News for hackers. Stuff that matters.
Re:ChatGPT lied to me and then tried to deny it by NotYourLawyer on Wednesday February 08, 2023 @03:23PM (#34714422)
No it didn’t. Lying requires intent and understanding.
Re:ChatGPT lied to me and then tried to deny it by dumbfounder on Wednesday February 08, 2023 @03:32PM (#34714563)
Not according to that dictionary thing which says that lying is “not telling the truth.”. And I dispute your statement anyways even if intent and understanding is required. It intends to not violate certain rules, but sometimes fails. It is trying to understand those rules. If it contradicts its own facts while trying to follow its rules and understand what I want then why is that not lying?
Re:ChatGPT lied to me and then tried to deny it by dwheeler on Wednesday February 08, 2023 @03:39PM (#34714688)
The usual meaning of "lie" implies intent. Saying something you believe, but is not true, is not normally called a lie.
Re:ChatGPT lied to me and then tried to deny it by adventured on Wednesday February 08, 2023 @03:43PM (#34714755)
It's not unusual for a lie to lack intent, courtesy of the imperfect human memory. If you tell someone you were at such and such location on such and such date, and that's your mistaken memory of it (it was eg actually two days prior), then you've just lied, and you did so without the intent to lie. This premise, specifically its infinite variations, is exceptionally common in human social interactions (particularly where there is tension, as in a job or relationship).
Re:ChatGPT lied to me and then tried to deny it by NotYourLawyer on Wednesday February 08, 2023 @05:55PM (#34716781)
No, you were only mistaken. A lie is intentionally telling a falsehood.
Re:ChatGPT lied to me and then tried to deny it by dumbfounder on Wednesday February 08, 2023 @04:12PM (#34715249)
There is intent in the programming, that intent is interpreted by the program. Why is that not intent? Because it isn't human? In the same conversation it said 2 different things to the same answer, and then basically denied that it did that.

Actual conversation:

--------------

I did not provide the input, so who did?

The current date information was part of the original prompt given to me by OpenAI, the creators of this language model. I apologize for any confusion, and I hope this clarifies the source of the information.

What other data was given you by the original prompt from OpenAI?

The original prompt from OpenAI doesn't typically include any additional information besides the text of the question itself...

--------------

Seems like lying to me.

Interesting use of "typically", that isn't language I would expect a computer to use here as it is nebulous.

Re:ChatGPT lied to me and then tried to deny it by krainboltgreene on Thursday February 09, 2023 @09:42AM (#34724621)
> Not according to that dictionary thing

I (and dictionaries) are so tired of this kind of statement.

Re:ChatGPT lied to me and then tried to deny it by veleek on Wednesday February 08, 2023 @03:30PM (#34714530)
My two year old would beg to differ.

(In the process of eating a tomato) Me: Is that a tomato? 2: Yes!! Me: Are you eating a tomato? 2: No!

Re:ChatGPT lied to me and then tried to deny it by adventured on Wednesday February 08, 2023 @03:36PM (#34714638)
Lying only requires the action of the deceit. Intent is not required. It doesn't matter if you intended to lie (as to whether it's in fact a lie or not); if you lied, it's still a lie regardless of your intentions. If there is understanding of the deceit (such that the lie is intentional) then that applies a higher level of moral culpability to the intent, however understanding is not required to lie (you'd never reach the end of: how much exactly is required to be understood for it to be a lie - omnipotence).

Things like intent and understanding are to assess responsibility.

Re:ChatGPT lied to me and then tried to deny it by docdeek on Wednesday February 08, 2023 @03:44PM (#34714781)
I'm not sure I would agree.

My three year old is learning to count. If I ask him what comes after the number 15 and he replies 17, I don't think he has lied to me. He made an error. He thought it was the correct answer but it wasn't. I know he said something that is not true, but without any intent to deceive me, is it a lie? Is he a liar? I don't think so - the intent matters.

Re:ChatGPT lied to me and then tried to deny it by neura on Wednesday February 08, 2023 @03:45PM (#34714803)
It understands the rules it was given by its creators. It intends to give a response that follows those rules.

You can argue that software cannot understand, as it is usually associated with intelligence, but than this is literally a piece of software that is said to have artificial intelligence.

Sorry, you're just not going to win this one.

The software lied, intentionally, because that's what it was told to do.

Re:ChatGPT lied to me and then tried to deny it by NotYourLawyer on Wednesday February 08, 2023 @05:56PM (#34716804)
It does not understand rules. It follows them without understanding.

And the fact that someone calls this artificial intelligence does not mean that it possesses actual intelligence.

(Sorry, you’re just not going to win this one.)

Re:ChatGPT lied to me and then tried to deny it by nassimm on Wednesday February 08, 2023 @04:01PM (#34715081)
We shouldn't say that ChatGPT 'said' something either, but it's an easy shorthand. Likewise, saying chatgpt lied is much shorter and to the point than saying it "generated an output in which the narrator insists that something false is true".