A newsletter version of this newsletter
A newsletter version of this newsletter
Trying out a different format while I work on a larger piece.
Programming note: The semester just started at GWU this week, which is going to cut down on the time I have available for blogging. My ambition for the next few months is to publish one essay per week here. I’m working on a bigger-picture essay right now, something I’ve been tinkering with for several months and I finally think is almost done. So I might not finish a long-form piece this week.
I’ve also been thinking about adding an actual newsletter format into the mix, once-a-week, with links and brief commentary to items that connect with the subjects I write about.
Let’s see how that goes. Here are five things that caught my attention in the past week:
1. AI Hype Train
A new Substack just launched, AI Snake Oil, by Princeton Computer Scientists Arvind Narayanan and Sayash Kapoor. The first issue is really good, and I’m looking forward to seeing what they do with it.
I think I’m going to pick up a new hobby — just watching serious deep learning/machine learning/AI researchers throw shade at the Artificial General Intelligence crowd. It’s delicious.
On a related note, apparently “Mystery AI Hype Theatre 3000” is a thing now? Yes, please.
2. Silicon Valley Ideologues
Y Combinator has a new President, Garry Tan. I’d never heard of Tan before. Apparently that’s because he preemptively blocked me on Twitter.
Anil Dash made a good point (below). Seems as though Tan blocks everyone who isn’t sufficiently devoted to the YC worldview — tech accelerationism, techno-optimism, founder adulation, faith in “disruption” as a force for good, and distrust of government regulation.
I think the YC worldview is ultimately pretty bad. And, between the techlash years and the current deflation of the tech bubble, we’re hopefully heading into a moment where it can be productively challenged, or amended, or maybe even discarded.
It isn’t a great sign that the new head of YC is so generous with the block button. Here’s hoping it’s a sign of their continuing slide into irrelevance.
3. Longtermism, again
Eric Levitz interviewed longtermist philosopher William MacAskill. It’s a good, thoughtful interview, well worth your time.
That said, the interview does nothing to alleviate my concerns about what I call “Jackpot Longtermism.”
MacAskill isn’t a Jackpot Longtermist. He’s a moral philosopher who has genuinely concluded that longtermism is the strongest philosophical position. But he’s also cast himself in the role of longtermism’s John the Baptist.
At several points in the interview, MacAskill sidesteps serious concerns by insisting that longtermism is not a monolith. (Q: “why wouldn’t union busting be permissible [under longtermism]?” A: “I guess the key question there is what you think about union busting.”) In effect, it is like a religious order that can only be truly understood by fellow devotees through hard study and reflection. It effectively means he can trumpet all of the most appealing implications of longtermism while hand-waving past all the flashing red warning lights. (Is Peter Thiel a real longtermist? No. Could he use the moral logic of longtermism to prop up dangerous political projects? Yes, but that’s not MacAskill’s concern.)
And there’s the disjuncture. I don’t much care about longtermism as a branch of moral philosophy. I’m a political scientist and communication scholar who studies digital technology. I care about the way that longtermism can be easily grafted onto tech elites’ existing preferences and used as a shield for socially-corrosive endeavors.
That’s not the type of conversation that moral philosophers normally engage in. The longtermist response to my fretting over Jackpot longtermism boils down to “well that’s not the specific version of longtermism we endorse. Please don’t ask us to defend a version of longtermism that we do not support.” But MacAskill is on a huge publicity tour. He is attempting to extend the reach of longtermism far beyond the confines of classroom thought experiments. And, to the extent that he succeeds, that means we have to examine it through different lenses than we usual apply to treatises on moral philosophy. We have to worry about how the coarse reading of the theory will be deployed by people with power and a political agenda.
Longtermism appears to either have a naive theory of power or no theory of power at all. (“Here’s a thought experiment. Let’s earnestly discuss it and then apply it to our own lives and to society writ large.”) It needs a better one.
4. Is the streaming economy in worse shape than I thought?
HBOMax has pulled 200 Sesame Street episodes, along with dozens of animated programs and the Batgirl movie.
Amazon Prime has sunk a billion dollars into “The Rings of Power,” and is sending signals that it will reevaluate the future of the streaming service based on how the Lord of the Rings prequel does.
Netflix hasn’t given The Sandman a Season 2 green light yet, even though it has been their top show for the whole month of August. Why not? “Because Sandman is a really expensive show.”
Back in July, I wrote about the economics of the streaming wars (“Will Netflix Be Alright?”). My impression at the time was that (1) it’s the end of the unlimited-spending golden-age of streaming television, (2) that was inevitable, it couldn’t last forever, and (3) Netflix would ultimately be fine.
I’m starting to wonder if I underestimated just how much debt the big streamers have built up. Things are falling apart a lot faster than I expected.
Each of these individual cases can be explained away.
Discovery just bought Time Warner/HBO Max, so they suddenly have a ton of extra debt to discharge. CEO David Zaslav is tearing everything to the ground to find cost savings. (If he ruins Sesame Street, I will learn the mystic arts and cast an incantation that haunts his dreams for a thousand years. Do not mess with Sesame Street.)
Netflix is probably going to renew Sandman. If they don’t, that’s a big negative signal.
Amazon Prime is probably just raising the stakes for the new series that they have bet absurdly big on.
That said, I’ve been operating under the assumption that Amazon Prime, Disney Plus, and Apple TV+ were the three streamers that would effectively be immune to the changing valuation of the streaming economy. Disney is a content empire. They can monetize their content in a dozen different ways. Amazon and Apple are tech behemoths. They can lose billions on their television services and still have it make good business sense, so long as they build up soft power and public goodwill, that makes it marginally less likely Congress will pass laws that interfere with their real business models.
If The Rings of Power turns out to only have a niche audience, and Amazon responds by scaling back their original content protection, that’s a big sign that the foundations of streaming video are worse than they seemed.
(All I really want is a second season of Paper Girls on Amazon Prime. That show us too good to become a victim of unfortunate timing.)
5. Big Money Ruins Everything - Elon Musk edition.
There’s a passage from one of Matt Levine’s recent newsletters (“Musk Tries a New Way Out of Twitter”) that I keep thinking about. This was in reference to a NY Post story about Jack Sweeney, the teenager who built a Twitter bot that tracks Elon Musk’s private jet flights. Musk asked him to stop, the kid said he would stop for $50,000, Musk refused. Then, in a twist, some guy who runs a smart air conditioning company offered to pay the $50K on Musk’s behalf, but Sweeney declined because the money wasn’t coming from Musk himself. Here’s what Levine wrote:
The layers of this one. One day, when I teach a business-school class on the Elon Markets Hypothesis, I will spend a day on this series of proposed transactions. Consider:
Musk’s perspective: If you are the richest person in the world, as Elon Musk is, you never need to spend any money. When you want something — a mansion, $50,000 to pay a kid to take down a Twitter account — you can just say “huh I wish I had that thing” and people will fall over themselves to buy it for you.
The air conditioner guy’s perspective: If you are a person in the business world, and you can solve some pain point for Elon Musk, that is so obviously good for your career that you should just do it, you don’t need to talk to Musk first, you don’t need to have any deal in place, you don’t need to have any particular mechanism in mind to spend $50,000 doing it. “The guy hoped that Elon would notice and buy his company,” sure, whatever, reasonable bet. We talked about this above and last week, but right now Elon Musk has one very obvious problem — he needs to find some evidence that Twitter Inc. is a massive fraud, so he can get out of his $44 billion deal to buy it — and I suspect there are a lot of people working to solve it. Some of them (his lawyers, etc.) work for him, but most of them are doing it on spec. If you solve that problem for Musk, surely he’ll buy your air-conditioner company or whatever.
Sweeney’s perspective: He’s a college kid, presumably he is not rolling in cash, and he is willing to take his Twitter bot down for a $50,000 payment from Musk, but not from the air-conditioner guy. A $50,000 check from the air-conditioner guy is worth $50,000. A $50,000 check from Elon Musk is priceless. You can tell that story — about the time you blackmailed Elon Musk for $50,000 — forever. That story will get you all sorts of jobs in the tech industry, possibly even at SpaceX. “The way finance works now is that things are valuable not based on their cash flows but on their proximity to Elon Musk,” I have argued, and here is an example. Fifty thousand dollars from Elon Musk is worth much more than $50,000. And: “If he let me fly with him on his jet, record it and talk about it — and maybe not even pay me the $50,000 — I would take it down,” the Post quotes Sweeney saying. Sure! Flying on a private jet with Elon Musk is worth so much more than $50,000! That is serious proximity to Elon Musk. Do you know how much the air-conditioner guy would pay for 10 minutes on a private jet with Elon Musk?
This is just a great example of how extreme wealth inequality warps incentive structures and degrades everything it touches. Musk’s wealth has gotten so extreme that it has its own gravity field. It leads to objectively weird scenarios like this one in which everyone is behaving rationally and the results are at best bizarre/funny and at worst flat-out awful.
Anyway, a wealth tax would fix this. And also, Matt Levine’s newsletter is just really good.
That’s all from me, folks. I’ll be back with another essay-type post in the next week or so.
Minor note but Sesame Street should be fine since it’s not actually owned by HBO. It’s owned by an independent nonprofit (Sesame Workshop) which currently has a distribution deal with HBO, but episodes are also still available on PBS (both broadcast and streaming).
Your regular format is a 5 course meal, and this is a nice steak and salad, I love a nice steak and salad, so more of this when called for. Big yep on the streaming thing, being one of the few winners was gonna be equivalent to being ABC/NBC/CBS, so it made sense to go all in, but at some point its gotta pay off, and it looks like some point is now. I wonder how long the Tubis and Plutos can hang on.