Often i think of a novel idea or solution to a problem, but use AI to communicate or adjust what I already wrote out so it’s more comprehensible. Sometimes when I write, it’s hard to understand.

▲

davebranton 9 hours ago | parent | next [-]

The more you write, the less this will be true. The more you write, the better you will become at it. Using an LLM to write is like sending a robot to the gym for you.

The more you use an LLM to write for you, the worse you will become at writing yourself. There is simply no other possible outcome. It's even true of spellcheck - the more you use a spellcheck the worse you become at spelling. I know this for a fact because I can no longer spell for shit. However, spelling is to writing as arithmetic is to mathematics. I also can't add up, but I have a degree in pure mathematics.

LLMs are a cancer on human thought and expression.

▲

briantakita 7 hours ago | parent [-]

> LLMs are a cancer on human thought and expression.

LLMs help to express what many people dont have the energy or ability to express. It also has a broader scoped view of protocol...It does not have emotions, which often leads to less than optimal discourse.

In many ways, it help those who are challenged in discourse to better express themselves...rather than keeping silent or being misunderstood.

	▲	rozal 4 hours ago \| parent [-]
		[dead]

▲

jamiek88 8 hours ago | parent | prev | next [-]

How do you expect to get better at it then if you avoid the hard work and emotional weight of fixing it?

▲

yellowapple 7 hours ago | parent [-]

So if you want to reply to a comment you read today, and you don't feel like your writing skill is up to snuff, you should be content with expecting to wait the requisite weeks or months or years of practice before even considering replying to it?

This seems especially relevant for non-English-fluent commenters, who are increasingly using LLMs to be able to communicate more effectively on an English-only site like Hacker News than they'd otherwise be able to do.

▲

rukuu001 6 hours ago | parent [-]

I've noticed a considerable drop-off in HN commenters who are unable to deal with the substance of a comment if it contains errors in spelling or grammar, so I don't think this is the issue it used to be.

It's still daunting posting in a second language, and LLMs are an attractive solution to that (depending on your definition of 'solution').

	▲	yellowapple 3 hours ago \| parent [-]
		Is that an actual drop-off in commenters, or in comments? The latter is readily explainable by “commenters who would previously call out the errors now choose to not engage with those comments/posts at all”. In any case, I don't think it's a bad thing to want to communicate as clearly as possible, and if an LLM helps you do that, I ain't one to judge. Sure, ideally I'd want to read folks' thoughts without the LLM-induced layer of vaseline smoothing them over, but even that's better than not reading them at all :)

▲

sharken 10 hours ago | parent | prev [-]

In that sense AI is a tool much like a dictionary, it enhances and I'd say improve the end result.

	▲	verdverm 8 hours ago \| parent [-]
		The difference is that I will retain what I drew out from the dictionary the next time. If people use Ai this way for writing, great! What many of the "enhanced-by-ai" arguments sound like is that this will be an indefinite outsourcing. Use them to get better, like how reading good writing directly (not summarized) will also make you a much better writer. Learn from the before and after so next time there isn't a need to reach for Ai.