3 Comments
User's avatar
Cyril Simonnet's avatar

INTERNET OF THE MIND @ SERENITY CAFE, your point about disputed sources and the human challenge of seeking truth together hits something I keep running into on the security side. When you talk about Claude Mythos and authentic intelligence in an age of distrust, you're describing the same failure mode I see in threat detection: the model confidently synthesizes from a shared corpus, and nobody checks where the confidence breaks down. My piece agrees with yours but frames it from the operator's angle. The edge isn't in having AI, it's in knowing exactly where your AI is wrong, because everyone trained on the same research. Curation of blind spots is the real trust problem.

https://cyrilsimonnet.substack.com/p/everyones-ai-was-trained-on-the-same

Infinite Improbabilities's avatar

I will admit that when I opened this article, I sort of wanted to read it angry. I opened it with all the empathy and understanding of someone getting another letter for shady credit card pre-approval in the mail. And then I realized that if that was the attitude I was going to approach this with, I'd probably better take a breather and remember one of my own deeply held morals. I don't set my definition of a higher good by an established god or school of thought, so I'll describe this piece of it directly. "Sometimes you can disagree with someone on something and still think that they're intelligent, and also that they are more intelligent than you about certain things that aren't what you firmly don't agree with". We all specialize. I personally come from a bias of vehemently shutting down any conversation that I feel like starts with "just trust me" whether or not someone may actually know what they're talking about.

In an attempt to live in a world where I can still have my biases but be comfortable enough that you can't force me not to, I set down my reactive anger. And I'm glad I did that at least long enough to see that this article touches on a lot of what had gotten me stirred up enough to challenge you so heavily on your last post, and it does it in a way that reminds me that you're not necessarily plugging you ears, you're focusing on something true and important from an angle that requires the assumption of a belief I don't have. Maybe other people who see these comments I've gotten into the habit of leaving will relate.

I think that's just about exactly why I feel the need to challenge you on your perspective that says we don't need resolute answers. While I agree with you that matters of real importance require deep slow thinking and cross referencing sources carefully, I also think it's highly relevant to the conversation here that we're having this discussion in a world (at least in the USA, which for me might as well be "the world") that has a strong "emergency culture". We're having these conversations from the context of a large population of people who live in survival mode. I'm not going to try to act like I live in growth mode as often as I would like, and so I'll instead tell you that I think any conversation on AI that happens slowly and calmly to a wide diverse audience of people may be reactive too, is going to have to bridge that gap.

I personally have found that a big part of what moves me from drama to growth is the acknowledgement that there are very few societal problems right now that you would benefit from panicking about. Not because we're powerless, but because we aren't. Real change comes from people who believe they can accomplish something and work together to do it. It's easier to hold that conversation that asks what we want to do as an empowered people when it's started with a real and grounded perspective that says "yes this is a problem, and we aren't dead yet, and while it's not likely someone is going to come up with a perfect unanimous answer by posting on the internet, we can all bounce answers around until we get one we think is good enough to be better than nothing".

I think that avoiding certainty is excellent, but avoiding explicit ideas on what could be better and how isn't going to help as much. Maybe that's the nature of a platform that doesn't have many different voices yet, and that you're trying to make sure new voices even if they may differ are still welcomed. At the same time, in a world where not everyone is going to be comforted by the divine, and with an understanding that a big issue here is dysregulation, we have to ask questions from a human scale of what we as human people can do.

Faith in the divine isn't avoiding certainty either, in my opinion. It's just getting your certainty needs met by a god that not everyone believes in. I don't thinks that's going to be enough for everyone in the country that makes it explicitly clear that faith is a choice. I feel like it might be relevant to ask less "where do we seek our comfort from about frightening news" and more "how do we turn down the heat on the frightening news" so that we can all think. At the same time, both matter. I just happen to focus more on the second thing, and so it's more irritating to me in my fear and reactivity when you don't agree with me on where to start.

I look forward to seeing this next article you intend on writing. The defiant child in me looks forward to picking apart more ideas, and the wise adult in me will make sure that the pursuit of a good intellectual debate isn't overshadowed by the very human need to feel like I can protect myself from "bad ideas" just by shouting my own also biased ones over them louder.

Don Carter MSW, LCSW's avatar

This comment is exactly what we need (IMHO): "Real change comes from people who believe they can accomplish something and work together to do it. It's easier to hold that conversation that asks what we want to do as an empowered people when it's started with a real and grounded perspective that says "yes this is a problem, and we aren't dead yet, and while it's not likely someone is going to come up with a perfect unanimous answer by posting on the internet, we can all bounce answers around until we get one we think is good enough to be better than nothing". Bravo! Thanks for sharing.