The Definitive Ranking of Tech Company Takeability

Friday, November 21, 2025

Welcome back to Sharp Text. If you’re reading this via email, thank you for subscribing. For anyone reading on the web who would like to receive weekly posts, you can sign up here.

Good morning!

Today’s focus will be tech companies and takes. Having hosted podcasts of all kinds for more than ten years, takes are near and dear to my heart. So the question today is simple: which tech companies are doing the most for the take economy in 2025?

To be clear, these aren’t judgements on the quality of these companies. Someone or something that’s particularly valuable to the take economy is not necessarily good or bad, but usually provocative, and preferably has a story that’s still being written, leaving plenty of ro0m for half-baked theorizing. For example, we can frame today’s exercise with recent news from Hollywood: the takeability spectrum runs from Robert Redford’s career (uncontroversially excellent, universally respected, what can you say?) to Sydney Sweeney’s career (your mileage may vary!).

In any event, you get the gist. I won’t be posting next week because of Thanksgiving, so what we’re celebrating today are the companies that all tech content producers, or content consumers, should be most grateful for this year. With that in mind, here are my notes…

15. Microsoft. The cloud business is fine, Office lock-in isn’t going away, and sure, the Activision deal looks like a disaster and the Copilot efforts are lagging a bit, but this is still a mature and extremely profitable business with an enterprise software moat that has survived for more than three decades. Even the years-long tension with OpenAI ended with Microsoft making a perfectly sensible decision to hedge against OpenAI’s upside rather than bet on its own ability to make great AI products. This is Microsoft under Nadella: smart, unsexy, and by not overtly trying to dominate the world, they wind up (ever so quietly) continuing to dominate the world. Unfortunately, none of is very takeable. Crush Microsoft and you will sound like kind of a moron; evangelize about Microsoft and you will sound like someone who needs to get out more.

14. Amazon. Microsoft in e-commerce clothing (complete with a massive cloud business). There may be conversations to have about why the stock has underperformed the rest of big tech over the past few years, or what the company is doing wrong in A.I., but the core business is fine, and the logistics moat is impossible to beat. More importantly, we’re living through one of the wildest eras technology has seen in 25 years. Does anyone really want to spend their time on a deep dive into the Andy Jassy era?

13. Nvidia. Seems like it should be higher, and definitely would be higher if this list were focused on the companies that produce the most content for Sharp China. In America, though: CUDA lock-in is real, efficiency advantages are only going to become more important, there are 500 ASIC start-ups and not one that matters, and demand for Nvidia chips still outpaces supply every single quarter. It’s very, very tempting to play devil’s advocate on a $4 trillion company that conquered the world in the blink of an eye — people do it on Twitter every single day! — but none of those arguments seem to hold water. Nvidia is great at what they do, what they do is boring but incredibly valuable, and now they are well-positioned to dominate for the foreseeable future. Actually a very frustrating company from a takes perspective, because all the most fun, skeptical takes tend to be wrong.

12. Netflix. Would have been much higher on this list five years ago! Unfortunately, most of the best Netflix debates have been settled for at least two years. Spending from rival Hollywood studios didn’t erode Netflix’s dominance, and in fact some of those studios are now selling their content to Netflix. Netflix did offer an ad tier, and no, it didn’t kill the subscription business. There’s maybe a medium term question about whether YouTube poses a threat to Netflix, but that argument feels strained in 2025. Netflix is so powerful today that its audience can generate a new franchise for studios, its churn rate is five times better than some of its competitors, and there’s not really any question anymore as to whether its formula is sustainable. And speaking of the formula, this is a good Netflix take:

11. TikTok. A company that has generated more takes (billions) than profits (???) over the course of its short history, and one that would have been much higher on this list a few years ago. The problem is that TikTok debates are very familiar now: yes, the the CCP could almost certainly use TikTok to manipulate algorithms and access user data; no, the U.S. wouldn’t be violating the First Amendment if it banned the app; and yes, Trump ignoring Congress and refusing to enforce the law is probably his most egregious abuse of law during his second term (a whole different realm of takes). There are also more basic questions — is this app making an entire generation dumber? — that have their own, obvious answers. That said, I’ll happily have any of those conversations. Some of them are important! And while we’re at it, I’ll throw out one more for consideration: TikTok pioneered addictive short-form video and forced Meta to push short form videos of its own, thereby training an entire generation to consume content this way, and further eroding the attention spans of millions of young people. To the extent this trend plays a role in diminishing the influence of of Hollywood and the traditional entertainment industry, which has been a source of U.S. soft power for roughly 100 years, isn’t that also a pretty massive victory for China? It would be a bigger coup than, say, convincing people in Taiwan that the Houston Rockets don’t exist.

10. The Other Model Makers. Why is Anthropic doing ads on Ringer podcasts? Do coders love The Big Picture? And what are these aesthetics? The company is pushing millennial-core minimalism that looks like a Hims ad, on one hand, and feels like it’s from 2015, on the other. It seems like the target audience for that “thinking” campaign is “people who want use AI but also lament its existence,” which would at least be consistent with every interview Dario Amodei has given for the past three years. Elsewhere: xAI sometimes looks like a very, very expensive hammer in search of a nail. No one knows what Ilya Sutskever saw or what his new company is trying to do. And how is Mira Murati’s new company, without revenue and without a product, raising funds at a $50 billion valuation? People bag on OpenAI’s spending and revenue projections, but at least there are users, products and cogent goals. This is the economy that makes no sense. All these players have to be on the list, however, because a) everyone enjoys teeing off on overvalued companies and/or effective altruists; and b) these companies are good for takes, because a world where OpenAI, Meta and Google are the only major AI players is far less interesting.

9. DeepSeek. 10 months ago, might have landed at number one on this list! For about 36 hours in January, the sheer volume of takes surrounding this company seemed like it might pop the AI bubble, and in D.C., DeepSeek’s success was treated as dispositive evidence that chip controls were hopeless, on one hand, or not being enforced properly, on the other. Today, DeepSeek’s R2 release has reportedly been delayed because the company was encouraged by the government to run on Chinese chips, the American AI market is frothier than ever, and the narratives surrounding DeepSeek’s costs ($5 million!) have been debunked. However… Deepseek is high on this list because it’s still used as the rhetorical stand-in for a Chinese open source AI ecosystem (Qwen, Kimi, et al) that might very well be a long-term problem for a variety of U.S. interests (chips, models, U.S. tech leadership around the world). Additionally, with now-close oversight and support from the CCP (with apologies to Meta, rumor has it that Vice Premier Li Qiang has banned rival companies from poaching DeepSeek talent), the next few years will be a fun experiment to see whether the party’s control freak tendencies will impact China’s national champion in AI.

8. Uber. Very similar to Netflix in that this company had entire armies of skeptics for almost a decade, and it’s harder to sustain any of those arguments in 2025. Unlike Netflix, though, a real disruption threat looms. If we assume that autonomous vehicles will be eventually ubiquitous in a variety of markets, Uber is like a team that doesn’t control its own destiny to make the playoffs. Essentially, Uber will need Waymo, Tesla and perhaps others in the autonomous space to strike out at monetizing their significant hardware investments — via their own apps, or maybe delivery — thereby forcing them to turn to Uber for access to customers, and preserving Uber’s negotiating leverage as an aggregator that can lean on fleets of other companies’ cars to deliver rides, extracting the same margins they would on any other trip. And look, it’s entirely possible that’s how it plays out; everyone already uses Uber and breaking those user habits is tough. But a) it’s easier to create new user habits when you’re offering an entirely new or better service, which Waymo sort of does; and b) while Netflix won by betting against the ability of companies like Warner Brothers, Comcast, and Paramount to thrive in a new technological paradigm, Uber is betting against Google, Tesla, and perhaps Amazon. A tougher row to hoe!

7. Meta. The core value proposition in the modern era — serving up addictive, mindless content from people you don’t know, surveilling user habits more effectively than any private company not run out of Beijing, and using that data to serve ads not too far off from that scene in Minority Report — may be looking increasingly bleak, but it’s also looking increasingly stable. We’re past the era of semi-regular panics over Meta’s lack of a moat, and the dominance looks as entrenched as ever. (And in fairness, Instagram ads are more useful than every other ad you’re assaulted with on the internet.)

All that noted, it’s everything else that Meta is doing that provides tremendous fodder for takes. Reality Labs has been happily lighting money on fire for a full decade with no end in sight. Every new pair of Smart Glasses is cool, yet no one in tech seems to understand that normal people don’t want to walk around with a computer on their face. Meanwhile, the AI division gets reshuffled every two months and is now full of poached researchers, on max contracts, who are improving Meta AI products that no one uses, as well as internal AI tools that should improve Meta’s ad products. We’ll see where that goes. Vibes would have been great for takes if the horrified reactions to that product hadn’t been instantly eclipsed by a rapturous reception for Sora. Speaking of which, it’s seems like Meta senses that OpenAI and ChatGPT are a real threat, and is responding by driving up the price for talent and hardware, while also releasing open source models that could theoretically undercut OpenAI’s long term pricing power. It’s defensive, but proactive, and at times very entertaining.

On the other hand: Will any of this work? Has Zuck in Founder Mode actually been additive to Meta’s business for the past 10 years? And outside of buying Instagram and WhatsApp and copying features from Snap and TikTok, when was the last time Meta built something new that people liked? These questions are becoming evergreen, and are at least more interesting than discussing Instagram’s ad load.

6. SpaceX

5. xAI and X

4. Tesla. I’m grouping all of Elon’s companies together because in practice, discussing one Musk company usually invites discussion of a few others. Related: The number of well-capitalized misses from Zuckerberg at Facebook should underscore just how hard it is to marry vision, timing and execution at scale, which Elon has done over and over again.

There’s an alternate timeline in which Tesla lands in the Netflix zone of this list, having successfully disrupted to the American car industry for the first time in 100 years, proving all its doubters wrong, and now sitting entrenched as by far the most dominant EV maker in the US (although China is making the rest of the world a lot more interesting). What’s great about Tesla, though, is that the company is pushing several orders of magnitude beyond that goal. From The Information:

Optimus is Tesla’s biggest long-term bet. Musk has said there will eventually be more humanoid robots than cars in the world, and that Optimus will one day be responsible for about 80% of Tesla’s market capitalization. Inside Tesla, he’s pushed the Optimus team to find ways to use the robot in tandem with another big, nearer-term bet: the Cybercab, according to a person with direct knowledge.

That includes Musk’s desire to have the Optimus robot sit in the Cybercab so it can deliver packages. That should be possible: newer versions of the Optimus robot are capable of consistently lifting and moving around with roughly 25-pound objects for three to four hours on a 30 minute charge, another person with direct knowledge said. But the connection between the robot’s torso and legs isn’t flexible enough to allow it to seamlessly get in and out of a Cybercab, according to the first person.

Ben wrote more about this on Tuesday, but let’s marvel at the insanity of that report. What’s envisioned above involves autonomous technology that doesn’t fully exist yet, Cybercabs that don’t exist yet, a market for automated delivery that doesn’t exist yet, and robots that do exist, but are not functional enough to drive around delivering packages. So, a lot of moving pieces! On the other hand, if anyone could will that science fiction vision into reality… And see? This is how you end invoking SpaceX catching a rocket as it re-enters orbit, as well as explaining that xAI will be critical to realizing any of those visions.

Elon is a maniac, with a Twitter account that’s completely intolerable to all and politics that are intolerable to many, but that shouldn’t distract anyone from the sheer volume of genuine breakthroughs that he’s pushing toward at all times, making technology more interesting for everyone. He’s the antidote to Meta building a trillion dollar business on the strength of Reels. And the more grand his ambitions and accomplishments become in the physical world, the funnier it becomes that he also owns X, and that him buying X had a bigger impact on American life than all of his other businesses combined.

Additionally, in the middle of all this, Musk still finds the time to sue Sam Altman every three weeks.

3. Google. What a two year ride. In 2023, the company that invented the transformer was behind in AI and rushing out hamfisted demos in Paris, their safety team celebrated Black History Month by accidentally making all history black, and ChatGPT was being hailed as the first credible threat to the search business since the Bush administration. Two years later… maybe this is going to be most powerful company the world has ever known? The search business continues to generate obscene profits and a Federal Judge was recently too scared to disrupt their distribution deal with Apple despite explicitly finding it illegal. Gemini 3 is state of the art (though it’s unclear how much that means at any given time), Veo is far and away the best in AI video, and Waymo is expanding to 5 more cities as of this week. YouTube is the most dominant entertainment platform on the planet, which will not only help Google own the future of consumption, but should also help the company’s AI efforts. YouTubeTV could be the cable bundle of the future if the company ever decides to make an aggressive play to own that market.

The advantages and opportunities in front of Google so are obvious and abundant that I’d be open to arguments that Google is too high on this list, as it’s now become clear to everyone that this company can and should win the future on a variety of fronts. On the other hand, I leave room for the possibility that Google and its massive bureaucracy will fail to capitalize on some or all of the foregoing opportunities, and I don’t know anyone in my personal life who prefers Gemini to ChatGPT. We’ll see. In the interim, I’m just enjoying the shift from the whole world asking whether Google needs to fire Sundar Pichai to now talking about him like he’s Keyser Soze:

2. Apple. This list is ultimately a celebration of takes, and in the current environment, is there any hotter take than just skipping out on AI spending altogether? I’ve come to admire it, and as a very satisfied iPhone 17 Pro Max user, I think it’s probably the right call. I don’t plan on switching phones anytime soon, neither do you, and besides, I don’t trust modern Apple to execute if the company were to invest hundreds of billions of dollars in AI infrastructure. Would they really better off flailing around in the wildnerness like Meta?

Elsewhere, Apple continues to collect a 30% fee on every app store transaction, an egregious practice that has generated a full decade of Stratechery hand-wringing and seems to face renewed scrutiny every six months. While I’d prefer if that behavior eventually led to meaningful consequences for a company that is now dangerously dependent on extortive Services revenue to drive growth (the app store fees, collecting a $25 billion toll from Google every year, and my favorite: launching an ads business just before deploying ATT changes that killed everyone else’s ad business), Apple is betting that regulators are too lazy to act, and companies are too desperate for Apple’s audience to do anything about it. Honestly, a reasonable calculus!

Finally, Tim Cook’s religious dedication to operations optimization was so extreme that it inspired one of the best technology books of the decade (mainly for its history of Apple) and one of my favorite Stratechery interviews of the year. This would be my guess for the controversial strategy that actually hurts Apple’s bottom line at some point in the next ten years, but we’ll see. Either way, Apple at once looks more vulnerable than it has for about 15 years, and more dominant than ever. Everyone has an opinion on what Apple should be doing and why, and it powers the take ecosystem all year long. However…

1. OpenAI. Listen.

OpenAI may or may not be the most important company of the future. There can be no doubt, however, that we are witnessing one of the most takeable enterprises in the history of the world.

From the day it was founded — with a non-profit corporate structure that sought to build AGI and then control it themselves “to ensure artificial general intelligence benefits all of humanity” — this company has divided the audience and invited either passionate support or aggressive eye-rolls.

10 years later, not a week goes by without a story that makes the leadership look kind of ridiculous (a Larry Summers snafu here, an Altman conniption there, and a friendly suggestion to the U.S. government over here), yet ChatGPT stands alone as the one AI product that practically the entire world continues to use and enjoy on a regular basis. And as fevered as the current landscape in tech can look, we should all be clear that the defining characteristics of that atmosphere — crazy lending, circular partnerships, overheated market reactions, uncomfortable social implications, and rapid uptake that makes anything seem possible — are all almost entirely attributable to the success and ambition of one company.

What I love about OpenAI is that it should be impossible for anyone to bet against a start-up with 800 million weekly users, but the company makes it so tempting. Not content to merely compete with Google, OpenAI is also trying to be a hardware company that competes with Apple (Jony and Sam saw you across the bar and like your vibe), the company’s also competing with Meta for attention and “time spent” on their app, and, according to Sam Altman in his Stratechery Interview, will compete for share in the enterprise space. Sam and co. are trying to do everything, and be everything, and will soon be serving users “mature content” (an arguably bearish signal if leadership is that desperate to juice engagement). The company also appears to be losing tons of money as it throws these ideas at the wall, though precisely how much is anyone’s guess.

That last bit is the crucial aspect of why this company continues to generate so many frenzied conversations, not only because OpenAI’s future prospects depend in part on how much it costs them to serve their product and that distinguishes ChatGPT from the stickiest consumer hits of the past (AI is, uh, not zero marginal cost). What’s most important is we don’t know. Incomplete information is a feature here, not a bug. OpenAI is a private company that doesn’t have to report out its costs and doesn’t have to answer to shareholders as it attempts five moonshots at once. The lack of transparency and accountability, coupled with a market that’s being redefined every month, leaves room for the whole world to theorize on what will happen next and why. Almost no OpenAI prediction in 2025 is definitively falsifiable.

Think OpenAI is the new AOL? I can buy that argument. But if it’s not, where might this story go? Will OpenAI hardware pose a threat to Apple, while its software becomes elemental to how the whole world works, shops, and learns? Do you trust Sam Altman? Will he still be running this company in four years? Is ChatGPT destroying higher education?

Whoever you are, wherever you are, I bet you’ve got a take.

A Taylor Sheridan Starter Kit

Brett asks:

As somebody who’s only dipped their toes into the Sheridanverse (Sicario + Hell and High Water) where do you recommend starting with the rest of his work?

Great question. Sheridan is an acquired taste and someone people never get there; my wife finds his shows to be too violent. The movies you mentioned are great in a more conventional, critically acclaimed way, whereas his TV shows take traditional TV formulas and heighten the drama with either violence, bizarre plots, or history that you won’t find elsewhere on TV. In any event, my recommendations would be the following:

1883. Probably his best television work to date. The one and only season chronicles the Dutton family ancestors (stars of Yellowstone) as they make their way across the American frontier. This show is brutally violent at times, poignant at others, and it’s all gorgeously shot. And no, you don’t need to watch Yellowstone beforehand.
1923. Another Yellowstone prequel. The pacing can be pretty languid at times, and there’s a sadist plotline that was probably a bit much, but I came to enjoy it over the course of the first season, and love it by the end of the second and final season. It’s his most sentimental work, and it features Helen Mirren and Harrison Ford having a great time. Maybe my favorite Sheridan project.
Landman. Great collection of stars — Billy Bob Thornton, Demi Moore (sort of), Jon Hamm, Ali Larter, a number of character actors from other Sheridan vehicles — and a killer concept for a show, unpacking the oil industry in West Texas. It has about six more storylines than it needs, there is a Jerry Jones cameo at one point, and when I wrote that Sheridan shows are often weird and fun in a way that’s refreshing, I was thinking of Landman.
Lioness. For people who loved Sicario, but also secretly enjoyed Sicario 2: Day of the Soldado. Drug cartels are again the enemy here, special ops are involved, and no, these kill missions have not been sanctioned. The show is gripping from the opening scenes of Season 1 and more or less maintains that pace throughout the run, though I still have lots of questions about Nicole Kidman’s marriage.

Yellowstone is also an enjoyable watch for the first two or three seasons, though I’ve been partial the work Sheridan has done since. And while we’re here… I was happy with how that Sheridan article came together, but as much as I celebrate Sheridan for taking more interesting creative risks than, say, The Bear, I don’t want to denigrate too much of modern TV, mainly because I still avidly consume lots of it. So, a handful of non-Sheridan recs for anyone who’s looking for something to watch over the holidays:

Task on HBO is the best show I’ve seen this year.
Industry on HBO is my favorite show of the decade (three seasons; the first one is a little bit weird and may be off-putting, but it’s the rare show in the streaming era that gets better with each episode, in every season).
Platonic on AppleTV is a great comedy to watch with your spouse. Much better than, say, The Studio.
The Gilded Age is another outstanding spouse show; nails the formula for compulsively watchable procedural.
America’s Team on Netflix. The best sports documentary I’ve seen this year.
Beckham on Netflix. The best sports documentary I saw last year, in case anyone hasn’t seen it.
Pluribus … Still making up my mind on this one, but I’m enjoying myself so far.

Letters of Recommendation

Empire of AI is wildly misleading about AI water use. Andy Masley with a piece that went viral on tech Twitter this week and was very gratifying to read. The notion that data centers use dangerous amounts of water is one of the more pernicious talking points distorting the modern AI conversation, and its persistence among the mainstream, particularly on the left, is good evidence that our information systems are still pretty broken.
The Bitter Lessons. Dean Ball with a really well-observed piece on the AI “race” between the US and China, and the ways in which that metaphor fails to capture the current dynamics.
What the Movies Need from Sydney Sweeney. Good, short column on the dignity of Sydney Sweeney’s (mostly terrible) movie career, and why Ross Douthat hopes she sticks with it.
How NOT to Negotiate with China. A brief note from Robin J. Brooks on how Trump should have handled the rare earths standoff, and why a modest tariff probably would’ve worked better than the maximalist strategy the U.S. side chose instead. He might be right, and he’s definitely right about the underlying dynamics in China.
Is Gambling Really Ruining the Integrity of Sports?. Jay Kang with a fair counterpoint to the reflexive argument that gambling is corrosive to the long term interests and integrity of sports leagues. I’m not sure I agree, but it made me think.
america against china against america. I recommended this on Sharp China Wednesday, but if you’d like a corrective to the sycophantic videos from Hasan Piker in China this week, this (very long) travelogue from Jasmine Sun was terrific, as she traveled to Shenzen, Shanghai, Yuyao and Hangzhou and visited a variety of tech companies along the way.

Also: this is not writing, but I thoroughly enjoyed this three minute video from my friend Spike Eskin on the “consensus of fear” in modern MLB awards voting (and lots of other places).

And with that, we’re done! Thank you to everyone who’s subscribed and written emails; the early response to this newsletter has been very encouraging. If you have questions for future newsletters, you can email me here. And since I won’t be publishing next week, have a great Thanksgiving!

Sharp Text is extension of the Stratechery Plus podcasts Sharp Tech, Greatest of All Talk, and Sharp China. We’ll publish once a week, on Fridays. To subscribe and receive weekly posts via email, click here.