Hacker News new | ask | show | jobs
by oytis 48 days ago
Its in the headline. Also he talks about the persona he assigned to his chat like "she" was conscious (e.g. "she was pleased")
1 comments

That's a good example of my point about reading comprehension. The headline is "When Dawkins met Claude Could this AI be conscious?".

That's a question, not a statement. By Betteridge's Law of Headlines, which states that any headline ending in a question mark can be answered "no", this would even justify claiming that he was denying that Claude was conscious.

But he isn't making either claim; instead, he's asking the much more interesting questions: if p-zombies are possible, should we expect them to be more or less likely to evolve? Why? What is the difference? Why does it matter to evolution?

They seem to have changed the headline. The one in the archived article the post quotes is "Is AI the next phase of evolution? Claude appears to be conscious". Again, "appears to be" is not exactly the same as "is", but the post in question also quotes his Twitter extensively, and it's clear that Dawkins is acting as if he believed in Claude's consciousness.
Citation needed. All of the direct quotes I've seen have clearly stated that he can not _disprove_ the claim of consciousness, and finds this fact interesting.
What gurantees Betterridge? What is the universal basis upon which no can safely be assumed in such formulations?
Obviously, there is none. The point is, assuming the answer to be "yes" isn't a slam-dunk, and (in general) may not even be a good bet.

(The mechanical reason for Betteridge being true more often than not is that if a journalist want to make a claim but can't (because the facts don't support it) they frequently phrase it as a question. If the thing they want to imply were true, they'd just say it.)

Ok, that makes sense