While the post is cringe inducing and dawkins is an ass, he doesn’t quite say that claude is sentient. More that the ways we use to define it are all basically met by an llm. And argues that we need a better way to define it.
It’s pretty simple, what’s the LLM thinking about when it’s not actively being prompted?
Does an LLM have the same “consciousness” when talking to someone else? Does the LLM even exist if it’s not actively being talked to? If everyone just stops talking to it, does it “die” until someone else logs in and submits a prompt?
I’m not defending AI consciousness/sentience (however we define them), but IMHO your arguments aren’t particularly relevant/convincing. Why couldn’t consciousness exist in brief flashes? Isn’t the “would it even be the same consciousness” same as any sci-fi “clone with identical memories” thought experiment: if a LLM were conscious then each conversation history would be a “self”. And yes, it would be in some vague state of death/hibernation/non-existance when not propmpted.
Has someone done a study on this? I’ve noticed everyone who claims to have a relationship with, if they’re male it’s female and vice versa. This includes married people and those in relationships
You can’t retain your level of intelligence forever, unfortunately. He’s probably heading the same way I’ve seen people I watched get older and slip into things that I know they would have laughed off years before. Now he may have been a bigot and an asshole before, but him not understanding LLMs and their limitations suggests a definite decline in reasoning, from the person who helped me grasp evolutionary theory and the scale of time with his books.
He posted some Tweets about how he believes ChatGPT (or whichever AI service he uses, I can’t recall) is sentient, and of course, female.
While the post is cringe inducing and dawkins is an ass, he doesn’t quite say that claude is sentient. More that the ways we use to define it are all basically met by an llm. And argues that we need a better way to define it.
Basically arguing about a philisophical zombie
https://unherd.com/2026/04/is-ai-the-next-phase-of-evolution/#comment-1031777
(paywalled, but can be mitigated by the firefox reader mode, for example)
It’s pretty simple, what’s the LLM thinking about when it’s not actively being prompted?
Does an LLM have the same “consciousness” when talking to someone else? Does the LLM even exist if it’s not actively being talked to? If everyone just stops talking to it, does it “die” until someone else logs in and submits a prompt?
I’m not defending AI consciousness/sentience (however we define them), but IMHO your arguments aren’t particularly relevant/convincing. Why couldn’t consciousness exist in brief flashes? Isn’t the “would it even be the same consciousness” same as any sci-fi “clone with identical memories” thought experiment: if a LLM were conscious then each conversation history would be a “self”. And yes, it would be in some vague state of death/hibernation/non-existance when not propmpted.
Claude.
It would be hilarious if he called ChatGPT Claudia though. If it was actually sentient, it might get offended.
Has someone done a study on this? I’ve noticed everyone who claims to have a relationship with, if they’re male it’s female and vice versa. This includes married people and those in relationships
You can’t retain your level of intelligence forever, unfortunately. He’s probably heading the same way I’ve seen people I watched get older and slip into things that I know they would have laughed off years before. Now he may have been a bigot and an asshole before, but him not understanding LLMs and their limitations suggests a definite decline in reasoning, from the person who helped me grasp evolutionary theory and the scale of time with his books.