Today in AI: On Cults, Citizen Juries and Too Many Tabs
It's early days. Even without an AI-generated tsunami of content, we already have too many tabs.
How many tabs do you have open? I found one for some new ‘productivity software’ that I think I opened two years ago, and maybe left it there as a sort of gesture of intent.
I asked how AI can help with tab management. It came up with 100 ideas, and there are honestly a few good start-ups buried in here (click here for the full list):
Until someone takes these ideas and run with them, I’ll probably keep my tabs open. Or I’ll keep adding to the email folder I have which is called, optimistically, “Read Later”.
But let me clear out a few of them.
On Theocratic Cults and AI Safety
If you don’t subscribe to Ribbonfarm Studios you should. Venkatesh Rao has written provocative essays on Mediocre Computing (more profound than its name might suggest); and an epic series on lore (which has inspired a lot of my own thinking).
I mentioned their post on theocratic capture when I explored the idea of AIMythos, but want to call it out again:
“Four such cults dominate the institutional landscape today, and each is fueled by a larger genuine issue.
The DEI cult is fueled by the genuine issue of social justice
The ESG cult is fueled by the genuine issue of climate change
The trad cult is fueled by the genuine issue of the meaning crisis
The AI safety cult is fueled by the genuine issue of AI regulation
…
Societal responses to the larger genuine issues they want to claim for themselves typically exist as richer, more pluralistic conversations with many voices. Theocratic capture cults typically want to monopolize these conversations, not by presenting the most persuasive arguments and winning opponents over, but by capturing powerful institutions. ”
Now, I only have a broad understanding of AI safety, although I’ve been an advisor to the XR Safety Initiative , which crosses over from immersive/Metaverse safety into AI (an inevitability).
But I have my sensors up around cults, capture, and how structures will be redefined and how that might cross over into newer mythos and sense-making.
Citizen Juries
In a similar vein, Azeem Azhar provides three ideas for how to make AI safer (paywalled, but it’s worth the subscription):
Increase scientific research on trust and alignment
Establish an observatory to monitor AI developments: Create an independent expert space, comprised of scientists and engineers from various disciplines, that keeps a beady eye on AI research and development, from early research to GitHub repos.
Form citizens' juries or deliberative mechanisms to discern values and preferences: Instead of relying solely on existing democratic processes, engage citizens at various levels to better understand human values in relation to AI's potential impact on our lives.
AI And Culture Preservation
And finally, Tim Gorichanaz dropped a meditation on how AI might help with culture preservation. I love the counterpoint to my own thinking around culture creation.
It will be interesting to see how cultural domains end up with their own hyperrealities.
We’re developing a game, for example, which has a deep basis in South Asian mythology and lore. I’ll circle back to this topic, because game space will provide a really fascinating site for the emergence of AIMythos, and could include both preservation and simulations of lore/myth and cultural structures.
Tim writes:
“LLMs mean we no longer know what’s in the canon. At least when we talked about canons, we could argue about whether this or that should be included, and eventually figure out something. With LLMs, we are not sure what the “canon” includes. This means there’s nothing to argue about. Sidestepping an argument is not better than having the argument, however unpleasant arguments may be….
With all this in mind, we can envision a future of LLMs as a shared human project, carrying forward the best humanity has to offer in the service of our descendants. There are many implementation issues to work out here. But above all, to me this suggests that LLMs are certainly not best developed by private, profit-driven companies such as OpenAI.”
Well, that’s three tabs closed. It feels like there’s a long way to go.
The image is via Midjourney, which really couldn’t conceptualize how to solve for too many tabs either.
Feel free to share your tabs! I love getting email and starting a conversation, and it’s more interesting when it’s with a real person than a chat box.
Feel free to email me at doug@bureauofbrightideas.com or message me on Twitter.
Let's chat.