2026-09-25 22:21:21
Aside from out-Googling Google with their AI personal assistant Muse, it's arguably more impressive that Meta appears to have out-Apple'd Apple by turning the just-okay Quest into what sure sounds like a better version of the Vision Pro. Yes, Apple's device undoubtedly still has nicer touches here and there, but that doesn't really matter when you completely flubbed the big stuff – quite literally. With the caveat that I've not yet tried it, the Meta VR Glasses sure sound like what Apple should have waited to ship as the v1 of Vision Pro...
🔗 Meta Unveils the Vision Pro Apple Should Have Shipped • Spyglass
Inklings is a newsletter featuring links and commentary from M.G. Siegler on timely topics found around the web.
🛩️ The Copilot "Super App" Lands – Well, more like it will be arriving soon-ish – some elements will be in limited preview "in the coming weeks", other elements will be in preview later this month. It feels like Microsoft wanted to get this out there now with all the other agentic activity – and perhaps a few more options dropping soon, including from key frenemy OpenAI. Overall, this sounds good on paper – unifying what has been a mess of Copilot options and brands into three tabs: Home, Code, and Autopilot. The latter is the rebranding of their Scout agent, and as I've said before, is great branding when paired with Copilot – a huge achievement for Microsoft! Basically, 'Home' is for chat (and 'Cowork' which is funny since Anthropic just unified these), 'Code' is for coding, and 'Autopilot' is for agentic work. And they're not trying to go after Muse (though there is a name/cute face option for your agent – not the cloud blob guy nor Clippy), this is all about work (but might those worlds collide at some point?). So it's a question of how much the deep Office integration will matter alongside the ability to mix and match models from other vendors. Again, that sounds compelling on paper – certainly for orgs that are Microsoft shops. But that was the case with the other instantiations of Copilot and it hasn't really mattered to date. I would say that perhaps the product is key here, but Teams has long been a crap product and that hasn't really mattered, so it's gotta be something else. Again, maybe it's the model picker. Because the business model still sounds a bit convoluted (usage-based in the app but license-based for the Copilot integrations elsewhere). [Verge]
📺 Premium Streaming Ads – A change to the Disney+ subscriber agreement would seem to suggest that the service may soon try to inject ads into their higher-priced tiers that have typically been ad-free. Hopefully this is just around sports, where such practice is obviously the norm due to how those broadcasts are constructed (which is, around ad breaks). But it also feels inevitable that ads will creep into all tiers eventually. Death. Taxes. Ads. Everywhere. (Even Apple.) There's simply too much money to be made – especially when those paying the highest prices are likely the most lucrative base to advertise to. That was always a mismatch. So will prices come down as a result? LOL. I'm sorry, you must be new here. WWND? What Will Netflix Do? [Polygon]
⚖️ DoJ Rides to Apple's Rescue – At first glance, this may seem like some political funny business – which this current DoJ clearly isn't above! – did Tim Apple get his buddy in the Oval Office to help here? But actually, this may be far more in the technical legal weeds. Apple has been arguing that, again, technically, they shouldn't be able to be held in contempt over their 27% commission funny business because that rate was never a part of the ruling. The Judge would undoubtedly argue that Apple violated the spirit of her ruling with their (again, comical as it actually raises rates on developers when they have to take care of their own payment processing separately) actions. But her move to put that rate at 0% was obviously over-the-top in the other direction, and subsequent appeals made clear some kind of commission was likely permissible. But if Apple can get this all further delayed... well, that may just save the App Store business and thus, the all-important Services business for Apple. For now, at least. [9to5Mac]
"I've seen Steve say it, and I've seen Tim say it. And so having the opportunity for me to say it was pretty magical. Kind of goosebumps."
– John Ternus on his first "one more thing" moment in an Apple keynote.
Speaking of Big Tech meddling in Hollywood, I meant to link to it a couple weeks back, but you can pretty clearly see why Amazon dumped Artificial – aka the OpenAI "Blip" movie – to indie distributor Neon: the trailer alone is dark as fuck. This is, um, not going to help with the overall AI narrative challenges. But if you can look past the guns in the bunker, Andrew Garfield sure seems to have Sam Altman's gait down perfect.
🎶 Listening to "Cleveland!" by Taylor Swift
🍺 Enjoying a Northern Monk Amy Hastings IPA
🇬🇧 Sent from London, England
2026-09-25 05:32:41
Two and a half years ago, just weeks after the launch of the Vision Pro, it was clear to me that Apple had made a mistake. They simply had no business launching the product in that current state. For all the talk of the company missing Steve Jobs' product vision, they clearly at least missed his ruthless feedback and curation here. The product that shipped should have been a dev kit – at best.
Yesterday, we saw the first glimpse of the product that should have launched. The only problem for Apple is that it's made by Meta.
To be clear, I'm still not sure how much of an actual problem this will be for Apple. Simply because I'm still not sure what the market for VR headsets looks like in 2027. But I know that a $1,299 device seems about three times more enticing than a $3,699 version. And a 100 gram pair of glasses that you can slip on rather than a 700 gram brick that you strap on to your face sounds about seven times more enticing. You'll still look like a complete jackass, but at least you won't feel as much like one!
That is to say, while the Meta VR Glasses still aren't perfect, they at least seem like a reasonable thing for consumers to buy. The Vision Pro was not. Even the second iteration – speed boosted, and heavier – was not. To be fair, this is also not Meta's first attempt in the space – far from it – but they've been willing to stick with it and keep iterating relatively fast whereas it seems like Apple has mainly moved on.
Will this Meta product bring them back to the table? Maybe, but they have other fish to fry, namely in trying to combat Meta's actual hit product in their smartglasses. Granted, they're not an Apple-sized hit – far from it – but Apple clearly feels the need to mobilize before they get such scale. Cupertino clearly wants to run their playbook and take such a product mainstream. So the Vision Pro gets left in the rearview for now.
But again, the Meta VR Glasses look like the v1 Apple should have shipped of the Vision Pro. It was almost exactly two years ago that I wrote about the obvious mistake Apple made it not offloading the onboard compute to a puck as Meta's prototypes were doing – especially if you were already forced to have a tethered external battery pack! My god that was a stupid design decision. I'm sure there was more to it with latency and what not. But come on.
Further, it seems like Meta was actually able to take some learnings from Apple to figure out what people want out of such a device. In a not-to-subtle callback to Jobs' original iPhone unveil, Meta kept reiterating that the VR Glasses were three products in one: a private cinema, a game console, and a computer.
That is exactly what the Vision Pro should be, yet isn't. The Quest wasn't better than Vision Pro, but the Meta VR Glasses very well could be.
Further, here's Mark Gurman reporting on Meta's product for Bloomberg:
In an interview, Meta Chief Technology Officer Andrew Bosworth said the company owes Apple “a little bit of a debt of gratitude.” He said many entertainment partners first became excited about immersive computing because of the Vision Pro and credited Apple with giving the industry “permission to explore split architectures,” helping inspire Meta’s external-pack design.
“I think they definitely were pioneers on some of those pieces for us and did the heavy lifting,” Bosworth said. “But otherwise, there’s not a huge comparison,” he said, noting the lower price and weight.
“We all know they missed the mark in that you just couldn’t wear that thing for very long,” Bosworth added. “It wasn’t glasses — you looked goofy wearing it.”
Again, to be clear, from what I've seen, you will still look goofy wearing the Meta VR Glasses. Well, that's a bit unfair. You will look like you are wearing medical glasses due to a loss of vision. If anything, Meta's glasses look more like those that the vision impaired wear because they are glasses and not a headset. But they'll also likely still be mainly worn at home. Or they'll be about a million times easier to carry on and use on a plane.
But I digress. Boz is correct in his other points. Meta can run here because Apple crawled.
Apple was so fucking slow in rolling out content tailored for the Vision Pro, but now that it's starting to roll in, Meta is right there to take advantage. Sure, things shot for Vision Pro may not directly translate to Meta's glasses, but there are probably ways to make it all work pretty well. Back in April 2024, I wondered if Apple would strike a deal with IMAX to more directly translate that format to their headset. Well guess who did?
Apple thawed the content producers mindset around making and tailoring such newfangled content – including with sports. Meta will now benefit from Apple's too-early mistakes as they largely sit back and watch from the sidelines. How very Apple of Meta.
2026-09-24 20:15:54
This was, without question, the most comfortable I've ever seen Mark Zuckerberg on stage. Over the years, he's oscillated from awkward to cocksure or worse, some combination of the two, which largely seemed to reflect the state of Meta itself. If that's the case, this year's Connect keynote may indicate a company coming back into its own.
Sixteen years ago, when the company was called Facebook and the conference was called f8, my overall takeaway from the keynote was, "I Think Facebook Just Seized Control Of The Internet". While that wasn't quite the height of their social networking powers – Facebook had a "mere" 400 million users then – they were clearly getting there. Confident enough to attempt to challenge Google for control of the web. Depending on your vantage point, it either never fully worked, or worked too well, but it felt like roughly the company's last peak, in confidence, at least.
Two years after that keynote, Facebook bought Instagram – for, Dr. Evil voice: one billion dollars. While many found that crazy at the time, with the benefit of hindsight,1 it seemed more like an acknowledgement that they could no longer build what's next, they needed to buy it. Two years after that, they repeated the trick by spending – Dr. Evil voice: nineteen billion dollars – on WhatsApp.2
In between those two deals, they tried to buy Snapchat, but when Evan Spiegel rejected their offer - for, Dr. Evil voice: three billion dollars – Zuck moved into kill mode, launching a half dozen efforts to end their younger rival. None worked, but by 2016, they finally got the model right – perhaps by sucking it up and owning that copying was exactly what they were doing with Stories inside of Instagram.
Anyway, sorry for the history lesson, but my point is that it has been a long while since there's been this kind of feeling around the company – again, at least when it comes to keynotes. And after a decade in the wilderness searching for What's Next™ – private messaging, crypto, and yes, the metaverse – they seem to have found it. Zuckerberg acknowledged as much on stage yesterday. While he's stated before that they perhaps erred in thinking that the VR (and AR) stuff would get here before the AI stuff, he finally acknowledged that "we've shifted our focus".
I mean, duh. But before this keynote, it wasn't entirely clear that their massive AI bet – no Dr. Evil voice needed, as we're talking hundreds of billions of dollars – was actually going to fully pay off. Honestly, the scale is such that we probably can't be sure for another several years, but the um, vision, seems crystal clear now.
What's wild is that the real linchpin of the entire keynote only came into being in the past couple of weeks! Muse.
Yes, it's early. Again, two weeks! But the early data Meta is seeing is clearly enough to push Zuck all-in here. He kicked off talking about Muse and ended talking about Muse. With a whole lot of Muse sprinkled in between.
Muse. Muse. Muse. Meta. Mark. The four-letter M-pire.
If Facebook was mentioned, I'm struggling to recall it. I think Instagram maybe got a couple shout-outs at most. And WhatsApp perhaps one (in relation to security). Meta has moved on. The social networks and services remain the glue holding the actual business together, but this is a very different company now.
To that end, I thought Zuckerberg's opening address was good! He was cool, calm, and collected. And yes, confident. He was close last year in his attempt to reframe the company around AI, but it was too predicated around the smart glasses. Those are clearly going well for Meta, but they're one endpoint3 – despite Zuck's clear obsession with pushing the iPhone into the background which is still not going to happen any time soon, if ever – AI is the means, and Muse is the harness.
Now the new 'Meta Audio Glasses' – take that, PervPods! – can be another endpoint. And the new 'Meta VR Glasses' another still – take that, Apple! Yes, even the perhaps two-times-too-cute 'Muse Charm' – take that, OpenAI!4
In this opening, Zuckerberg bridged the old Meta (Facebook) with the new Meta (Muse) by saying it has always been about empowering people (while sticking it to the Doomers). I mean, he has said this a lot over the years. But it has always felt a bit hollow, until now. I'm just not sure that using Meta AI to make what many consider to be "AI slop" or "social slop" was really helping people, but you can certainly see how Muse can!
I might also point out that it seems indisputable at this point that Meta's social products, while technically helping billions to connect, had some consequences – unintended or not – that hurt many. Perhaps millions. And while a settlement, legally, isn't an admission of guilt, it's certainly an acknowledgement that they'd like to move on.
But the reality there is that you can't really move on until you give people something else to talk about. Even a full-on name change didn't work. Aaron Sorkin has not forgotten. A true reinvention requires invention. As impossible as it seems now, maybe in 50 years we'll remember Meta as the AI company and not the social media company. Nintendo was once a playing card company. Nokia was a pulp mill.5
I acknowledge that such ideas and commentary will make people angry right now. I get it. A lot of people really, really dislike Meta. Some even hate the company. Even more distrust it, which is obviously a big problem for a product like Muse. Which is why security was such a focus from day one. But that's going to take time.
To hear Zuckerberg tell it right now, they don't even intend to monetize Muse through advertising. That's practically Meta's entire business at the moment! To be clear, he left the door open by noting that their current plan is to take "a small fee" from transactions facilitated by their agents (and undoubtedly to upsell usage credits to the "Whales"), but if nothing else, I suspect that Feed inside Muse will prove too enticing.
But the point is that these are the lengths which Meta is going to try to distance themselves from the old company. Again, they still need those profits, so they can pour them right back into the AI build-out, but those services are more or less in maintenance mode – or rather, milking mode.
Milk, another four-letter 'M' word. Funny that.
"Stay optimistic out there," was Zuckerberg's short and sweet sendoff – again, taking a swipe at the current AI narratives. There finally seems to be reasons to be optimistic about his company once again.
1 Though I'll just note that the deal didn't seem crazy to all of us at the time :) ↩
2 It's wild to think back to how much money $19B seemed like at the time, and how quaint the amount seems now in the Age of AI. We're seeing talent deals – "hackquisitions" not full acquisitions – happen for more money. ↩
3 One which many, including yes, Apple, will be gunning for soon. ↩
4 Meta was in full-on Godfather take-out-enemies mode, also announcing a partnership with Walmart (and a handful of other partners) days after Amazon blocked Muse. ↩
5 Yes, I'm purposefully avoiding the companies with far more notorious and problematic beginnings for obvious reasons, like Hugo Boss and Volkswagen. ↩
2026-09-23 21:31:15
Muse has likely kicked the AI agentic race that was already well underway into a higher gear, with seemingly all of Big Tech and Big AI set to respond. Or should I say, Big "SI" – you know, "Super Intelligence" the bigly branding that everyone is clearly going to adopt ASAP... Read on for thoughts on...
2026-09-23 17:51:30
🔗 Meta Has Been Testing a 'Human Concierge' For Its New Personal AI Agent, Muse • Reuters
Basically impossible for me to ignore this headline after yesterday's "Soylent Googlebook is Android". As Katie Paul reports:
Meta has been testing a “human concierge” for its new personal AI assistant, Muse, which entails having human contractors quietly handle some of the phone calls placed via the digital agent, according to internal company posts seen by Reuters.
The Facebook and Instagram owner told employees about the test last week, shortly after the company publicly launched a phone-calling feature for Muse.
The Muse agent, which has topped US app download charts in the two weeks since its debut, is designed to autonomously perform tasks like sending emails, shopping and booking travel on a person’s behalf.
With the phone-calling feature, a Muse user can tell the agent to dial phone numbers for US businesses and talk with people on the other end to handle errands like booking haircuts, checking whether a store has an item in stock or getting quotes from different contractors.
On one hand, this could not be any less surprising. This type of outsourcing-AI-to-humans happened a lot in the previous iterations of AI.1 As the story goes on to note, Meta's long-forgotten first attempt at a virtual assistant from over a decade ago – so long ago that they were still called Facebook – 'M' was leveraging this hybrid system.2 The difference, I suppose, is that everyone actually believes in the AI to not only work, but to be able to do these tasks autonomously now. It's quite the dichotomy to read headlines about AI being on the verge of destroying humanity, while at the same time reading stories that it can't figure out how to make a phone call.
The bigger problem here for Meta, of course, is the trust issue. Muse is a great product. That doesn't mean it will ultimately work as the 'AI personal assistant' holy grail product – a space with just as many corpses as any other over decades of attempts – but I think it has the best shot of anything I've seen to date. And again, that's in part because the AI is real this time. Well, real enough! But, as my inbox can attest, a lot of people have trouble envisioning using such a product from Meta – because it's from Meta. And so when there are headlines that actually the AI, which they promise has your data and information sandboxed away and safe, may be leveraging humans – contractors at call centers, no less – that feels... not great. Sort of like a bait-and-switch.3
And, well, Meta's own employees recognize this issue:
Some employees have raised privacy concerns about humans handling those calls, warning that the approach could result in sensitive information being shared unintentionally with contractors in call centers, the internal posts showed.
“It’s baffling to me why we think this feature is worth the risk,” one employee wrote in an internal post. “We are one bug away from unnecessary information being leaked to human callers.”
To be fair to Meta here, this feature was very much in beta testing mode and not live for all Muse users. But they were also clearly rushing to get it out there, undoubtedly to compete with other offerings from products like Instinct (which leads one to naturally wonder how they were pulling off the phone calling features...). Would they have disclosed the human-in-the-literal-loop if it launched? I mean, hopefully! Instead, they decided to roll-it-back for now, it seems.
The bigger picture remains that we're probably going to see more such hybrid approaches at the edges of AI as agents rise and try to crack real world tasks. The AI is better now, but there are still too many things in the real world that sort of require humans in some capacity – at least until the actual robots come.
1 How about 'outsourcing-AI-to-I'? ↩
2 'M' of course, reminds me of James Bond, which seems fitting given my own Muse bot's name... ↩
3 How about 'bAIt-and-switch'? ↩
2026-09-22 21:36:38
I was there Gandalf. I was there 3,000 years ago when Google unveiled ChromeOS. Back in 2009, this was a big deal and my headline for TechCrunch at the time, 'Google Drops A Nuclear Bomb On Microsoft. And It's Made of Chrome.' made some (shock)waves. The basic idea was that the web had disrupted the dominance of Microsoft, and so building an operating system completely around a web browser would be the final nail in the coffin. But the reality is that ChromeOS didn't destroy Windows – because Microsoft was busy destroying Windows themselves.
By the time Satya Nadella rose to CEO in 2014, it was clear that the future of the company would be the cloud and services, not operating systems. A year earlier, Apple started giving away macOS (then still called OS X) for free – well, provided you bought their highly priced hardware, of course. That's where their money was. And the iPhone rose to become a far bigger business than the Mac ever was.
At best, ChromeOS was disrupting yesterday's market. And the rise of Android just drove this home, and was awkward for Google. Chromebooks did well in taking over the education market – largely from Apple! – but the market that truly mattered was the smartphone.
Which brings us to the Googlebook. While the company framed it around AI during their initial unveiling back in May, all the coverage around the launch this week makes it pretty clear that it's far more about Android. Oddly, Google opted not to call it that, and keeps trying to talk up the more hybrid nature of ChromeOS and Android here, but every single story I've read focuses on just how Android this all is. In all but name, as David Pierce writes for The Verge:
Even the ultimate name of the devices and the software, Googlebook and Googlebook OS, hint at the extent to which Google hopes Aluminium brings everything Google-y together in one place. “We realized pretty quickly that this is about the whole Google experience,” Kuscher says. “If you have a phone that runs Google, and a car, and a laptop, that’s what this is about.”
(The naming here is a delightful bit of Google chaos, by the way. Googlebooks run Googlebook OS, which is mostly Android, but with some ChromeOS trappings thrown in. Get it? Me neither. They’re Android laptops.)
I had been wondering what they might call the OS, and I think I gave them two great options with 'Google OS' or 'gOS'. Alas they chose the bland, seeming straightforward yet somehow still confusing route in 'Googlebook OS'. But honestly, reading all of this, they probably should have just gone with some variant of Android to finally unify the OSes. Instead, now we have Android, ChromeOS, and Googlebook OS. Because why have one project when you can turn two into three?This is the (Google) way.
The reality is that ChromeOS probably isn't too long for this world. Maybe it continues to exist for that education market, but what consumer buys a Chromebook now? Someone looking for a dirt cheap laptop, perhaps. But the component pricing situation is straining that market and the MacBook Neo looks awfully attractive, even now at $699 ($599 for students).
Meanwhile – shocker – all the PC makers are suddenly launching their own colorful cheap Windows-based machines. And so this may be a hedge of sorts, with Acer, Asus, Dell, HP, and Lenovo all on board here to launch Googlebooks, undoubtedly with more coming – still no Samsung?! Or Google themselves?
And it may be a smart hedge because these machines are clearly built for people who use Android, which much like the iPhone overtook the Mac, Google's mobile OS is far more widely used than Windows – perhaps two times over. So if you're used to using that OS as your primary computing system, making it the computing system on your PC should make quite a lot of sense to many people. Even with the confusing branding.
But you also can't help but wonder if Google isn't sort of a step behind yet again. Yes, they played up AI and Gemini in that May preview, but again, that doesn't seem like what Googlebooks are or Googlebook OS is. Because the project was clearly started before the current Age of AI really took off. Google will undoubtedly try to cram it in wherever they can – just as Microsoft has been trying to do with Windows and now even Apple is in macOS – but it's not its essence. Perhaps the fact that it's simply new will prevent any sort of backlash like Microsoft has seen while trying to inject AI everywhere. We'll see.
At the same time, Microsoft is trying to push the broader PC world into the AI Age. The first Copilot+ PCs don't seem to have lit the world on fire, so it's try, try again. This time with a focus on massive memory and chips built for inference – by NVIDIA, no less. Apple has naturally morphed Mac Studio and even Mac mini to play here as well thanks to their early smart bet on unified memory and continued 'neural engine' work.
Anyway, my bet would be that there is no truly native 'AI PC' and that instead, people will run AI on the current computing paradigms as long as they're in the cloud. As smaller models become more capable, some workloads will move to the edge, but these may actually be on newfangled devices and wearables like AirPods and Apple Watch. And yes, the iPhone. Which will remain the actual computing hub for quite some time, and why I suspect we'll see other players try to come back into the smartphone space – even if they don't call it that.
That leaves Googlebooks as a sort of interesting PC alternative for the Android-native crowd, but not exactly the future of computing. Incidentally, it may be yet another nail in the Windows coffin, but it also still won't totally nuke that market, thanks to the strongholds of gaming and enterprise.
1 Maybe, just maybe, it will spur Apple to bring back the glowing Apple logo on MacBooks?! Nah. But a boy can dream... ↩