Killer robots writing poetry. This week OpenAI began sharing access to GPT-3, its new language-generating AI, with a select handful of developers . Poetry, stories, essays, and bits of work generated by the program leaked to Twitter, and the general consensus was “holy shit, our replacements have arrived.” Core to the human being is our self-conception as fundamentally creative, which we contrast with every other lesser intelligence on the planet. Now here’s a robot in that game with us. Or… is it? Sam Altman, one of the cofounders of OpenAI, took a stab at explaining the awed reaction to GPT-3 here:

I agree the natural language link is probably key to the reaction, but the programming aspect seems a somewhat separate issue. Most people haven’t even used GPT-3, we’re all just judging the AI by its creative work. This got me thinking: is everything generative really creative? The relationship between language and sentience is something we all intuitively feel. What is desire — the state of wanting things — without expression of desire? Language is in some very real sense the window into who and what we are. Expression without desire feels empty because it is, and I did feel a kind of emptiness in GPT-3’s work. But I think this has less to do with the program than it does with people. Separate from the incredible technological achievement in GPT-3, I was reminded of two things: 1) we are still very far from general AI, which Sam and others freely admit (a stark departure from the enthusiastic era of The Singularity is Near, which these days feels further off than ever), and 2) most people aren’t especially creative.

The GPT-3’s poetry struck me as most interesting in that it was bad precisely in the way most contemporary poetry is bad. In creative writing, people often hide behind the word “experimental,” an expression of intelligence rather than… whatever it is, exactly, that other human quality — that thing we all know unique to us, that spark, that something else, that intuition, or inspiration, when we take nothing and turn it into something. But most of the time human beings aren’t doing that, are we? GPT-3 has been able to make as many strong predictions of our writing as it has because so many of us aren’t really saying much. So while I don’t see the robots taking over any time soon, this probably is more bad news for tortured literary fiction writers. Sorry, guys.

A colleague of mine over at Founders Fund did a good job explaining GPT-3 this week. He’s pretty excited about what is absolutely an important technological advance, and it’s worth checking his piece out (though I do wish he would have had the AI write it).

Blue check down. There was a massive Twitter hack this week. Dorsey was fairly ambiguous on how exactly it happened (“social engineering” ???), and we still aren’t completely sure who is responsible (though a handful of journalists purportedly know, and are protecting the identity of the perpetrator(s) in a manner not afforded popular bloggers who, in stark contrast, have not committed spectacular feats of crime). Kind of off the beaten path, the piece of all this I found most interesting was the apparent cultural divide between “Blue Checks,” or verified users on the platform, and the rest of us. In a smart bit of damage control Twitter briefly prohibited verified accounts from posting, and so began a flood of happy memes. A personal favorite:

While the repeatability of Zimbardo’s famous prison experiment has been challenged, I’ve got to say the Blue Check divide really does have this energy. A handful of people have been anointed by Dorsey as voices of authority. Many of these people have quite apparently internalized the feeling that they are in some sense special. The unverified pick up on this feeling, and resent it. It’s a kind of classically classist dynamic, which is why it’s surreal so many of the verified media accounts are in such apparent favor of another French Revolution. Someone really needs to tell them they aren’t the proletariat. They’re the royalty.

The Blue Check divide doesn’t really seem that serious, but it is another kind of content curation, which Twitter seems to do a lot of compared to other social media companies. For folks on the political right already frustrated with the company, the hack didn’t help relieve them of the feeling they’re being targeted. Section 230 is absolutely going to become a battleground for tech. Is Twitter a platform, or is it a publisher? And if Republicans aren’t defending the company in Congress, who is Dorsey hoping will be on his side? Economically leftist political parties have not historically come to the defense of massive corporations looking to self-regulate. Another drama for another day.

Substack stacks the deck. Bari Weiss dipped out of the New York Times this week, describing a hostile work environment in her resignation letter, and once again pointing out the new reality for most media companies in the country:

"Twitter is not on the masthead of The New York Times. But Twitter has become its ultimate editor."

There’s a mob online policing speech, and many of Bari’s former colleagues are carrying pitchforks. The link between what is often positioned as a grassroots identity movement (religion?) and the institutions the mob purportedly loathes was once again made. Worth asking: if your view is the dominant opinion at every academic, media, and corporate institution in the country… can you really say your voice is marginalized?

Well, flying a little too close to the sun, here.

The culture war reignited, and of course we all retreated to our predictable tribal positions (GPT-3, is that you?). But let’s talk about the future of media for a second. Bari made absolutely no mention of Substack, which did not stop folks from wondering if she was about to join the party and launch a newsletter of her own. Two interesting things: 1) the overwhelming sense both haters and fans of Bari’s had was she was about to start something of her own, a dramatic departure from the days it was assumed a popular writer departing some popular paper would just go work for some other popular media brand, and 2) Substack is the de facto new alternative. If you have an audience, it’s easy now to sell subscriptions to your work, and to build a direct relationship with your readers. Then, on the heels of the Bari story, Andrew Sullivan — who has implied heavily he was fired from New York Magazine for not adhering to contemporary identitarian orthodoxy — did exactly that.

The subscription-based newsletter is increasingly seen as a somewhat safer form of media for any writer taking risks. No one wants to have their audience erased, which is explicitly what the cultural authoritarians at Twitter and the legacy media institutions are asking for. But the interesting thing about Substack is even if it’s ultimately taken down by the identitarian dogma now destroying so many once-great American institutions, there’s only so much they can do.

I’m not naive. No media platform can be fully trusted. But Substack built a way for me to not have to trust them. Or, I certainly don’t have to trust them as much as I have to trust Twitter. In some future world where I am finally canceled for my many crimethinks, the company could strip me of my monetization. They could delete my account. But I’d still have my email list, which means I’d still have my audience.

China’s bad, but that Mark Zuckerberg! I give the New York Times a lot of grief, but this isn’t because I hate the institution. It’s because it used to mean so much to me. And still, there are great reporters at the Times doing great work. Most recently I’ve been impressed by their coverage of the Chinese Uyghurs. If you haven’t been following this story, go ahead and take a look now. I only wish editorialists at the Times would do the same, as they seem clearly not to get it yet.

The Chinese government shackled Hong Kong — that’s millions of free people lost to totalitarianism, and we barely blinked. It invaded disputed border territory with India. It waged an international propaganda campaign against the West after crippling the globe by lying about a pandemic that in my opinion pretty obviously began after a Chernobyl-style accident at a Wuhan Lab. And it is now engaged in what appears to be an actual genocide. We really do just need to be on the same page, here. There is no moral equivalence between China and the United States, and certainly not between China and Facebook. FFS. We need to wake up.

-SOLANA