What Kind of Future Do SubStack & Twitter Have?; & BRIEFLY NOTED: For 2023-04-10 Mo
Superclown chaos monkeys gonna super clown, gonna chaos, & gonna monkey; 1979 TTSS; software cost revolutions in our future?: implementation as the key link; Bair, Basu, Wolf...
FOCUS: What Kind of Future Do SubStack & Twitter Have?:
The keen-eyed and incredibly sharp Ben Thompson makes two critiques of SubStack here, and I have a bone to pick with one of them:
Ben Thompson: Substack Notes, Twitter Blocks Substack, Substack Versus Writers: ‘Substack has rightly earned the affection of a lot of writers… [who] want Substack to succeed…. [But it] is effectively asking for a donation from an audience that almost by definition doesn’t know any better. That doesn’t seem very writer friendly! Nor, for that matter, does this fight with Twitter…. [The] product bet… makes a lot of sense: Substack needs to take big swings if it’s ever going to reach its valuation. Writers, though, who need Twitter’s distribution, didn’t sign up for this fight; they are simply stuck in the middle…
On the first of these complaints, I think Thompson is 99.99% correct. SubStack can hypnotize itself into thinking that for its writers as a group (although not for any individual writer) money they contribute toward extending SubStack’s runway is a net plus, given that they are already substantially reliant on SubStack. So SubStack can hypnotize itself into thinking that it is not exploiting the goodwill of its writers by valuing any buy-in they make at peak bubble valuation. But they are certainly not sharing the upside the way they should be since we are now not in a -1%/year but in a +1%/year long-run safe-real interest-rate world. Bubble-time valuations were crazy to being with. And duration-risk alone has cut the rational fundamental valuation of a 20-year duration investment in half since the start of 2021. Asking people to invest without dividing peak valuations by four is… not sharing, and SubStack really wants to be in a sharing relationship rather than an ens---ification relationship with its writers:
.
On his second complaint… I think Thompson is only 0.01% correct. First of all, SubStack did not pick a fight with Twitter. SubStack has a beta tab in its app that it is experimenting with. The beta tab is a microblogging platform (as opposed to the macroblogging platform that is its flagship email-web-app tab prodcut) that looks like Twitter (rather than, say, microblog), but that could develop into any of a very wide array of things, only one of which is a full-out attempt to move Twitter’s social-interaction graph and flow of messages over to SubStack.
Now Elon Musk reacted to this very strangely. He reacted by trying to ban links to and mentions of SubStack on Twitter, thus creating a substantial Streisand effect as millions of people who would never have heard of SubStack Notes now have it in their minds as a potential someday-replacement for Twitter that they might like to check out, and then backing down from his bans, while probably continuing an algorithmic-timeline shadow ban of SubStack from Twitter. Net effect on Elon Musk and Twitter? Unambiguously negative, to yet further reinforce the true and apposite meme that a SuperClown has bought what was already a clown car that fell into a gold mine. (The last part of which is a thing that Mark Zuckerberg may or may not have said.) Net effect on SubStack? Unambiguously positive, to substantially raise its salience. Net effect on SubStack’s writers? Probably negative, as the SubStack runway extension from the publicity is probably outweighed by diminished usefulness to them of Twitter as an audience-attraction and audience-communication mechanism because of the algorithmic shadowbanning.
And yet Ben Thompson sees this as SubStack’s fault? The implicit model of Ben Thompson’s reasoning here seems to be:
There are people who have, or think they have, or want to have durable moats around a key location in a tech value chain very valuable to the consumer in which every other link has been commoditized, in the manner that Microsoft and Intel held their dual monopolies in the x86 PC era.
ONE MUST NEVER DO ANYTHING TO MAKE SUCH PEOPLE THINK THAT YOU ARE CONSIDERING ATTACKING THEIR MOAT!!!!!!
Why not? Because then they will undertake steps to destroy the value that you and your users gain, whether or not those steps actually help them.
Thus Ben Thompson seems to think not just that (i) SubStack has a primary duty to its writers but (ii) bears responsibility for the value capture strategies undertaken by others up-and-down the value chain; but also that (iii) SubStack will (a) build a Superintelligent AGI, (b) build a working time machine, (c) transport that Superintelligent AGI back into the present, and (d) use that Superintelligent AGI to build an accurate predictive model of a malevolent SuperClown Chaos Monkey newly at the wheel of the clown car that fell into the gold mine.
Perhaps this is demanding too much of SubStack?
Certainly I would not have expected the SubStack Notes beta to have triggered the banning and shadowbanning that it has.
My view is that the right attitude to take is this:
SuperClown Chaos Monkeys are going to clown, going to chaos, and going to monkey.
Given that, everyone not just adjacent to, but in the same neighborhood, as Twitter needs to be working as hard as they can.
Working at what? At building escape hatches, fallbacks, workarounds, and alternatives.
SubStack Notes is supposed to be one such
SubStack Notes is suppposed to be a way that SubStack can keep the process by which people discover its newsletters robust in a post-Twitter flameout world.
A Twitter flameout seems more likely with every passing day.
As my podcast co-host Noah Smith (and when are we going to record another hexapodia?) said, now more than a year ago, he was rooting for Elon to buy Twitter because:
Twitter was in a bad equilibrium.
An Elon shake-up might well improve it a lot.
If Elon did not improve it it would flame out, and far healthier microblogging frameworks would be very likely to grow in the ashes.
I think it is to say that the possibility of (2) is now off the table. But (1) and (3) still look pretty good to me…
ONE VIDEO: 1979 Tinker, Tailor, Soldier, Spy:
ONE IMAGE: I Confess I Think SK Ventures Is Vastly Optimistic Here:
I see a tenfold increase in the productivity of the typical software engineer from AI systems over the next decade—and I also see us as having already had a tenfold increase in productivity since 1960:
MUST-READ: Implementation as the Key Link:
Lexington: Biden’s big bet on big government: ‘Top aides like John Podesta are racing the clock to transform America’s economy…. Republicans view Mr Podesta as a ruthless partisan and the climate spending as a fat target for investigation. That is a playbook Mr Podesta knows, and it seems far down his list of worries. Nor is he much concerned Republicans will legislate to undermine the climate law itself. New projects are already under way, as the renewable-energy industry begins investing against the legislation’s ten-year horizon. Marjorie Taylor Greene, a fire-breathing Republican representative from Georgia, may have called global warming healthy for the planet, but she has celebrated the new jobs coming to a solar plant in her district. “This is going to get rooted and be very hard to uproot,” Mr Podesta predicts.
Very Briefly Noted:
Maury Obstfeld: The Global Dollar Cycle: ‘dollar appreciation shocks predict economic downturns in EMDEs… are highly correlated not just with tighter U.S. monetary policies, but also with measures of U.S. domestic and international dollar funding stress that themselves reflect global investors’ risk appetite…
Danny Critchton: Industrial Policy: ‘Kicking Away the Ladder… Creating the Cold-War University… Asia’s Next Giant… The Politics of Large Numbers… The Collapse of Complex Societies…
Henry Farrell (2012): Hayek and the Welfare State: ‘Bruce Caldwell’s defense is not that Hayek didn’t claim that the welfare state was the slippery slope to gulags and jackboots—it’s that he didn’t say this in The Road to Serfdom, although he did say it in his later works…
Conor Sen: SVB’s Collapse Says a Lot About San Francisco and Seattle…
Tate Ryan-Moseley: AI might not steal your job, but it could change it: ‘I’ve spent time over the past two weeks looking at the legal industry and how it’s likely to be affected by new AI models, and what I found is as much cause for optimism as for concern…
Jacob Steinhardt: Complex Systems are Hard to Control: ‘Powerful deep learning systems such as ChatGPT… are complex adaptive systems, which raises new control difficulties that are not addressed by the standard engineering ideas of reliability, modularity, and redundancy…
Shane Delamore: Setting ChatGPT's Temperature: Introducing the ChatGPT API…
Erik Hoel: The gossip trap: ‘How civilization came to be and how social media is ending it…
Jonathan V. Last: A Brief History of Media and Audiences and Twitter and The Bulwark: ‘Elon Musk didn't attack Substack. He went after the subscription model for journalism…
Rotten Tomatoes: Renfield…
¶s:
Robert Bair: Putin, Trump, Ukraine: how Timothy Snyder became the leading interpreter of our dark times: ‘Historians aren’t supposed to make predictions, but Yale professor Timothy Snyder has become known for his dire warnings – and many of them have been proved correct…. Snyder’s dire warnings were easy to caricature… yet Trump’s attempts to overturn the 2020 election allowed him to claim vindication…. Snyder’s view of Putin was still more ominous. In Putin’s Russia, Snyder sees a corrupt autocracy that has turned to neo-fascism in an attempt to regain… imperial glory…. In his opinion, the full-scale invasion that started last year was not, as some saw it, a minor regional conflict, but rather an atrocity of epochal significance: “It is about the possibility of a democratic future”…
Kaushik Basu: How Singapore Impresses: ‘There are many reasons for Singapore’s rapid growth over the past six decades, but its authoritarian regime is not necessarily one of them. Instead, the city-state’s economic miracle has more to do with its unique brand of multiculturalism and enviable social cohesion…. Singapore’s rapid growth… high savings and investment rates, first-rate education system, and intelligent policymaking. The intelligent policymaking can come from an authoritarian leader, but there is no causal link there…
Martin Wolf: The EU’s future in a world of deep disorder: ‘In adjusting to today’s crises, the bloc needs to decide whether it wishes to be an ally, a bridge or a power…. How might the EU, liberated from the internal obstacles created by a sovereignty-obsessed UK, respond to a global environment so different from the one it hoped for some three decades ago?… A role as a bridge would come naturally to an entity committed to the ideal of a rules-governed order. The question, though, is how to be a bridge in a deeply divided world in which the EU is far closer to one side than the other. The third alternative is to seek to become a power of the old kind in its own right, with resources devoted to foreign and security policy commensurate with its scale. But for this to happen the EU would need a far deeper political and also fiscal union. The obstacles to that are legion, including deep mutual distrust.
As a software engineer I may be a little biased, but I think the idea that software costs will fall to nothing is beyond laughable. Most of software engineering is figuring out what you want, whether it’s possible, and whether the thing you made actually does what you want. The making of the thing is already pretty easy. So 10xing the making part, which I can imagine, is only going to take a fraction of the time. And it’s just not the case that there’s a bunch of people who can do the requirements and feasibility and fit-for-purpose assessment parts and are only missing the ability to code. (Or rather, there actually are some of those people, but they already have high-paid jobs, often as engineering or product managers, because those skills are the ones in demand.)
The ability to code is not sufficient. It’s like saying Stable Diffusion will make architects obsolete. It’s a productivity tool, not a replacement.
Maybe I know the wrong kind of lawyers, but trial lawyers are about producing a product that appeals to a certain set of consumers, juries. They have to satisfy judges and cow the opposition, but their real job is to sell the case to a jury. There is an entire literature on juries, and the more I've read of it, the less I think there is a place for AI in assessing them. The current state of practice is to set up focus groups, mock juries and mock trials, and then figuring out what worked and what didn't. These are always full of surprises, so one four day job in Little Rock, for example, turns into a series of three day gigs testing out alternatives. There's good money in running those mock trials which has the advantage of payment not being contingent.
The rest of their practice seems to be responding to documents with appropriate counter-documents, ideally in minimal time, and AI might help a bit with the mechanics, but the challenge is aligning the strategy and the tactics. When someone slams the judge with a plea or a response at 4:45PM on Friday, the response has to address both the surface and structural attack, and it has to be sent out by 4:55PM. There's also the matter of telling people you will be late to their dinner party, something you really shouldn't outsource to an AI.