Post

Preprinted! Been working on this for ~1 year w wonderful coauthors Nancy & Mark: > Understanding Artificial Neural Networks: Mysterianism about Known Mechanism is Mysticism doi.org/10.5281/zeno... Have you noticed many experts say they don't understand AI? What's going on? 🤯 Let us tell you... 1/

Olivia Guest · Ολίβια Γκεστ

@olivia.science

8 May

Here are some flagrant examples, although the appendix is full of them of them too — basically experts and journalists alike seem to repeat this mantra that we do not understand how contemporary AI works. See appendix at the end here for more: doi.org/10.5281/zeno... 2/

Olivia Guest · Ολίβια Γκεστ

@olivia.science

This hangs with the other statements in fig 1: 1️⃣ ANNs (artificial neural networks) are inspired by human organism 2️⃣ ANNs are a black box we don't understand 3️⃣ We understand neither ANNs nor cognition therefore they are similar All 3 are misplaced! We unpack this... doi.org/10.5281/zeno... 3/

8/5/26, 11:53 am47 reposts236 likes

Mian Wang

@windssw.bsky.social

1 Sept

I experimentally demonstrated that neural networks are classical stateful dynamical systems. In my framework, training, learning, inference, and feedback are different parts of the same state evolution. Paper: arxiv.org/abs/2608.20965 Experiments/code: github.com/wind342/gfg-... Discussion welcome

arxiv.orgTraining, learning and inference: unified dynamics of neural systemsWe define an atomic generation fact f=(u,tau,omega,z;rho), recording the origin, realized transformation, concrete occurrence, generated result and relation role. Compiled into a Generation-Fact Graph...

Olivia Guest · Ολίβια Γκεστ

@olivia.science

8 May

Pausing and will be back later (to avoid getting overwhelmed as well as because I am hungry)! But if you're itching to know more check the PDF obviously. 😌 4/

Olivia Guest · Ολίβια Γκεστ

@olivia.science

9 May

OK, let's talk about 3 important concepts: ANNs, LLMs, & AI: "Definitions of AI and LLM are hard to pin down [&] ANN is an umbrella term for models" Black box: Really not what people think because actually you can open it (read on in the PDF to find out) by design! doi.org/10.5281/zeno... 5/

Olivia Guest · Ολίβια Γκεστ

@olivia.science

9 May

Leibniz' Mill: "argument against the idea that mechanical components suffice to understand cognition" All are important & misused (1st 2) or forgotten about (last) to pretend we don't understand AI. OK & lil peek ahead to appreciate the pun this enables in section 4 (doi.org/10.5281/zeno...)! 6/

Olivia Guest · Ολίβια Γκεστ

@olivia.science

9 May

Before we use those 3 to tease apart what's going on, however, we have to contend with where we are. In 2010s we actually did avoid these traps, yes, even industry! “the most extreme promises of AI are based on a flawed premise: that we understand human intelligence” doi.org/10.5281/zeno... 7/

Tim Crane

@timcrane.bsky.social

8 May

Thanks Olivia this looks great

Olivia Guest · Ολίβια Γκεστ

@olivia.science

8 May

Oh, thank you!!

Tim Crane

@timcrane.bsky.social

8 May

Looking forward to reading the paper

Mark

@aguasonic.com

15 May

3️⃣ We understand neither ANNs nor cognition therefore they are similar 🤣

Tony Durham

@rhamdu.bsky.social

29 May

re [2] did you see my comment bsky.app/profile/rham...

Tony Durham@rhamdu.bsky.social

Is AI mysterious? No and yes. We know exactly how machine learning works. But what the machine eventually learns can be mysterious. AFAIK we don't yet have a human-intelligible account of how Alphafold predicts protein configurations.

Olivia Guest · Ολίβια Γκεστ

@olivia.science

29 May

I wrote everything I think in that preprint if useful

Craig Walmsley

@craigwalmsley.bsky.social

13 Sept

3 is such a logical howler it is painful to see in print. But I’ve actually seen notionally professional philosophers aver such rank twaddle. It’s all a bit depressin.

Olivia Guest · Ολίβια Γκεστ

@olivia.science

13 Sept

Yeah, it's a type of argument from ignorance or false analogy.

Olivia Guest · Ολίβια Γκεστ@olivia.science

Also it is consistent with false analogy, which assumes two things are similar (when they aren't) in deep ways and causes immense confusion if not detected in one's thinking. doi.org/10.1007/s421...