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

The computers are getting better at writing

newyorker.com

Post reply on HN
Print view View original

81–90 of 165 posts

Re: The computers are getting better at writing

#81

Post by Smithalicious »

bertil wrote: post #69
caconym_ wrote: post #66

The lesson we should take from GPT-3 is that we spend way too much time reading text that isn't really saying anything, and/or failing to find meaning in what we read. It's a neat trick, and I think it could absolutely replace much or most of the writing on the internet today. But is it good writing? No, not really. At least, not in the sense of writing as producing anything more substantive than slick prose with a v…

If you want to echo the last paragraph of the article, the system also doesn’t know when to stop. Which leads to a simple improvement: teach the model when to stop, what sections to remove. There’s already great ML-based summary engines, and systems able to answer questions from descriptions.

I wonder if that could be a viable approach, having a separate "writer" and "editor" AI

Re: The computers are getting better at writing

#82

Post by ht_th »

nxpnsv wrote: post #77
ht_th wrote: post #68

Earlier quoted context omitted.

I am using LanguageTool. Not the website, but a command-line tool. You can configure it with all sorts of rules to check all sorts of spelling, grammar, style, and language issues. However, as far as I have noticed, LanguageTool is not a very advanced language-help tool. Still, it did improve my writing a lot. If nothing else, my writing has gotten more consistent.

The cli from here? https://github.com/languagetool-org/languagetool

Yes. You can build the command-line application Java Jar from that repo. I also combine it with vim (https://github.com/dpelle/vim-LanguageTool).

Re: The computers are getting better at writing

#83

Post by wpietri »

caconym_ wrote: post #66

The lesson we should take from GPT-3 is that we spend way too much time reading text that isn't really saying anything, and/or failing to find meaning in what we read. It's a neat trick, and I think it could absolutely replace much or most of the writing on the internet today. But is it good writing? No, not really. At least, not in the sense of writing as producing anything more substantive than slick prose with a v…

> we spend way too much time reading text that isn't really saying anything, and/or failing to find meaning in what we read

Absolutely. It helps make clear that the function of words isn't always what we think. It's the sort of window into the machinery of the mind that helps us reverse-engineer what's really going on.

It reminds me of the Oliver Sacks essay "The President's Speech", which Wikipedia describes as being "about a ward of aphasiacs [people who have lost the ability to understand words] and agnosiacs [here, people who understand only the words, as if reading a transcript] listening to a speech given by an unnamed actor-president, 'the old charmer', presumably Ronald Reagan. Many in the first group laughed at the speech, despite their inability to follow the words, and Sacks claims their laughter to be at the president's facial expressions and tone, which they find 'not genuine'. One woman in the latter group criticizes the structure of the president's sentences, stating that he 'does not speak good prose'."

Glibness has a purpose, but it's not really about conveying information. It's for something else.

Re: The computers are getting better at writing

#84

Post by DyslexicAtheist »

caconym_ wrote: post #66

The lesson we should take from GPT-3 is that we spend way too much time reading text that isn't really saying anything, and/or failing to find meaning in what we read. It's a neat trick, and I think it could absolutely replace much or most of the writing on the internet today. But is it good writing? No, not really. At least, not in the sense of writing as producing anything more substantive than slick prose with a v…

> But is it good writing? No, not really.

agree. I'd say it's impossible to solve. Where would we even place our yardsticks:

Has society been able to get any better (than we used to) at assessing and categorizing what "constitutes" great writing? If so in what relation could that be measured to our past performance especially that many no longer read today due to our loss of attention span? (most won't even read this comment because it's way too long).

People are terrible at spotting some of the best writing when it happens. History is full of writers who's work could only be understood and appreciated after their death.[1] Death makes their work a finite commodity. There is no more chance asking the author "what they meant". Others get shunned because they were too radical, (or too ahead of their times).

Great writers as in "a radical new style" or as "for their radical ideas" get only credited after decades (if not centuries) post mortem. When it's too upsetting (because it's true) we need the distance of history and later generations to appreciate it[2].

Even "what" we comprehend and then appreciate changes over time, and within the same person with every time we read it[3].

[1] 6 people centuries ahead of their time https://www.oxford-royale.com/articles/6-people-centuries-ah... [2] The Last Messiah by Peter Wessel Zapffe https://philosophynow.org/issues/45/The_Last_Messiah [3] James Joyce Dubliners https://archive.org/details/dubliners00joycrich / https://en.wikipedia.org/wiki/Dubliners __

I'm not saying AI is incapable of better writing than AI is now, but it would only remain mediocre and never outperform humans. Also the writing can't stand or be judged by itself. It can only be valued and appreciated through human mind. Art (and many books are art) has the same problematic relationship with AI. John Berger's "Ways of seeing" talks about this (but so does a lot of philosophy) https://www.youtube.com/watch?v=0pDE4VX_9Kk

Re: The computers are getting better at writing

#85

Post by anthk »

watwut wrote: post #64
dumpsterdiver wrote: post #42

Earlier quoted context omitted.

Well, it's not just women, I have read and enjoyed such writing. But you are correct, writing that lacks substance intended for people who expect nothing more. Sex is easy to write about. It doesn't take much to turn people on.

I used to read a lot of junk paperbacks - mostly horrors, but also stories about truck drivers, detectives and western stories. They were schematic to the extreme. Looking back, it was the same story over and over and over. I liked them back then. I used to read also a lot of repetitive sci-fi: those small start trek books. I never read Harlequin, but it being repetitive is not something special. Back when people rea…

>I used to read a lot of junk paperbacks - mostly horrors, but also stories about truck drivers, detectives and western stories. They were schematic to the extreme.

He, ditto in Spain. In most cases the (same) authors would rehash a detective story as some sci-fi based short novel by just changing some devices and lore and call it done.

Well, in the end cyberpunk it's just futuristic noir.

Re: The computers are getting better at writing

#87

Post by notahacker »

causality0 wrote: post #4

GPT-3-generated writing is a combination of comically absurd and incredibly deep. The weak infants of the woman next door who had to be picked up with the tongs and thrown into the dustbin is quite silly but then there's He wanted to crawl away from it, but there was no place to go. which is a surprisingly deep statement about the revulsion he feels towards himself. That a computer can generate that says very uncomfo…

Is it "surprisingly deep" though? Looks like two standard cliches of human writing often used in close proximity and the wider context suggests literal motion (which is also present in the silly line about the weak infants!)

I did like "fuzzy belly" tbf, and Kafka himself might have appreciated the purely serendipitous irony of a line about "the bugs his father used to bring him as a little boy"

Re: The computers are getting better at writing

#88

Post by dave84 »

nxpnsv wrote: post #76
ayushchat wrote: post #72

I've actually used tools like copy.ai and converstion.ai for a couple of copywriting projects and found them to be very useful. Of course the output isn't perfect, human intervention is required, but it is way better than doing the writing myself from scratch. I think going forward we are going to see a lot of human + AI writing projects instead of purely human or purely AI.

* you mean https://www.conversation.ai/ ?

I'd imagine it's https://www.conversion.ai/

Re: The computers are getting better at writing

#90

Post by WesolyKubeczek »

I have a suspicion that, since GPT-3 gorged its enormously swollen fuzzy digital belly on a lot of real prose, it might, when prompted, occasionally throw in a piece of the prose in question, or a whole original passage, because to its fitting function, it would be a suitable thing to do.

But then again, aren’t we all doing the same, regurgitating bits and pieces that we heard or read before and liked?

Don’t we all sometimes, when busy with our own thoughts and worries, occasionally spout out oddly appropriate responses to what our spouses are speaking, without actually listening to and processing the conversation, maybe because we heard it all so many times we can readily predict what they are going to say? And then we agree that we need to go take the alligator from the repair shop.

However, if anything, GPT-3 is mostly showing how little meaning there is, but in how many words.

Never mind me, I’ll pour myself another one.

Post reply on HN