Joseph Carlson — Prepare For The Earnings Week Ahead

Netflix has fallen roughly 50% from about $140 to about $74, and the press says it's in trouble: people are watching less, quitting shows after the first season, and the company is "getting desperate" — talking about always-on channels (like cable) and bundling other services. Carlson thinks the scary story is overblown. His key point is that "adding new kinds of content" isn't panic — it's literally what Netflix has always done, from licensing old sitcoms, to making its own hit shows, to documentaries, stand-up, Korean dramas, live events, and now podcasts. Each expansion looked desperate at the time and each one made the subscription more valuable.

On the "turning into cable" fear: Netflix has quietly tested always-on channels for five-plus years as a way to help indecisive viewers — and it would only ever play Netflix's own shows, not outside cable content. He agrees the engagement dip is real, but "fixing problems is what Netflix does" — it reversed a subscriber slide back in 2022 and added 70 million subscribers since, and today has 325 million subscribers, very low cancellations, and a fast-moving, data-driven management. So he treats the sell-off as a media-driven story, not a broken business: "if Netflix drops from the 70s into the 60s, I'll be once again increasing my stake."

Meta is one of Carlson's largest positions ($176k) and he's "very bullish." It jumped 21% in 15 days after Mark Zuckerberg announced a new, cheap, capable AI model ("Musepark 1.1"). Why is a cheap model good news? Because Meta doesn't plan to make money by charging for the model at all — it makes money from its enormous distribution (3.5 billion daily users across Instagram, Facebook, WhatsApp, plus glasses and assistants). So Meta actually wants AI models to become cheap, ordinary commodities that everyone can use, because that undercuts the companies (like OpenAI) whose whole business depends on selling a special, expensive model.

His bigger bet is a framework (below): in the long "tug-of-war" between companies that want AI models to be commodities and companies that need them to stay premium, Carlson believes the commoditizers — Meta, Google, Amazon — win, because they have "endless compute and endless money" to make their models just as good over time. That makes Meta's heavy spending strategic, not wasteful.

Carlson's organizing idea: sort the big AI companies by how much they want AI models to be commodities. Meta wants models fully commoditized (cheap, interchangeable) because it earns money from distribution, not the model. Google is in the middle — mostly a distributor, but it also sells a premium model. Amazon and Microsoft want models cheap too, because they earn from renting computing power (AWS, Azure) around whatever models exist. On the other side, OpenAI and Anthropic need models to stay premium and unique, because the model itself is what they sell (via APIs, agents, and subscriptions).

These goals are directly at odds. Carlson's call: over the next few years models get cheaper and harder to tell apart, and "when you can no longer distinguish between one model and the next, the companies with the motivation of commoditizing the models win" — the hyperscalers (Meta, Google, Amazon), not the premium-model labs.

In Carlson's framework Google sits in the middle — "partially commoditized, partially premium." Like Meta it has huge distribution (billions of users across Search, Gmail, Maps, Android, YouTube) and wants the cost of running AI to be very low, so most of its money comes from that distribution and its own cloud. But unlike Meta it also keeps a premium, paid-for model on the side.

He lumps Google with Meta and Amazon as one of the "commoditizers" he expects to win the AI tug-of-war, and separately notes Google (alongside ASML) has been "a massive winner in the portfolio for the past year." So it's both a framework winner and a proven holding.

Amazon and Microsoft play the same role in Carlson's framework: they want AI models to be cheap, competitive and interchangeable, and they're not focused on building a flagship model of their own. That's because they make their money renting out computing power — Amazon through AWS, Microsoft through Azure — so they profit no matter whose model wins, and they benefit when models get cheaper and more of them run on their cloud.

Carlson groups them with Meta and Google as the "commoditizers" who, in his view, ultimately beat the premium-model companies (OpenAI, Anthropic) as AI models become nearly indistinguishable over time.

Microsoft "fits firmly in the same category as Amazon" in Carlson's framework — it wants AI models cheap and interchangeable and monetizes AI through its cloud (Azure) and in-house platform rather than by charging for a special model. It's part of the hyperscaler group he expects to win as models commoditize.

The big banks report this week, and Carlson expects strong results. His reasoning is simple: the market has been very volatile, and banks make a lot of money when trading is heavy (both their own trading desks and the fees from handling clients' trades). They also earn big from credit cards — JPMorgan, for example, "makes a lot of money from Visa" — and banks generally do better when interest rates are higher, which they've stayed. Put together, he thinks "the big banks are going to do well."

ASML has been one of Carlson's best holdings — a $122k position that's up about $90k. He still loves the business ("the moat continues to be extraordinary"), but he did something notable: he trimmed it twice in the past month, selling some shares at around $1,900 and $1,750. The reason is purely valuation — after a huge, fast run the stock trades at about 45× earnings, which is very high even for a great company, and ASML still has some cyclicality (its demand rises and falls with the chip cycle).

His discipline: "it's warranted to take some off the table when you see valuations rise this high this fast." He isn't bearish and isn't selling out — he's just moving some money from an expensive winner into cheaper quality names (Uber, Netflix, Meta) because he expects the red-hot semiconductor trade to cool.

TSM reports alongside ASML and Carlson expects a great quarter — he's "still very bullish fundamentally." But like ASML, he's "not making any big bets on it this quarter." After a huge run in the whole semiconductor group, he'd rather add to cheaper, out-of-favor quality companies than pile more money into semis right as he thinks the trade is about to slow. So it's a hold-and-admire, not a fresh buy.

Apple sued OpenAI after a former senior Apple design executive allegedly moved to OpenAI and, over weeks, deliberately stole detailed designs for Apple's hardware. Carlson stresses this isn't Apple overreacting to a couple of documents — the theft looks extensive. Apple filed what's been called a "thermonuclear" lawsuit, and Carlson reads Apple's real goal as tactical: use the theft to tie OpenAI up in court for as long as possible so it can't release its own devices.

The tell, in his view, is that Apple clearly considers OpenAI a genuine threat in hardware — it knows OpenAI is building devices and has strong AI models — so Apple wants to keep a dangerous future rival stuck "in legal purgatory." He's describing the strategic fight, not making a buy-or-sell call on Apple stock.

Paramount is trying to buy Warner Bros. Discovery for about $110 billion, and 12 states (led by California) are suing to block it. The states' worry is concentration: this would merge two of the five big movie studios, and afterward just a few companies would control most film distribution. Carlson thinks the lawsuit probably fails. His main reason: the U.S. Department of Justice already looked at the exact same deal and reached the opposite conclusion, approving it.

He also thinks the states' argument is weak because box-office market share bounces around wildly year to year — one studio has a blockbuster (like Oppenheimer) one year and a flop the next — so it isn't the steady, durable market power you'd see in something like cable. Bottom line: he expects the merger to survive the state challenge.

OpenAI is Carlson's "fail of the week" on two counts. First, Apple's lawsuit: a former Apple executive who joined OpenAI is accused of systematically stealing detailed Apple hardware designs over weeks, and Apple is now trying to trap OpenAI in years of litigation to keep it from shipping devices. Second, in his AI framework OpenAI is on the losing side — it needs its model to stay premium and unique so it can charge for it directly, but Carlson believes the deep-pocketed commoditizers (Meta, Google, Amazon) grind that advantage away over time.

The rest is theater — Elon Musk calling him "scam Altman," Altman firing back about Musk's "space data centers" — which Carlson enjoys but treats as billionaires behaving badly, not investment substance. OpenAI is private, so there's no stock call; the stance reflects his skeptical read of its position.