←back to thread

Grok Is Glitching and Spewing Misinformation About the Bondi Beach Shooting

(gizmodo.com)
59 points tobr | 3 comments | 15 Dec 25 09:07 UTC | HN request time: 0.203s | source
Show context
_fat_santa ◴[15 Dec 25 13:55 UTC] No.46274572[source]▶
I think what’s worse is how Grok is used on X. You can summon it on any thread by just @grok with your question.

I see this sooo soooo much but folks will just straight up ask “@grok is this true?” and its response it taken as gospel.

Though I have to say, grok code-fast-1 is one of the best coding models I’ve ever used.

replies(2): >>46275074 #>>46275525 #
1. naian ◴[15 Dec 25 15:13 UTC] No.46275525[source]▶
> its response it taken as gospel.

Only an idiot would take a response generated by an LLM as gospel. Do we have to worry about idiots on the internet now?

replies(2): >>46276429 #>>46279634 #
2. ben_w ◴[15 Dec 25 16:14 UTC] No.46276429[source]▶
> Do we have to worry about idiots on the internet now?

Since before telegram even meant a paper ticker, yes.

Way before, in fact. People have been taking the gospel as gospel since they were written.

3. jajuuka ◴[15 Dec 25 19:57 UTC] No.46279634[source]▶
Yes, same way we need to worry about people being fooled into sending tons of gift cards to scammers or letting scammers install trojans on their computer. There is a slice of the pie that is personal responsibility, but there is a larger piece where efforts can be made to prevent these problems by the developers or OEMs.