What AI (doesn't) teach us about the mind
Also: Fake views, recent gun sales, the persistence of Christian Identity, protecting artists from AI, and platform notes
We’re in a bit of a holding pattern on content right now, as I just sent back the peer revisions for The Last Twitter Census, and I am close to finishing the first draft of the first installment in the Lawful Extremism project. The former will arrive in your inboxes sooner, but still needs formatting and other prep work, so a little while yet. I am hoping the latter will be out before year-end.
As Twitter continues to slide into the abyss, I’m focused on trying to recreate some portion of the flow of news and research, incoming and outgoing. I am not convinced this is my forever platform to do that—although I will note we’re closing in on 1,000 subscribers, which I gather may have some momentum-enhancing effects.
It’s not that I don’t like the tool; it’s great, feature-packed and with room to grow. But Substack has a lot of problems, some of which are very similar to Twitter’s problems, and others that are arguably worse. I’m not yet ready to make a change; there’s been a lot of chaos in my life, online and off, and something is constantly jumping the queue ahead of this issue. For now, I am going to try to keep using this platform to recreate some of the “old Twitter” magic, as well as sharing longer pieces, but I am actively thinking about and researching a variety of alternatives.
I have been trying to maintain some old-Twitter-like news links as previously presented in the Intelwire Daily Brief. Until now, that has been hosted on Flipboard, which is another flawed platform. I am experimenting with a hosted solution, which you can see at https://www.jmberger.com/headlines. Nothing about this page, including the branding, is final or official. It’s a rough sketch, more than a rough draft. Check it out over the next week, and let me know what you think. I like this better than Flipboard because I have much more control over the content and sources. If I get good feedback, I may switch this over to a dedicated website, either under the Intelwire brand and/or in conjunction with CTEC, in the near future.
There are platforms and combinations of platforms that can better execute my vision for the future of my online presence, but I want something that I know is going to last, that can be executed efficiently in the time I have available, and that doesn’t cost me anything (at minimum) or even pays me something, if only a pittance.
In the meantime… let’s start droppin’ some knowledge.
Reclaiming AI as a theoretical tool for cognitive science
This paper is a thoughtful examination of the current state of so-called artificial intelligence, or “spicy autocorrect” as it is called by those in the know. The angle here is that neural computing research potentially has a lot to teach us about how the brain works and performs cognition, but the current crop of AI pretenders don’t.
Lies, Damned Lies, and Social-Media Metrics
This is a very good article, not academic but in New York Magazine, on the fictional nature of many-to-most engagement numbers you see online. This will not come as a huge surprise to those of you who have read Optimal (and if not, why not), or my article a while back in The Atlantic. The NY piece is a great summary of the mess in which we find ourselves sitting, but a little short on consequences. Aside from juicing numbers for PR braggadocio or to pump up ad rates, these manipulations are fundamentally a tool of social control, and a pretty effective one at that. This is a much bigger deal than anyone really cares to admit. These fake engagement numbers move the economy, public health, and politics in ways that are measurable.
Who bought a gun during the COVID-19 pandemic in the United States?
Gun sales surged during the pandemic, and if you think that sounds alarming, well, congratulations, you’ve got good instincts. A new study finds that new gun owners skewed White, male and Republican, no big surprise, and that they showed strong tendencies toward “QAnon beliefs, pro-gun attitudes, Christian Nationalism, approval of former President Donald Trump, anti-vax beliefs, COVID-19 skepticism, … intimate partner violence, antisocial behavior, … desire for power, belief in a dangerous world, low agreeableness, low conscientiousness.” You may be thinking this is also no big surprise, but it’s a BIG change from pre-pandemic gun-buyer profiles, as the authors show. Buckle up, 2024 is going to be a hell of an election year.
AI Art and its Impact on Artists
Some recommendations for policies that may or may not help resolve the crushing ethical dilemmas surrounding AI art. As an amateur artist who has spent some hands-on time with AI generators, I’d love to imagine some ethically sound path forward, but color me deeply skeptical that such a thing exists.
Christian Identity Ideology and its Links to Antisemitic Mobilisation
For GNET, a short research note on the continuing relevance of Christian Identity ideology. While the organizational aspect of the movement has been in decline for many years, CI casts a long shadow in various corners of the online and offline world. As noted here recently, Christian Identity is overrepresented in the language models used to train AI. Some years back, I wrote a detailed history of the movement’s ideological development here, and my dissertation is going to look in more detail at some of the mid-1800s precursor ideas that contributed to CI beliefs.
Increasing Demand for Fact-Checking
Fact-checking often works—if people read the fact-check. Unfortunately, people who love misinformation often refuse to read fact-checks. This study looked at three successful ways to encourage the recalcitrant and to make fact-checking efforts more effective overall.
