2026-01-13 01:48:08
No surprise, but now it's official:
Apple is joining forces with Google to power its artificial intelligence features, including a major Siri upgrade later this year, the tech giants said on Monday.
The multi-year partnership will lean on Google’s Gemini models and cloud technology for future Apple foundational models, according to a statement obtained by CNBC’s Jim Cramer.
Sort of weird that they would announce such a big deal this way rather than official releases/interviews/etc, then again, the talk has been – at least on Apple's side – to downplay the partnership. We get it, it's sort of embarrassing to have to outsource your work in such a key aspect of technology, let alone one you believed you were at the forefront of not that long ago, at least with regard to Siri. And one you promised would get grandma home from the airport soon, only to fail to launch. So now you're stuck outsourcing that work to not just someone else, but one of your chief rivals for years and years. Ouch.
Of course, even as Android battled the iPhone, the two companies remained wedded in Search – one of the most lucrative and divisive deals of all time. And this deal undoubtedly expands upon that one. While they're declining to comment on terms, Mark Gurman of Bloomberg pegged it at around $1B a year back in November. That seems low, especially when we know the Search deal itself is $20B+ a year. But there are a bunch of details we don't yet know. Apple's actual statement on the matter is interesting:
“After careful evaluation, we determined that Google’s technology provides the most capable foundation for Apple Foundation Models and we’re excited about the innovative new experiences it will unlock for our users,” Apple said in a statement.
This makes it sound as if Apple won't run Gemini straight-up, but instead will use Gemini to train (distill?) their own foundational models. Unless those models are really just a white-labeled version of Gemini, which they may be at first.
Another option may be to pipe Gemini into Siri as an option alongside the current ChatGPT partnership – which Apple said isn't changing with this news. Perhaps such integrations are even more pronounced in a future build of iOS this Spring...
If you squint, you can see a three-step strategy from Apple here:
1) Place Gemini as an option for Siri alongside ChatGPT
2) Use different Gemini flavors to help train Apple's own Foundation Models
3) Work on your own custom Foundation Models without Gemini
Again, if they go with #1, I have to imagine they make the placement much more prominent. Even if not in the UI, perhaps they'll just make Siri default to Gemini and/or ChatGPT (depending on which the user chooses?) much more often – perhaps for basically all but the system-level queries (setting timers, etc).
This would buy them some time to work on #2, getting their own new models up to speed. Perhaps for iOS 27 in the Fall, or perhaps even later. Presumably, they'll get first access to new Gemini work from Google to stay at the cutting edge with this deal. And it seems like Apple will probably become one of Google Cloud's biggest partners, if they're not already? As I wrote last November:
There are probably a few other interesting wrinkles in there – one of which may be Apple's willingness to do this because a lot of their cloud infrastructure is already running on Google Cloud. So this may not be as heavy of a lift and as big of an ask as it may seem on the surface. And while the "walled off" aspect is clearly a must for Apple here, you could also imagine that the company may be willing to share some data – fully anonymized, of course – back to help constantly improve the model. And that may speak to why Google would want to do this deal (well, that and the money). Apple has devices in the wild at a scale that basically no one can match. Maybe Samsung, but this potentially unlocks a totally new user base.
Then with #1 and #2, or some combination, #3 would give Apple even more time – years – to completely rebuild and rework their own in-house AI that's less dependent on others. Apple doesn't even have a leader in place for that work yet to replace John Giannandrea and the remaining team has been gutted by Meta so... they need some time. As a bonus, this deal gives access to the best technology right now while they can take their time to figure out if LLMs are fully worth doing on their own, or if other, newer types of models/tech comes into favor...
One more thing: it is interesting that Apple emphasizes the "careful consideration" aspect of the process here. Presumably that means they weighed continuing to do this on their own, but also the possibility of partnering with Anthropic, as has also been reported previously, or going deeper with OpenAI. Not surprising that Apple would go with Google – who yes, just passed them in market cap – over a startup. Still, this must be disappointing to OpenAI given the current partnership. With Anthropic, it just seems like something they only would have done if Apple made it make sense monetarily, which they clearly couldn't agree upon. Back to what I wrote in November:
Apple may not have wanted to pay Anthropic $1.5B a year to use Claude but $1B a year to a partner that is paying you $20B+ for that Search deal? That can be just an in-kind deal! "Google, you know that $25B you owe us this year? Make it $24B, but we'll take a custom build of Gemini. Deal?"
Deal.
Update: There is now an official joint statement, at least on Google's blog.
Similar to the one Apple gave to CNBC, but a bit longer:
Apple and Google have entered into a multi-year collaboration under which the next generation of Apple Foundation Models will be based on Google's Gemini models and cloud technology. These models will help power future Apple Intelligence features, including a more personalized Siri coming this year.
After careful evaluation, Apple determined that Google's Al technology provides the most capable foundation for Apple Foundation Models and is excited about the innovative new experiences it will unlock for Apple users. Apple Intelligence will continue to run on Apple devices and Private Cloud Compute, while maintaining Apple's industry-leading privacy standards.
That last bit seems key. The wording is vague (undoubtedly on purpose), but it seems to suggest that Apple will be able to train Apple Intelligence with Gemini to still run locally on devices – and for the non-local queries, presumably Private Cloud Compute is already running in Google Cloud, even if they don't exactly tout it.
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
2026-01-12 21:58:05
There are two camps. Either you're in the camp that thinks AI is absolutely the future of everything and anyone who says otherwise is a moron. Which includes the other camp: those that think that AI is the worst thing in the world and is going to ruin everything. There are varying degrees within those camps, of course – for example, the "Doomers" who think that not only is AI a problem, it could quite literally be the end of the world – but you're on some spectrum in one of those camps...
2026-01-10 23:42:37
As the CEO of a major streaming player, you'd think Anthony Wood would rather not get Hollywood all riled up. Maybe he's taking a page out of Ted Sarandos' former playbook?
“I have no idea if there’s an investment bubble, but I know that AI is going to be huge, and it already is,” Wood said during a headliner conversation with Variety‘s co-editor-in-chief Cynthia Littleton at the Variety Entertainment Summit at CES Wednesday. “So it’s going to affect lots of different industries, but in our industry, it’s going to lower the cost of content production. I predict within the next three years, we’ll see the first 100% AI-generated hit movie, for example.”
The first part is obviously a reasonable position. Most people think it is an investment bubble, but also that it's perhaps a needed one given the prospects of the technology, which is where I generally fall (outside of a handful of outlandish deals). But Wood is not in the business of investment advice or thoughts. Again, he is in the entertainment business and so not only talking about the prospects of "100% AI-generated" movies, but predicting that they'll be here within three years, producing hits, is, um, controversial.
I also tend to think it's silly. Will we be able to produce 100% AI-generated content in three years? For sure. But we can already do that now. It's what is flooding social network feeds as we speak. What started as AI-generated images has morphed into video. And what started as short clips is naturally morphing into longer ones. But really all you need to do in order to make longer ones is simply stitch shorter ones together, obviously. And so the AI-generated trailers that have been flooding our feeds ranging from fake Avengers movies to fake Zelda movies are just one step removed from full movies. Well, maybe a few steps. But really, it's just time. Anyone can do that right now.
The bigger question is if there will be an audience for that content. And that's the controversial part of Wood's comment. It's not the creation – again, that will happen – but the notion that it would lead to a "hit movie". I have a hard time believing that will happen for a range of reasons, including the current powers that be.
I suppose it depends on how you define "hit". But I take it to mean a movie that millions watch via "traditional" methods, be it in theaters or now on streaming services (or on "regular" television, I suppose). Three years is a long time in AI, but in the grand scheme of entertainment, it's short. Movies are just being greenlit right now that will come out in three years. The mechanisms are already established to distribute those movies. So where is this AI hit movie going to slot in? Certainly not in a movie theater!
Okay, so what about a streaming service? Maybe. But again, the most obvious such service to create hits, Netflix, is in the midst of trying to buy a movie studio and as such, Ted Sarandos is now busy backtracking from his aforementioned previous comments. Certainly Netflix, like Roku, will use AI to enable talent (and yes, drive down costs), but I have a hard time believing they'll go forward with a full AI-generated movie in that timeframe. Also, just imagine the lawsuits?
Other services? Maybe! But it certainly won't be Disney. I don't think it will be Amazon. HBO will either be owned by Netflix or Paramount, and again, it won't be them. So it could be a smaller player, but will they be able to make a movie a true "hit"? Could Roku?
The wildcard would be YouTube. Because of the UGC underpinnings of the service, you'll certainly see movie-length AI-generated content uploaded. Especially from those outside the current systems – that's a good thing! But will any of it actually "hit"? Maybe as a novelty some will get a ton of "views", but it's hard to imagine it would be considered a true "hit" movie. More just like a proof-of-concept type thing, perhaps.
People will point to music and note that there are some AI-generated "hits" popping up. Two things there. First, I still suspect there is a novelty aspect to this. Second, a three-minute song is a different beast than a two-hour movie. Many people are willing to give up three-minutes of their time. Two-hours is a commitment. One other element that our AI-enabled future is going to highlight: time is the ultimate premium.
Now, longer than 3 years for that AI-generated hit movie, all bets are off. But that window Wood gives seems too short. Not because of AI, but because of the way distribution currently works.
I would also just say that I continue to bet that while such AI-generated content is coming, I think once that novelty factor wears off, you'll be hard-pressed to find big hits there in the way you do with human-created movies and television. And, counterintuitively, I think as AI-generated content continues to flood every surface, it will raise the demand (and thus, value) for human-generated content.
I do think there will be "hybrid" content – part human-created, part AI-created – but I think people will prefer to watch work that was created by other people, because they will value the time and care spent on it. We'll slowly realize it's just as much about the input, as the output.
To be fair, Wood does hit on some of this in the same chat:
“I think people underestimate how dramatic that’s going to be,” Wood said. “I mean, obviously I don’t think people are going to get replaced. Humans are still the creative force behind creating content and hit shows, but the cost is going to come down dramatically, and that’s going to change a lot of companies’ business models. So I’m focused on, how do we take advantage of that? That’s a big opportunity for us.”
Again, that's reasonable. The AI-generated movie topping the box office in 2029? Less so.
One more thing: I continue to be intrigued by Roku's "Howdy" play:
For Howdy, Wood describes the $3-per-month Howdy offering as “not designed to replace a major streaming service like Netflix or Disney, it’s designed to be an add-on service.”
“The opportunity for Howdy was, if you just look at what’s going on in the streaming world with streaming services, they’re getting more and more expensive,” Wood said. “They keep raising prices, and they keep adding larger and larger ad loads. So the part of the market where it actually started, low cost and no ads, is gone now. There’s no streaming services that addressed that portion of the market. That’s the opportunity for Howdy. It’s three bucks a month and no ads and it’s doing extremely well. Just like we built the Roku Channel using the promotional power of our platform, that’s what we’re doing with Howdy. We’re using that to grow it. But Howdy has very broad appeal. There’s lots of people in the world that want a $3-a-month streaming service with no ads. So we’ll start on Roku, but we’ll also take it off platform as well. I think it’s going to be a really large business.”
I've been critical of Roku with regard to ads in the past, but this makes a lot of sense to me. And I suspect we'll see others start to copy this model too.
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
M.G. Siegler
2026-01-09 23:53:06
Last March, I set out to map the OpenAI "Constellations" – that is, the startups not only in OpenAI's orbit, but those directly tied to it, founded by entrepreneurs that had previously been at OpenAI. No surprise, given both our moment in time with AI and the current funding environment around AI, there are a lot. So I set a threshold at the $1B valuation mark. By my math back then – again, a whole 10 months ago – those entities had a combined valuation of around $200B. And when you added in the valuation of OpenAI itself at the time, we were right at the $500B mark.
Just six months later, those numbers had changed drastically, so I checked back in on those valuations. At the end of August, that $200B aggregate valuation was past $400B. And with OpenAI's updated valuation, we were right around $1T in combined value.
Well, we're not quite six months since then, but given that Anthropic, xAI, and OpenAI itself – the three main drivers of those valuations, obviously – have all either just raised or are in the process of raising again, let's check back in...
Given that these two companies formed by those formerly affiliated with OpenAI are valued above $600B by themselves, it's probably worth breaking them into their own category, and we know how much AI companies love to append "Super" to everything, so let's go with "Ursa Super Major".
These combined numbers add up to over $100B now, just for Ursa Major, without the 'Supers'!
So that's a solid haul for this group just over $10B.
Okay, so grouped all together, we're looking at an aggregate valuation of just over $730B – a number which has nearly quadrupled in less than a year. And nearly doubled just since August. Obviously it's top-heavy, but even taking out the 'Supers', you'd still have a group valued collectively at $123B. Would that alone be the highest valued group post-PayPal Mafia? Or even if you remove Elon from the full equation, you're looking at almost exactly a $500B aggregate valuation for the OpenAI diaspora.
And if you add OpenAI itself into the mix, with their newly targeted $830B valuation, well, you're truly in the stars now: just over $1.5T.
That's awfully close to Meta's market cap. And if/when OpenAI and Anthropic go public, both may shoot past that mark by themselves given likely investor demand (depending on timing). And if/when xAI merges with Tesla, making them an AI company as well...
M.G. Siegler
M.G. Siegler
M.G. Siegler
2026-01-08 22:56:09
Do you like email?
It is, of course, a rhetorical question as no one likes email.1 It's a fact we're reminded of each and every new year after the holidays are over when you open your email service and your face is met with a fist in the form of your inbox. And while you might think that Google likes email because they run the most popular email service in Gmail, with 3 billion plus users, there's obviously a very real association risk here. If a company makes the product you most hate to use... there are negative halo effects, just ask Meta.
Anyway, the good news for Google is that they may finally have the appropriate tools to combat the email problem. While they've long been inserting AI into Gmail here and there – the auto-completion/correction features were probably some of the first AI that a lot of the world engaged with in a forward-facing manner – Gemini now seems robust enough to insert it everywhere. They've obviously been ramping up doing that in Search so as best to disrupt themselves before someone else does, and now it would seem to be Gmail's turn.
In a blog post today, the company outlines what Gmail will look like in the "Gemini Era". A lot of it you've undoubtedly already seen in the form of 'AI Overviews' – though this is seemingly getting a nice expansion to your entire inbox based on queries, and not just individual overviews at the top of emails. But there's also the more standard and straightforward: 'Help Me Write', 'Suggested Replies', and, of course, 'Proofread'. But the real key is something you haven't seen to date: 'AI Inbox'.
While Google is just starting to test it now, you can see what it looks like and how it will function in their post. They describe it thusly:
Your inbox is filled with updates; some are critical, others are just noise. The new AI Inbox filters out the clutter so you can focus on what’s most important.
AI Inbox is like having a personalized briefing, highlighting to-dos and catching you up on what matters. It helps you prioritize, identifying your VIPs based on signals like people you email frequently, those in your contacts list and relationships it can infer from message content. Crucially, this analysis happens securely with the privacy protections you expect from Google, keeping your data under your control. This lets high-stakes items — like a bill due tomorrow or a dentist reminder — rise to the top. We’re giving trusted testers access to AI Inbox before making it more broadly available in the coming months.
The first part sounds like just an expansion of what Gmail has long done with algorithmic sorting. I turned this off long ago as I found it not that useful when it worked and insanely frustrating when it didn't. But the second part is the key. From the sound (and look) of it, 'AI Inbox' is going to completely blow up your inbox as you've known it to date.
To be clear, that old inbox will still be there. Google isn't crazy enough to force 3B+ users into this new reality – he says this and then immediately remembers Google+ being shoved in the faces of billions of users – but apparently there will be a new 'AI Inbox' in the sidebar above your regular, old, hated inbox.
And in this new inbox you'll find not email, but information. Things you need to know or do, automatically surfaced for you from that dreaded old inbox. The key, of course is Gemini. Google's AI may finally be good enough to fulfill the promise of killing your inbox. Or at least beating it into submission. A place you go from time to time when you want to remember the pain you left behind.
But let's not get ahead of ourselves. Plenty have promised such solutions in the past. For this to work, it has to be incredibly accurate and useful. It has to overcome old habits and falling back into the "I'll just reply to the email myself" mentality. Sure, it may take just a few seconds to respond to that email, but in aggregate, we're all spending hours every week and as such, days every year on email. Days. Completely consumed by email. Gone, never to come back. Because of the scale of your inbox.
Google can use that scale to their advantage to train their AI to make this work. No, they're not reading your emails to do this, but using other signals and data at scale to construct this new type of inbox. Honestly, they may be the only ones who can do this.
And if they do, it's step one to a potential true end of email. I mean, it will always stay around as a sort of fall-back – a true cockroach of the internet – but the way you interact with it will change drastically. I wrote about this topic last June in a post entitled: With AI, Email May Actually Morph Into a Task List. A couple passages I'll highlight since it was for members of The Inner Ring:
Moving this concept over to email, if AI can write your email and read your email, it's easy now to joke that in many jobs, such as the ones I've done in my career, you could just go on vacation. Of course, you couldn't really do that for the same reason as mentioned above: at some point, someone – a human – has to be in the loop about something. I mean, honestly, for a lot of email back-and-forths, probably not – but for some, there are real-world tasks to be completed. By someone. One day, bots – robots – may be able to handle those too. But in the more immediate future, I think this looks more like emails being distilled into actual action items.
Yes, your inbox will go from being a de-facto to-do list (generated by someone else), to an actual to-do list (generated by your AI).
To some, this will sound like absolute hell. But I suspect it may lead to actual productivity gains for most people because again, the scaling of the email inbox has become untenable. Boiling email down to its essence could work.
That's what Google is creating here with 'AI Inbox'. But again, it's just step one. The next step is obviously using AI Agents to do many of these tasks/to-dos created,2 including responding to the messages:
Anyway, the point is that I can see a world where AI actually does lead to the end of email, in a way. It won't eliminate it, but it could eliminate your need to do it. Or, at the very least, cut back on it quite a bit because it will be abstracted into an AI layer above it, where you talk to your AI assistant about tasks and to-dos. Sure, you'll still be able to send it, but it will be more like disengaging the autopilot to fly manually. For most things, it will become something the bots do on your behalf.
Speaking of Agents:
And this antiquated technology could become the ultimate fallback for "agentic" communication when various newer protocols don't align for whatever reason. In many ways, email already is that fallback for many things today. Being the cockroach of the internet has some advantages...
Wouldn't that be fun? A sort of "have your agent email my agent"...3
One more thing: even beyond the whole end-of-email thing, there's another upside I see here, a broader one:
To others, this will sound downright dystopian. We're taking human-written email and turning it into bulleted action items generated by AI. But again, it feels inevitable. And, in many ways, needed.
Oddly, it may lead to a world in which letters – old school letters – make a comeback. I've long been of the notion that I think one of the second order effects of AI is that human-made creations will *increase* in value, and we might see something along these lines with handwritten notes.
Yes, this may lead to a new sub-economy where people are paid to write these personalized notes a la Theodore Twombly in 'Her'. Yes, this is a 'Her' reference without mentioning Samantha. Until now. Damnit.
Killing email to restore the value of writing to humanity. Who says no?
1 I'm sure some have hated email longer than I have, but I'm very well documented... ↩
2 We used to call these "bots" part 1. ↩
3 We used to call these "bots" part 2. ↩
2026-01-07 19:39:06
When looking back at the origins of the "hackquisition" model for acquiring startups, Microsoft was clearly the godfather of such deals. Their March 2024 deal to not buy Inflection, but instead to bring on much of their team was groundbreaking in this regard, and it has led basically all of their rivals to copy (or expand upon) the same general structure.1 But actually, there was almost a "hackquisition" that Microsoft did before the Inflection deal... OpenAI.
I've long thought that the Inflection deal structure, and thus, the "hackquisition", was born out of a deal Microsoft nearly had to execute under duress. In November 2023, OpenAI's board ousted CEO Sam Altman and the ensuing weekend saw Microsoft, led by Satya Nadella on down, scramble to save the situation. The company had invested around $13B into OpenAI at the time – a lot of money in those days2 – and they were worried it was going to go to zero overnight.3 So they implemented a plan. Either OpenAI would reinstate Altman, or Microsoft would just hire the entire company.4
Sound familiar?
But wait, don't "hackquisitions" also involve some level of IP rights? You know, "non-exclusive licensing agreements"? For show, if nothing else? Yeah, Microsoft didn't need those here because they already had far more rights to OpenAI's IP, thanks to their deal terms following their massive investments. So yes, this would have been the first "hackquisition". And now we know how much it would have cost.
Thanks to text messages from OpenAI COO Brad Lightcap, uncovered as part of Elon Musk's lawsuit against the company, to Nadella and Altman, we have the number. $25B. Or $29B if they wanted to include Ilya Sutskever. Which of course Microsoft would have, but given his role in said coup d'état of Altman, it's not clear that would have happened...
As relayed by Rocket Drew of The Information:
Lightcap disclosed Sutskever’s financial stake after the OpenAI board fired CEO Sam Altman and Microsoft CEO Satya Nadella prepared to hire Altman and OpenAI colleagues. Lightcap wrote to Altman and Nadella that paying OpenAI employees for their equity would require $25 billion excluding Sutskever’s equity, or $29 billion including Sutskever’s vested units. (It’s not clear how much Sutskever had in unvested equity.)
The day following the exchange, Microsoft chief technology officer Kevin Scott said publicly that Microsoft would hire and match the compensation of every OpenAI employee who chose to resign from the startup, in the event that Altman was not reinstated at OpenAI. Those hires never happened because the OpenAI board rehired Altman later that night.
So yeah, Microsoft would have paid at least $25B, and perhaps $29B to do that would-have-been "hackquisition". That's not too surprising since OpenAI was valued at... $29B during secondary sales that were taking place at the time. But the math wouldn't have been quite that straightforward.
The key bit to remember here is that no one actually owned equity in OpenAI at the time. This was long before the conversion to the public benefit corporation, and everyone from Microsoft on down simply had percentages of the rights to future profits. Microsoft was clearly willing to "true up" those employees and match the on-paper value, and seemingly assume that those employee shares equated to 100% ownership of the company. Of course, there were other investors in OpenAI, even beyond Microsoft. So the bigger question remains if they would have given those other investors anything.5
That is, of course, another key component to the "hackquisition", paying out investors to ensure they go along with the deal. Given how freaked out those investors were during "The Blip", it seems reasonable to think they believed they'd be getting nothing for their investments. But it's also probably reasonable to think that Microsoft would have given them something to avoid future litigation, if nothing else. Again, this is what Microsoft did with Inflection investors several months later.
And if that's the case, that probably would have pushed the deal well past $30B – especially if you consider compensation for Altman, who famously had no direct equity – or even "Profit Participation Units" – in OpenAI at the time (and apparently still doesn't). But if Sutskever was going to get $4B... Microsoft probably would have wanted to keep Altman happy, in order to keep him.
Anyway, there was one massive shareholder Microsoft didn't need to pay out: Microsoft. Everyone now knows they own 27% of OpenAI, but again, at the time no one had any equity. Microsoft had the right to 49% of those profits (though actually a much higher percentage until their initial investment was paid back) and undoubtedly could have and would have argued that meant they "owned" 49% of the company. Unless it was regulators who were asking.
Had anyone else tried to "hackquire" – or even actually acquire – OpenAI, they probably would have wanted at least $15B for their troubles, if not more. And it highlights just how killer of a deal Microsoft would have gotten here.
Call it $30B - $35B to effectively buy OpenAI in late 2023. Given that the company is currently raising around an $850B valuation just over two years later...
Alas, by the end of the weekend, Altman was back in place atop OpenAI and "The Blip" was just that, a blip. This is clearly the outcome Nadella and Microsoft wanted, undoubtedly due to regulatory concerns if nothing else. But the mad scramble also pretty clearly planted the seed that would grow into the "hackquisition" model.
One more thing: if Sutskever's stake was worth $4B back at that $29B valuation, the simple math indicates it would be worth $117.2B at that $850B valuation. Of course, per above, it's not that simple. For one thing, there's been a lot of subsequent funding – to say the least! – and thus, dilution since then. For another, there was that whole conversion of the company to a public benefit corporation, which also turned the profit share percentages into actual equity – we know how Microsoft converted (from 49% of profit shares to 27% of equity), but we don't know how anyone else converted. Sutskever also stayed another six months at OpenAI after "The Blip" so he undoubtedly vested more equity before he was off to start Safe Superintelligence. Anyway, his equity could be worth anywhere from tens of billions all the way up to $100B – assuming he hasn't sold in any of these tenders...
Update January 8, 2026: Further testimony and messages reveal even more about Satya Nadella's motivations after Sam Altman was ousted: he was teriffied that Altman and Greg Brockman might spin-up a competitive lab – epsecially since they were getting calls from the leadership at Google and Meta. So he quickly had offers drawn up for the pair...
1 Notably bringing on board OpenAI co-founder and Inflection CEO Mustafa Suleyman to head up Microsoft's AI efforts, presaging Google doing a similar move with Character.AI's Noam Shazeer and Meta doing that with Scale AI's Alexandr Wang... As an aside, the $650M price point seems quaint today! I mean, NVIDIA just did a "hackquisition" for $20B! ↩
2 Also sort of wild that Microsoft apparently hasn't really committed more money since that point. Sure, they've "participated" in some subsequent fundraises, but presumably only token amounts. That shows just how clearly they've outsourced the support of OpenAI to SoftBank and others post "Blip". They clearly decided they needed to try to own AI themselves... ↩
3 To be clear and fair, a lot of that money was in compute credits, and future capital commits. So Microsoft undoubtedly wouldn't have actually lost $13B, but still many billions. And arguably a lot more if you take into account where OpenAI is valued today... ↩
4 Well, anyone who wanted to come aboard. Which, judging from the immediate reaction from the employee base, seemed to be pretty much everyone. ↩
5 And I won't even go into the "grants" Elon Musk made to OpenAI at its inception. He clearly – clearly – would be suing Microsoft right now over those, either to block/unwind any deal, or get paid. But it may have been a harder argument without any legal equity conversion... ↩
