Kids online; techno-authoritarianism; encoding values.
PLUS: What everyone is getting wrong about Sora, OpenAI's text-to-video tool.
Hi, it’s Charley, and this is Untangled, a newsletter about technology and power.
👇 ICYMI
📖 I published my second-ever ‘reading list’ for paid subscribers — the first focused on AI. This one curated and contextualized my favorite articles, papers, and books about crypto and blockchain technology.
📚 I published my second Tiny Book, Crypto Untangled. Now, a paid subscription gets you access to two e-books, the archives, special issues like The Primer and Technically Social, and every other issue, including podcast episodes.
🧠 Google just launched a new open-source AI model - Gemma. Read my latest essay to understand what the fight between open and closed AI is really about.
📁 From the Archives
OpenAI claimed that its new text-to-video tool, Sora, can “understand and simulate the physical world in motion.” News coverage then speculated that this represents a technical breakthrough in AI’s ability to understand the physical world; and that its understanding of physics is an emergent property of the model. Sora doesn’t understand the relationship between physical objects. Given a grouping of pixels, it predicts another grouping of pixels. If you want to dig into the technical details,
wrote a 3-part breakdown in his great newsletter, .I’m not trying to pooh-pooh a cool tool. The videos are impressive. But I suspect we’re about to make the same mistakes we made with ChatGPT, anthropomorphizing the technology, and assuming the tool has a model of the world and displays emergent properties. Want to ready yourself for this discussion? Here are a few relevant essays from the Untangled Archives.
On anthropomorphizing AI: “AI isn’t hallucinating. We are.”
On the metaphors we select, and why they matter: “What even is ‘information’?”
On the false notion that AI exhibits emergent properties: “Does generative AI have emergent properties?”
If you enjoyed this tour of the archives, subscribe to the paid edition and access it whenever you’d like.
🗨️Techno-Authoritarianism
In “The Rise of Techno-Authoritarianism,” Adrienne LaFrance argues that technocrats use the language of the Enlightenment but “in fact, they are leading an antidemocratic, illiberal movement.” LaFrance breaks down this authoritarian politics that, for example, “profess unconditional support for free speech, but are vindictive toward those who say things that do not flatter them.” At the heart of the autocratic movement among tech elites is the eccentric belief system that “technological progress of any kind is unreservedly and inherently good; that you should always build it, simply because you can” and that their power should therefore be unconstrained — that, according to LaFrance, they should be able to “impose these beliefs on the population, which is neither consulted nor usually, meaningfully informed.”
I write Untangled because:
We need to call out the belief systems that put technological progress above societal progress.
We need to make technological systems more democratic.
We need to describe the world we want and then map back to the kinds of technologies that might contribute to such a world.
This might sound abstract but LaFrance concludes by making a tangible call to action:
“We do not have to acquiesce to their growing project of dehumanization and data mining. Each of us has agency…Every day we vote with our attention; it is precious, and desperately wanted by those who will use it against us for their own profit and political goals. Don’t let them.”
If you’d like to dig into the “techno-determinist” mindset that LaFrance alludes to in her article, read this gem from the archives.
Or perhaps you want to consider what it would look like to imagine alternative worlds and map backward to the role of technology — in that case, start here.
🤬 Kids Online
Keep reading with a 7-day free trial
Subscribe to Untangled with Charley Johnson to keep reading this post and get 7 days of free access to the full post archives.