Welcome to deep_causality Discussions!

[marvin-hansen]

👋 Welcome!

We’re using Discussions as a place to connect with other members of our community. We hope that you:

• Ask questions you’re wondering about.
• Share ideas.
• Engage with other community members.
• Welcome others and are open-minded. Remember that this is a community we
  build together 💪.

To get started, comment below with an introduction of yourself and tell us about what you do with this community.

[zicklag]

Hey there! I'm @zicklag, who is interested in recently started doing investigation into AI, with more of an interest in learning AI and reinforcement learning than the general "recognize this image/video" type AI that most stuff is built on.

I've got a rough idea of most basic concepts, but nothing much in-depth.

So far it seems to me like active inference might be theoretically better founded than typical reinforcement learning, but I have yet to find an implementation that seems to demonstrate it much effectively for real-world robot-like environments like I'm looking for.

Deep causality is new to me and sounds like it avoids some of the shortcomings of the typical perceptron approach to AI, so I'll be interested to get into it more, and maybe try something out, if I ever get the time.

AI tends to be pretty low priority on my backlog, since I don't have a solid use-case for it yet. 😄

[marvin-hansen]

Thank you @zicklag

I've recently presented to the Linux Foundation and you may want to take a look at the slides to decide if this project is for you:

https://github.com/deepcausality-rs/deep_causality/tree/main/deep_causality/docs/slides/LF_2023

You are correct that DeepCausality addresses some of the rather fundamental issues with DeepLearning (IID, lack of context, lack of causal relationships), but at the same time it's all very early meaning currently models are crafted by hand because causality learning on geometric structures hasn't really been solved. I think it's all elaborated in the limitation document.

Feel free to ask any question that may come up.

Thank you
Marvin