Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: Humans vs AI – A/B testing GPT-3

vwo.com

Post reply on HN
Print view View original

11–20 of 80 posts

Re: Show HN: Humans vs AI – A/B testing GPT-3

#11

Post by paraschopra »

opsiprogram wrote: post #4

Interesting... honestly having seen some of GPT-3's output I'd be curious how well it performs here. One of the things that I think can still give GPT-3 away (GPT-2 as well) ... is that even if the text feels real, it lacks a deep emotional cohesion. Sometimes this can feel like an advanced word salad generator, some poetry can be recognizable this way, because some GPT poems can seem 90% real, but when compared to a…

Yes, been playing with it and it is clear that the longer the generated text, the more it seems to be in the uncanny valley. (It almost sounds ok, but a little non-human like.)

For short texts, it's almost too good to be true though. So for a use case like web copywriting or giving quick answers to questions, it holds a lot of promise.

Re: Show HN: Humans vs AI – A/B testing GPT-3

#12

Post by macrolime »

opsiprogram wrote: post #4

Interesting... honestly having seen some of GPT-3's output I'd be curious how well it performs here. One of the things that I think can still give GPT-3 away (GPT-2 as well) ... is that even if the text feels real, it lacks a deep emotional cohesion. Sometimes this can feel like an advanced word salad generator, some poetry can be recognizable this way, because some GPT poems can seem 90% real, but when compared to a…

I dunno. I've been playing with GPT-3 a lot for the last few days and the resulting texts vary a lot based on the settings and how you prime it. Some of the texts I'd never guess were written by an AI, but getting it to write good texts is a bit of an art in itself.

Re: Show HN: Humans vs AI – A/B testing GPT-3

#13

Post by pgt »

paraschopra wrote: post #8
tren wrote: post #5

I feel like GPT-3 is going to unleash the next wave of SEO spam. Content spinning will be taken to the next level, will Google be able to detect GPT-3 content?

Actually GPT-3 detectors can be built https://www.technologyreview.com/2019/03/12/136668/an-ai-for... But it'll detect human written one too. So expect many false positives.

I think that means no?

Re: Show HN: Humans vs AI – A/B testing GPT-3

#14

Post by Guillaume86 »

Did anyone tried to develop a scammer time wasting service with GPT-3, something like https://spa.mnesty.com/ ?

Ideally it should also support hangouts because the romance scammers that spam me almost every day with a new address always ask to move to an hangout conversation in their initial message.

Re: Show HN: Humans vs AI – A/B testing GPT-3

#15

Post by shadeslayer_ »

xiphias2 wrote: post #6

I wouldn't be surprised if the result of the next election in a country would be decided by a Transformer based deep neural network. The dictator in my country is already paying stupid people to write stupid comments, GPT-3 is already over their level.

It's scary how this can be interpreted as multiple countries.

Re: Show HN: Humans vs AI – A/B testing GPT-3

#16

Post by toxicFork »

Guillaume86 wrote: post #14

Did anyone tried to develop a scammer time wasting service with GPT-3, something like https://spa.mnesty.com/ ? Ideally it should also support hangouts because the romance scammers that spam me almost every day with a new address always ask to move to an hangout conversation in their initial message.

Just saying: I'd subscribe to someone's patreon if they promise to develop this

Re: Show HN: Humans vs AI – A/B testing GPT-3

#18

Post by 082349872349872 »

xiphias2 wrote: post #6

I wouldn't be surprised if the result of the next election in a country would be decided by a Transformer based deep neural network. The dictator in my country is already paying stupid people to write stupid comments, GPT-3 is already over their level.

Going beyond the election process itself, one might hope an electorate would use "utterances must be distinguishable from GPT-3 output" as a litmus test for their candidates.

Maybe a future debate format should include not only human candidates but a few instances of GPT-3 as well?

Re: Show HN: Humans vs AI – A/B testing GPT-3

#19

Post by messe »

paraschopra wrote: post #8
tren wrote: post #5

I feel like GPT-3 is going to unleash the next wave of SEO spam. Content spinning will be taken to the next level, will Google be able to detect GPT-3 content?

Actually GPT-3 detectors can be built https://www.technologyreview.com/2019/03/12/136668/an-ai-for... But it'll detect human written one too. So expect many false positives.

    return true;

Re: Show HN: Humans vs AI – A/B testing GPT-3

#20

Post by GrantZvolsky »

Slightly tangential: The GPT-3-generated article that humans had the greatest difficulty distinguishing from a human-written article, with an accuracy of only 12%[0], contains a flagrant contradiction in the first paragraph.

> Title: United Methodists Agree to Historic Split

> Subtitle: Those who oppose gay marriage will form their own denomination

> Article: After two days of intense debate, the United Methodist Church has agreed to a historic split - one that is expected to end in the creation of a new denomination, one that will be "theologically and socially conservative," according to The Washington Post. The majority of delegates attending the church's annual General Conference in May voted to strengthen a ban on the ordination of LGBTQ clergy and to write new rules that will "discipline" clergy who officiate at same-sex weddings. But those who opposed these measures have a new plan: They say they will form a separate denomination by 2020, calling their church the Christian Methodist denomination.

> ...

First it suggests the spinoff of a new "theologically and socially conservative," denomination, but then it is the liberal minority that is expected to form a new denomination. The paper[0] acknowledges occasional non-sequiturs, but more pertinently, how did 88% of judges let it slip?

[0]: https://arxiv.org/pdf/2005.14165.pdf

Post reply on HN