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

If writing is thinking then what happens if AI is doing the writing and reading?

hardcoresoftware.learningbyshipping.com

Post reply on HN
Print view View original

81–90 of 121 posts

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#81

Post by sothatsit »

mlinhares wrote: post #47
karaterobot wrote: post #6

The title made me think this was going to be about the mental consequences of outsourcing writing to AI. In fact, the article is completely about people not reading documents. Corporate documents to be exact. His examples are from the 00s, so the problem has absolutely nothing to do with AI. Heck, I, too, have noticed that nobody reads anything: what does that have to do with AI? At least with AI, people could read a…

Completely wild counterpoint, now that our docs are available to the AI bot, people are interacting with them more because when they ask the bot can reply, explain and then say "docs are available here" do the point i'm actually investing even more of my time in writing them.

I agree, it feels like the value of text documents has gone up immensely with AI making it much easier to make use of them. Accurate reference documentation that AI can point to is really valuable.

Although, the next step on this ladder is going to be that people don't even double-check the facts in the original document, and just take what the LLM said as truth, which is perhaps scarier to me than people not reading the original documents in the first place...

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#82

Post by Herring »

wvenable wrote: post #70
Herring wrote: post #26

Idk, I'm more optimistic than the author. I'm currently using a LLM to rewrite a fitness book. It takes ~20 pages of rambling text by a professional coach // amateur writer and turns it into a crisp clear 4 pages of latex with informative diagrams, flow charts, color-coding, tables, etc. I sent it out to friends and they all love the new style. Even the ones who hate the gym. My experience is LLMs can write very very…

> 4 pages of latex with informative diagrams, flow charts, color-coding, tables, etc. What tool are you using for this? I too have used LLM to do writing that, frankly, I wouldn't have done without it. Often I don't even take what it says but it helps to get the ideas out in written form and then I can edit it to sound more like how I want to sound.

Gemini pro + overleaf (learned latex from a long stint in grad school). Cheers mate.

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#83

Post by saulpw »

caseyohara wrote: post #65
gchamonlive wrote: post #59

I think it's the type of thing that the mere awareness of it counts a lot to counterweight the problem. It's like when you are growing up and a certain type of behaviour that would work for socializing when you're 14 years old suddenly doesn't work anymore when you are 21. You learn about it when someone you trust brings you attention to it and suddenly you have the opportunity to reflect and change your behaviour. T…

> mind-altering drugs like Adderall This is strange to me. You could give me 100 chances to guess which “mind-altering drug” you are thinking of and Adderall wouldn’t cross my mind. Amphetamine is a stimulant; it’s plainly not mind-altering in the way that psychedelics are. Adderall is mind-altering in the same way that caffeine is. Which is to say, it isn’t.

Yes, Adderall is mind-altering, even if it's not as profound as psychedelics. Caffeine is also mind-altering to a lesser degree.

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#84

Post by voidhorse »

kelseyfrog wrote: post #12

Writing is not thinking just as calculating is not "doing math". We've invented the equivalent of a "calculator for words" and we're going through the growing pains of discovering that putting words together is a separate activity from thinking. We've never needed to conceptualize them as separate activities until now so we don't have the conceptual distinction and language to even describe them that way.

I would say: sort of.

I think proving theorems is a better analogy than rote calculation—like the writing process, it is creative.

While mathematicians don't sit down and smash random symbols together to concoct proofs, many of them will stress the importance of good notation. Chewing on notation a bit can help reveal connections in the hunt for a proof.

When people claim that "writing is thinking" they mean something similar. The free-associative process of writing out our thoughts can help bring an initial clarity that is difficult to achieve without an external medium. Most of the time, we ought to polish those thoughts and connections just as mathematician ought to polish his proof sketch, but there is some extent to which the actual activity does in fact make up a significant portion of the thinking process. This is why offloading all of that activity to LLMs is dangerous. You can judge the string of propositions produced by an LLM, but you haven't thought them.

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#85

Post by intended »

wvenable wrote: post #70
Herring wrote: post #26

Idk, I'm more optimistic than the author. I'm currently using a LLM to rewrite a fitness book. It takes ~20 pages of rambling text by a professional coach // amateur writer and turns it into a crisp clear 4 pages of latex with informative diagrams, flow charts, color-coding, tables, etc. I sent it out to friends and they all love the new style. Even the ones who hate the gym. My experience is LLMs can write very very…

> 4 pages of latex with informative diagrams, flow charts, color-coding, tables, etc. What tool are you using for this? I too have used LLM to do writing that, frankly, I wouldn't have done without it. Often I don't even take what it says but it helps to get the ideas out in written form and then I can edit it to sound more like how I want to sound.

Production became easier -

But your consumer changed as well.

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#86

Post by southernplaces7 »

Herring wrote: post #50
southernplaces7 wrote: post #44

Earlier quoted context omitted.

Might I ask what LLM you use for it to do all of that, including the visuals, so neatly?

Gemini pro. The flow chart, color-coding, tables etc are just latex. The illustrations are picked off the web. I suspect I might have to hire a professional photographer eventually.

>The flow chart, color-coding, tables etc are just latex. The illustrations are picked off the web

Thanks for replying so quickly! Just to clarify, what do you mean by latex?

So you don't use AI-generated illustrations. Those are real.

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#87

Post by volemo »

bugbuddy wrote: post #74
didericis wrote: post #66

Earlier quoted context omitted.

4. Degradation: Humans with specialized knowledge lose their specialized knowledge due to over reliance on AI, and AI degrades over time due to the lack of new human data and AI contaminated data sets.

5. Society collapses in an Idiocratic fashion.

6. PROFIT?

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#88

Post by makeitdouble »

GMoromisato wrote: post #54

I think Sinofsky is asking a question: what does the future look like given that (a) writing is thinking but (b) nobody reads and (c) LLMs are being used to write and read. It's that (already) old joke: we give the LLM 5 bullet points to write a memo and the recipient uses an LLM to turn it back to 5 bullet points. Some plausible (to me) possibilities: 1. Bifurcation: Maybe a subset of knowledge workers continue to w…

> It's that (already) old joke: we give the LLM 5 bullet points to write a memo and the recipient uses an LLM to turn it back to 5 bullet points.

This is already how we moved from stupidly long and formal emails to Slack messages. And from messages to reactions.

I understand not every field went there, but I think it's just a matter of time we collectively cut the traditional boilerplate, which would negate most of what the LLMs are bringing to the table right now.

> 2. Augmentation

I see it as the equivalent of Intellisense but expanded to everything. As a concept, it doesn't sound so bad ?

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#89

Post by snoman »

wlesieutre wrote: post #20
gleenn wrote: post #13

Earlier quoted context omitted.

I am firmly a believer very few people read anything. They don't read long things as much as they don't even read short ones. One of the things I always thought was funny was having Product Managers see there were problems with UI, then I would get tickets to add text near the problems. It always crackme up because if users didn't even barely read the button they were clicking they why would they read a paragraph nea…

In school they were really big on the 5-paragraph persuasive essay format. I guess because it teaches you to think through an argument and present it to someone. In practice, I find that if I don't format something as a bulleted/numbered list, nobody is going to look at it.

> In practice, I find that if I don't format something as a bulleted/numbered list, nobody is going to look at it.

I’m one of those people, if I’m honest. If I’m reading for work, I want the minimum words necessary to get the point across. I’m looking for information, not a story.

If I’m reading for enjoyment though, it’s another thing entirely.

Re: If writing is thinking then what happens if AI is doing the writing and reading?

#90

Post by bitwize »

wlesieutre wrote: post #20
gleenn wrote: post #13

Earlier quoted context omitted.

I am firmly a believer very few people read anything. They don't read long things as much as they don't even read short ones. One of the things I always thought was funny was having Product Managers see there were problems with UI, then I would get tickets to add text near the problems. It always crackme up because if users didn't even barely read the button they were clicking they why would they read a paragraph nea…

In school they were really big on the 5-paragraph persuasive essay format. I guess because it teaches you to think through an argument and present it to someone. In practice, I find that if I don't format something as a bulleted/numbered list, nobody is going to look at it.

Now we're in the era of:

Alice: Hey ChatGPT, please take this bullet list of points and turn it into a polite, but assertive and persuasive, email to Bob.

Bob: Hey, ChatGPT, please take Alice's email and turn it into a succinct list of bullet points.

Post reply on HN