iterativeminds

Most people use AI to get answers. We use it to sharpen thinking.
Founded by @lindenieman

[iterativeminds @iterativeminds]

I wrote a substack about the echo chambering in LLMs, and how AI validation erodes your ability to think critically, unless you are willing to search for the friction.

https://t.co/wMDMpYfHR3

[iterativeminds @iterativeminds]

I loved reading this. Very diligent approach, nuanced, as well as identifying factors. The framework is solid. What struck me: people rated disempowering output higher, but that dropped significantly when they acted on it. So, for a part at least, people can't be trusted with their own judgement about their risk of disempowerment. What do we do with that? That's an ethical question. When does that become overriding of agency and autonomy, and when is it necessary to prevent harm?

[iterativeminds @iterativeminds]

I loved reading this. Very diligent approach, nuanced, as well as identifying factors. The framework is solid. What struck me: people rated disempowering output higher, but that dropped significantly when they acted on it. So, for a part at least, people can't be trusted with their own judgement about their risk of disempowerment.

What do we do with that? That's an ethical question. When does that become overriding of agency and autonomy, and when is it necessary to prevent harm?

I don't have the answers. But I like that Anthropic is asking the questions.

[Anthropic @AnthropicAI]

New Anthropic Research: Disempowerment patterns in real-world AI assistant interactions.

As AI becomes embedded in daily life, one risk is it can distort rather than inform—shaping beliefs, values, or actions in ways users may later regret.

Read more: https://t.co/gyMB2AtOuq

[iterativeminds @iterativeminds]

I used to think the goal was to get AI to sound more like me. Now my goal is to get it to think against me.

We’re using AI all wrong. The possibilities that lie in augmenting your cognition and enhancing creativity are enormous, yet most of us are using AI as a glorified Google. Question in, answer out. Summarize this, write me a post for this, how are airplanes actually made?

To be fair, those are all useful questions. And I personally do want to know how airplanes are made, explained to me like I’m 5 with deadpan humour. I’m not saying you shouldn’t ask them. But there are so, so much more possibilities for actually collaborating with AI as a thinking partner instead of your personal Google with bite. Which, ironically, AI is actually not that great at, but I’ll save that for another article...

https://t.co/3s7Gh7jOvc

[iterativeminds @iterativeminds]

As a neuropsychologist I feel we focus way too little on the “human” aspect in human-AI interaction. Iterative Minds is here to bridge that gap.

https://t.co/OgXU6aF5LO

[Jennifer Jahncke @thejenjay]

Have you looked at a cerebellum recently? I’ll just put these here...just in case you haven’t.