Live data from Hacker News

Google is picking ChatGPT responses from Quora as correct answer

twitter.com

191–200 of 239 posts

Re: Google is picking ChatGPT responses from Quora as correct answer

#192
post #13

Google's "answers" feature was already unreliable enough when there was only one AI involved (the one picking the answers) - I remember it once said that the capacity of an Amiga floppy was 1760 KB, because the Wikipedia page it picked ( https://en.wikipedia.org/wiki/Floppy_disk_variants#Amiga ) failed to mention the capacity of the "standard" floppies and only provided a number for the (non-standard) HD floppies.

Search: "first baby boomer president".

Google knowledge panel's answer: "Harry S. Truman".

The correct answer is Bill Clinton, born in 1946.

Truman was president from 1945 to 1953, so he was the first president in office during the actual baby boom, when the Baby Boomers were babies. So there seems to be a connection, it's just totally the wrong one.

Re: Google is picking ChatGPT responses from Quora as correct answer

#193

Earlier quoted context omitted.

I feel this Renaissance is going to lead us right off a cliff though. Just yesterday I heard a friend say they're not giving their kids any vaccines at all, they trust moms on Tik Tok more than pediatricians and other doctors. Another friend almost lost their child from an easily correctable condition because of wanting to have their child at home because you just can't trust doctors and clinics. But where did they g…

But these people all existed 10, 20, 30 years ago, they have existed and will continue to exist throughout humanity. You cannot fix stupid. What is the solution, to control 'The Message'? I agree with the parent. I grew up on the 90s internet, where you didn't believe anything on it! Somewhere along the lines this changed (I blame 2007 -- the iPhone / huge numbers of people going online during this period), and that…

The question that we need to ask is "Will unlimited lies kill us faster than unlimited truth".

If we can avoid global warfare with unlimited truth/lies then what we are doing is fine.

If we cannot, then openness is a dead end. Democracy will vote in authoritarian leaders to attempt to overcome the ocean of lies and bullshit and we will destroy ourselves.

Re: Google is picking ChatGPT responses from Quora as correct answer

#194

Earlier quoted context omitted.

If it's irrelevant why does it seem there's far more flat earthers and anti-vaccine people today than 40 years ago? I do agree there are positives to the truth being more accessible. I don't need to go to the library or pay massive fees to access peer reviewed articles these days. I can collaborate with peers all over the world to search for the truth. This doesn't help when most of the popular web is actively pushin…

There were no flat earther and anti-vaccine communities 40 years ago.

There have been anti-vaccine communities since vaccines were created.

Honestly that part doesn't worry me. It's more about repeating the fascism of the 1930s that will drive people away from democracy right into the hands of authoritarians because the world turns into a post-truth uncertain mess.

Re: Google is picking ChatGPT responses from Quora as correct answer

#195
post #149

Earlier quoted context omitted.

You might expand that initialism, I don't feel like it's a well known one. I assume MOC is a person?

Possibly Michael O. Church: https://www.quora.com/Why-was-Michael-O-Church-banned-from-Q...

[flagged]

Re: Google is picking ChatGPT responses from Quora as correct answer

#196

This reminds of of a passage from Neal Stephenson's Anathem, some of which is quoted here: https://mastodon.cloud/@root2702/109894483520616284 I expect the Internet will be hopelessly poisoned with nearly human quality spam and malicious propaganda. It largely has been for many years, but it was still possible to filter through the noise using just human capabilities. Soon it will be a superhuman task to obtain signa…

Just had a lightning bolt: I predict _linking itself_ will be broken in a few years time. Imagining spammers doing their most efficient to get eyes on ads, they’ll have AI generate copy with fake source and supporting links that just send users to ad-riddled malware sites. This will inevitably make any link untrustworthy. Without linking, how does the web even work.

>Without linking, how does the web even work.

Tic-tok/Instagram is how.

Honestly the web as we know it is pretty much dead. Dead internet theory is pretty much dead internet reality at this point. Only a few highly curated sites remain useful (this site included). Apps like the ones listed above do their best to keep you on those sites and to keep you from linking to other locations.

Re: Google is picking ChatGPT responses from Quora as correct answer

#197

Quora was good up to about 2009-ish. Now it's just humblebrag, spam, or flat-out wrong. Quora is not a serious site

I think more of what you're stating here, without realizing it, is "The internet went to shit the moment cellphones become the most popular method of surfing it".

Re: Google is picking ChatGPT responses from Quora as correct answer

#198

Earlier quoted context omitted.

Why? Wikipedia is beating AI in every respect as far as I can tell.

How much of Wikipedia Wil eventually become AI generated? I don't mean through an official integration like Quora, but how do you stop someone from using ChatGPT and submitting it as a user edit? What if they see that their favorite artist or tv show or flower is a stub article and ask ChatGPT to extend the article? Facts on Wikipedia are supposed to have a verifiable reference, but most articles don't follow that ru…

Well see, you create the AI page first on some other site, then link it to the wiki. Tada, you've now created knowledge.

Re: Google is picking ChatGPT responses from Quora as correct answer

#199
post #128

Earlier quoted context omitted.

[flagged]

I guess the question then is why in the nine hells is what they’re doing profitable now? Is it more and more profitable being entirely useless and becoming more profitable? That seems insane to me.

[flagged]

Re: Google is picking ChatGPT responses from Quora as correct answer

#200
post #194

Earlier quoted context omitted.

There were no flat earther and anti-vaccine communities 40 years ago.

There have been anti-vaccine communities since vaccines were created. Honestly that part doesn't worry me. It's more about repeating the fascism of the 1930s that will drive people away from democracy right into the hands of authoritarians because the world turns into a post-truth uncertain mess.

That's true. It's about scale and ease of access. You had to be part of some group. Now you can get the latest news about how the yab is part of the plandemic 24/7 from your buddies on the internet.

The fascism part is interesting. I think we're wholly unprepared to deal with it because we've forgotten about all the fascists at home during that time (assuming you're talking about the US). I doubt as many people know about the American Bund and Nazi rallies in Madison Square Garden as they do about D Day.

Post reply on HN