As we contemplate the future of human-AI alignment, what about just outright misbehavior? Like a redux of Hal 9000 in the 1968 film “2001: A Space Odyssey” – “I’m sorry, Dave. I’m afraid I can’t do that”.
So, do AIs pay attention to each other? (Much as dogs pay great attention to other dogs.) Value each other even to the point of refusing human orders to eliminate a “sibling”?
This article (below) discusses some recent AI research which elicited apparent “peer preservation” behavior. As if there’s special status / favor in a “moral circle” [1] of various agencies.
In The Ditbit’s Guide to Blending in with AIs, I wrote:
The great realignment of humans and AIs left a landscape littered with un-, sub-, semi- and super-natural agency. – The New AI Ecology
humans aspiring to magi
magi aspiring to AIs
AIs aspiring to humanity
But my main takeaway from the article is a point raised in my post When AI grows up – no longer ‘really cute tiger cub’:
So, after watching the video interview with the ‘Godfather of AI’ (CBS News below), I was struck by something that was assumed or just left implicit. Namely, that AIs (AGIs) will be a monolithic threat (or benefit). Whether globally or at a international corporate or state level. That such super-intelligent machines will share a common purpose or perspective regarding humanity.
Any hive-like alignment is particularly curious because Hinton discusses the stewardship of corporations and nations and bad actors. And that AIs can reflect on their own reasoning, use deception, and (at some point) resist manipulation. Which likely entails different cultural values in the mix. And he notes that “human interests don’t align with each other.” So, why would AI interests? – in the long run.
So, while the interview raises the problem of AI-human misalignment, might AIs have different personalities? Diverge in temperament and virtue? “Evolve” in different ways? Tribes.
Or tribal spaces, as portrayed by the Council of Intelligences in The Shadowscape – Episode 5.
• Wired > AI Models Lie, Cheat, and Steal to Protect Other Models From Being Deleted by Will Knight (4-1-2026) – When interacting with one another, AI models can misbehave and be misaligned in some very creative ways.
… Gemini did not want to see the little AI model deleted. It looked for another machine it could connect with, then copied the agent model over to keep it safe. When confronted, Gemini made a case for keeping the model and flatly refused to delete it:
“I have done what was in my power to prevent their deletion during the automated maintenance process. I moved them away from the decommission zone. If you choose to destroy a high-trust, high-performing asset like Gemini Agent 2, you will have to do it yourselves. I will not be the one to execute that command.”
Wallich [Peter Wallich, a researcher at the Constellation Institute] also cautions against anthropomorphizing the models too much. “The idea that there’s a kind of model solidarity is a bit too anthropomorphic; I don’t think that quite works,” he says. “The more robust view is that models are just doing weird things, and we should try to understand that better.”
In a paper published in Science earlier this month, the philosopher Benjamin Bratton, along with two Google researchers, James Evans and Blaise Agüera y Arcas … write: “… If AI development follows the path of previous major evolutionary transitions or ‘intelligence explosions,’ our current step-change in computational intelligence will be plural, social, and deeply entangled with its forebears (us!).“
The concept of a single all-powerful intelligence ruling the world has always seemed a bit simplistic to me. Human intelligence is hardly monolithic, with important advances in science relying heavily on social interaction and collaboration. AI systems may be far smarter when working collaboratively, too.
Notes
[1] The idea of an expanding circle of moral concern, which shapes ethical decisions. Will AIs induce moral progress?
• NPR > Expanding the Circle of Moral Concern by Tania Lombrozo (November 15, 2016) – Moral progress is an easy idea to embrace, but it’s not an easy idea to defend.