – obviously, they're creating a faux supply/demand situation, propping up a marketplace.1 But it's not addressing the underlying issue: seeing movies in a movie theater doesn't have the hold over the general public as it once did. And it never will again. Windows or not.
That's not to say movies – or Hollywood itself – is dead. In many ways, movies are more alive than ever. But that's thanks to streaming – and yes, smartphones (sorry, Mr. Nolan). We used to have to watch a movie in a movie theater in order to see it. Now we can watch it basically anywhere. To pretty much anyone that would look like progress. But not to Hollywood. And certainly not to theater owners.
But actually, buried a bit in this news is an element that suggests Hollywoods' heads might not be as buried in the sand as it seems...
One important wrinkle: Universal’s corporate sibling, Focus Features, which makes smaller-budget specialty films like “Hamnet” and “Bugonia,” will not change its theatrical exclusivity policy. Three weekends, or about 17 days, is all that Focus will continue to promise.
NBCUniversal said specialty films — one of the most challenged genres at the box office — required special handling. For some of these movies, a theatrical release has become valuable mostly as a marketing tool for what Hollywood calls “premium video on demand,” the sales window that immediately follows theatrical exclusivity. Focus’s movies also tend to open in a handful of theaters at first, expanding gradually to more markets as word-of-mouth demand builds.
“Universal remains a theatrical-first studio,” Ms. Langley said. “That’s proven by the breadth of our slate, our commitment to our filmmakers and the ongoing investments we make in the creative community.”
In other words, they're acknowledging that all movies are not created equal. Actually, that's not fair. The movies themselves may be, but the natural audiences for those movies is not. That's not harsh, that's reality. And again, it has been reality for years and years and years. In the olden days, people would go to the movies to see basically anything because they had little else in terms of choice. These days, we all have the opposite problem. As a result, only a handful of movies really warrant a long theatrical run. It's not a comment on quality or anything else, it's just reality.
It's like the age-old epistemological question: if a studio put a movie in theaters and no one goes to see it, did it actually play? These days, we can ensure that the movie plays in a place and in a way that people will actually see it. We have the technology.
Well, technically, Netflix has the best of that technology. But Hollywood didn't want to leverage it to the fullest extent – including, notably, helping with theatrical. In hindsight, I believe this will look incredibly stupid. But I get it. The key now is the acknowledgment that this new/old seven-week window isn't some sort of commandment handed down from the old gods of Hollywood, it's simply one path for a certain type of movie.
That's how I choose to read Donna Langley's statement. That she knows every movie can't do seven weeks in theaters. Obviously. What if it bombs opening weekend? Do you let the stink linger for six more weeks? Of course not. This is simply a guidance. And a game plan for the "tentpole" movies which will launch under the Universal banner. Other, smaller films will likely go through sub-studios, like Focus. This is a good first step for every studio.
And, by the way, Netflix should still come at this from the opposite end: if a movie is likely to do well in theaters, Netflix should push it wide (though yes, they'll have to partner to do that now without acquiring Warner's business). I've been saying this long before they tried to buy into theatrical. Now that they aren't, it's still the same rationale. I mean, maybe not seven week theatrical runs now that Ted Sarandos doesn't have to be held to such standards. But still, something. Where it makes sense!
And where it doesn't, for all studios, they should go to PVOD right away! Or streaming! It can even salvage clunkers!
This is all so obvious, and yet Hollywood fights it. Because it's not the way of the past, it's the way of the future. Netflix could have shown them the way more clearly, now they'll have to do it themselves. But it's the same path.
Fine, fix the window. Just fix the house too!
1 I also do think that 2026 is primed to be a good year at the box office just by nature of the slate. Hollywood will do a victory lap around this as if it wasn't mostly about timing of some blockbusters aligning of course... ↩
2026-03-12 21:00:19
"I'm here in case you succeed."
That's what Colonel Jack O'Neill (Kurt Russell) tells Dr. Catherine Langford (Viveca Lindfors) when she asks why he's a part of the 'Stargate' mission in the 1994 film of the same name. The implication, of course, is that in the unlikely event that the Stargate works, and opens a portal to another world, he's the fall-back option in case something goes wrong.
Well, the stakes are lower, but the tech project named after the movie – the massive AI data center build-out spearheaded by OpenAI, Oracle, and SoftBank – is no less complicated. Case in point: the many fits and starts of the real life initiative. And the latest twist and turn perhaps points to a very fundamental challenge of the project – and the broader data center build-outs in general.
The latest news dropped last week. Last Thursday, Bloomberg reported that OpenAI was backing out of a planned expansion of a data center project in Abilene, Texas. If you've heard of Abilene in recent months it's likely one of two things: either you've been listening to a lot of Ella Langley, or you've been following the aforementioned 'Stargate' project – that's the location of the first such site...
2026-03-11 23:15:09
It used to be the case that just about the hardest thing in the world was to get someone to download a new app. But in 2026, given the connection speeds we all enjoy, the actual downloading process is negligible. The hardest thing now is getting them to actually use the app. Such is the story with Sora.
As Stephanie Palazzolo and Sri Muppidi of The Information report:
OpenAI plans to soon launch its Sora AI-generated video capabilities in ChatGPT, according to people with knowledge of the effort. The strategy shift could increase the chatbot’s usage as well as its costs.
The move comes about five months after OpenAI released a Sora mobile app, which lets people quickly create videos of themselves and other people or animals doing mesmerizing or humorous acts. The app’s look and feel mimics the style of TikTok but doesn’t appear to have gained wide appeal.
Well, it did gain wide appeal, reaching the top overall position in the App Store and lingering in the top 10 for quite some time. The problem was staying there. And that's in no small part because the millions who originally downloaded the app largely stopped using it. At least regularly. I should know, I'm one of those users.
To be clear, when I do find myself opening the app every once in a while, I still get sucked in. There is a lot of content that ranges from amusing to hysterical – and from creative to copyrighted – but there's also less of it now, at least from people I know.1 And unlike TikTok, it was set up with some core social components in mind, notably through "cameos" (the problematic feature name due to its own trademark issues!). And whereas Sam Altman himself humorously filled any social voids in the early days of the product, now it feels more like nth-tier celebrities trying to latch on to an early movement.
Again, the movement has largely stopped, at least in that the app is no longer "breaking out" as it once was, and as such, doesn't have much of a shot at another mainstream moment. At least not naturally. If and when Disney content gets added into the mix, there will likely be another viral swarm back to the app. But unless OpenAI can keep that up at a regular cadence, they're probably going to need another answer here.
And that answer is clearly to bundle it back into ChatGPT itself.
Yes, yes, we're right back to the "bundle and unbundle" quote that I turn to so often. But here, the unique dynamic with OpenAI is that they're constantly bundling and unbundling at the same time. And I think it's probably the right strategy, because it's still the early days of AI and no one is quite sure what will work and what will not. Success is often stumbled into, from ChatGPT on down.
That's the thing, without question, ChatGPT is that mainstream hit. As it nears a billion active users, the service has become synonymous with AI itself. If you asked my older daughter about AI, it would be the first thing she mentions. I asked her about Gemini and Claude the other day, she didn't know what those were. But she's now constantly asking me to "ask ChatGPT" something (as she's too young to have access on her own). Despite the comically clunky name, it has become the "Kleenex" of the industry, at least for now. And that's brand value that money cannot buy.
So of course, as a company, you want to leverage that. I still think it was the right call to launch Sora as a stand-alone app to start. And I think the early viral success bears that out. But that also doesn't mean you can't pivot if and when that virality flames out.
If it's the case that the real value of Sora is less a TikTok-style feed and more about sharing amongst smaller social circles, you obviously want to tweak the app to ease that, as OpenAI has. But what if it's actually more about creation itself? At that point, you might pivot again to move it back within ChatGPT.
As the article notes, this may also be an attempt by OpenAI to boost ChatGPT above that elusive 1B WAU mark. But that doesn't mean it's the wrong thing to do and try. Especially as ChatGPT itself continues to morph from a simple chat interface into something more robust.
At the same time, my sense is that the group-chat element of ChatGPT itself isn't really working in a major way. That may be based on my own anecdata, but the fact that OpenAI isn't touting it constantly seemingly suggests it's not some huge hit – in the way that say, early Sora was.2 The answer here may be the opposite: to give the group chat feature its own app/service. That alone won't be enough, of course – again, it's hard to get people to download let alone use new apps – but OpenAI is great at coming up with these little product moments to spur usage. They could and perhaps should try that with group chats. Because I do think there's something that could end up important there. I mean, "chat" is right there in the name!
And you know what could pair well with Group ChatGPT? Sora.
Meanwhile, all eyes and hands internally would appear to be on Codex at the moment, as OpenAI tries to combat the rise of Claude Code (and Cursor, etc). Maxwell Zeff has a big story about that effort that just hit in Wired. It's a good read about how OpenAI missed the initial opportunity and what they're doing now to try to correct that. One of the big high-level takeaways would seem to be that many inside the company assumed that coding would just be a part of ChatGPT, and not necessarily its own product.
And that may or may not have been colored by the fact that Microsoft was so heavily using their models for coding within GitHub. Since their (at the time) core partner (and benefactor!) was tackling that use-case head-on, OpenAI focused on the myriad other things they wanted to work on. And by the time they recognized the mistake, and tried to buy Windsurf (after Cursor turned down their overtures), Microsoft stepped in to scuttle that opportunity. Ouch.
Anyway, now there's Codex, which is right there, front-and-center in ChatGPT's sidebar. But clicking it takes you to yes, another app.
At the same time, OpenAI is hustling to sell Codex into enterprises. And, unsurprisingly, they're using their own strength to make inroads:
Several major enterprises have signed on to use Codex too. “The fact that ChatGPT is synonymous with AI gives us a massive advantage in the B2B market,” says Fidji Simo, OpenAI’s CEO of applications. “Companies want to use technologies their workers are already familiar with.” OpenAI’s strategy to sell Codex is largely based on packaging it in with ChatGPT and other OpenAI products, Simo said.
Ah yes, the old bundling playbook! Deemed illegal at times for companies like Microsoft because it's so effective! Next up? The government!
Another product that's rather oddly unbundled at the moment? Translation. You'd think that would be an easy one to include in the ChatGPT bundle, but OpenAI is clearly using it as lead gen. As a wedge to expose new customers to the magic of AI.
Health could be next. Simo has already talked about the high-level strategy here of making it a part of ChatGPT to start (in its own area). But don't be surprised if they try it as its own service/app at some point. Obviously, it will depend on how it works within the broader app and what features resonate with users.
Don't forget apps! As in third-party services that are bundled to run within ChatGPT. OpenAI tried an 'app store' approach before for GPTs that didn't really work. And this doesn't seem to be working so well either, thus far.
All of this points to the obvious downside of bundling everything within one app – it gets crowded, and thus confusing, fast! OpenAI has already shown some Microsoftian tendencies with their model drop-down menus – since alleviated, thankfully – and cramming a ton of features into a Frankenstein's Monster-like product experience is obviously the fear here. At the same time, you have the most downloaded and most used AI app in the world. You should use that to your advantage, right? (Especially when you're trying to launch ads within that app...)
Again, I think the answer and strategy here has to be trying a bunch of different things. Sora as a stand-alone app, Sora within ChatGPT. Codex as its own app, Codex more tightly integrated. Group Chats within ChatGPT, Group Chats perhaps outside. Etc.
And this experimentation will necessarily need to expand as OpenAI breaks into actual hardware. They've tweaked voice mode within ChatGPT a few times, but when it's the main method of interaction with any newfangled device, how does that change the product and UI of ChatGPT itself?
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
1 I might also note that it's still only available in a handful of countries – one is not even the UK yet... ↩
2 Ditto with Atlas, the web browser launched to much fanfare. They've been awfully quiet about it of late... ↩
2026-03-10 23:21:01
Look, do I realistically think Apple is going to buy Anthropic? I mean, no. It would be by far the largest M&A deal of all time – undoubtedly more than double the next largest deal,1 and certainly more than triple the recently insanely large (and insanely leveraged) Paramount Skydance/Warner Bros Discovery deal. Anthropic is currently valued at $380B. Apple's largest deal ever was two orders of magnitude below that. And that deal, Beats, was done over a decade ago.2 This is simply not a deal that Apple – or probably anyone – does in a normal environment.
But we're not in a normal environment...