FilterHN

Ask HN: Are LLMs Just Philosophical Zombies?

2 points

1 hour ago

| 2 comments

The whole point of LLMs is to fake being human through following the likely paths text will follow. Is that not what a philosophical zombie[0] is: appearing to be human while having no actual intelligence?

[0] https://en.wikipedia.org/wiki/Philosophical_zombie

▲

jqpabc123

1 hour ago

[-]

It's not actually "artificial intelligence" but more like "fake intelligence".

Assembling a plausible sounding sentence doesn't mean that you know what you're talking about.

The number of people who fail to grasp this is mind boggling.

▲

pixel_popping

1 hour ago

[-]

And how do you explain that they can do advanced mathematics?

▲

jqpabc123

18 minutes ago

[-]

How do you explain when they can't? If they *really* understand math, why is the error rate so high?

According to a 2025 Stanford HAI report, large language models fail basic multi-step arithmetic up to 40% of the time without external tools.

https://medium.com/@dojolabs.main/why-does-ai-get-math-wrong...

▲

pixel_popping

7 minutes ago

[-]

2025... have you checked latest models? and even, this talk is irrelevant when we know in few months/years this exact topic will be solved.

▲

Trasmatta

1 hour ago

[-]

> appearing to be human while having no actual intelligence?

Isn't a P-zombie about consciousness, not intelligence?