Ask HN: How do you feel about AI assisted blogging?
8 comments
I actually started a survey on how people react to AI assisted blogging ~a week ago. It's at https://forms.gle/oskwkY12epMpXHAZ6 if you want to share there as well
It's probably going to produce a worse product and I don't have time to read mediocre blog posts.
My first reaction is to feel distain for that person, especially if it's a blog or a long post. But recently I read a post on X that was clearly written with help of AI but the content was very solid.
I realized that the author really thought about the topic but lacked writing skills to express it. So I think if the idea is solid, I don't mind if they used AI. Within reason of course.
I realized that the author really thought about the topic but lacked writing skills to express it. So I think if the idea is solid, I don't mind if they used AI. Within reason of course.
It depends how the AI is used.
I have no problem with something like "Check this text for grammatical errors, do not change the writing style: < chunk of text here > ".
If one is using an AI to do the actual writing, I'm not interested in wasting my time reading it.
I have no problem with something like "Check this text for grammatical errors, do not change the writing style: < chunk of text here > ".
If one is using an AI to do the actual writing, I'm not interested in wasting my time reading it.
There's a lot of "it depends" to this. But loosely speaking, I'll say that at the end of the day, I care more about whether the content is interesting/engaging/original/whatever, than the specifics of how it was created. So if, say, I read a blog post and I can't even tell that it was crafted with AI assistance, do I really care that they used AI? Not so much. But if I'm reading something that was clearly generated almost exclusively by AI? I usually find it distasteful and tune-out.
But keep in mind there is a continuum over which the idea of "AI assistance" ranges. You have your "one shot" deals where somebody just says "Hey, Gemini, write me a blog post on the merits of warrantless surveillance in the AI age" or whatever. And then you have ones where somebody uses AI to do research, brainstorm, etc., but writes all (or nearly all) of the actual text themselves. And everything in between. So yeah... "it depends".
But keep in mind there is a continuum over which the idea of "AI assistance" ranges. You have your "one shot" deals where somebody just says "Hey, Gemini, write me a blog post on the merits of warrantless surveillance in the AI age" or whatever. And then you have ones where somebody uses AI to do research, brainstorm, etc., but writes all (or nearly all) of the actual text themselves. And everything in between. So yeah... "it depends".
I have personally pondered and accepted the fact that anything I have to say, written or otherwise, is a result of others' work no matter how much I want to believe I have a truly unique perspective. I read a lot of blog articles, and it's truly rare that I don't recognize either borrowed words or concepts even though I enjoy reading the content.
I admit that AI generated writing is a little strange unedited. Putting that output into a writer's voice with personal experience is always going to help to humanize it and make it more readable.
But the reality is that AI is just the next phase of borrowing and rehashing of older materials and concepts--just like reading an encyclopedia or Wikipedia article and then putting those rehashed words into an essay.
My suggestion is that as long as you enjoy what you're reading or learning, it's probably not worth getting upset about provenance. Anything truly new and innovative can't come from AI anyway--and innovation is very rare.
There is nothing new under the sun.
I admit that AI generated writing is a little strange unedited. Putting that output into a writer's voice with personal experience is always going to help to humanize it and make it more readable.
But the reality is that AI is just the next phase of borrowing and rehashing of older materials and concepts--just like reading an encyclopedia or Wikipedia article and then putting those rehashed words into an essay.
My suggestion is that as long as you enjoy what you're reading or learning, it's probably not worth getting upset about provenance. Anything truly new and innovative can't come from AI anyway--and innovation is very rare.
There is nothing new under the sun.
The same way I feel about AI assisted anything - that the drive to strip all of the character and flavor out of human expression in favor of an SEO optimized, generic, machine-extruded "fellow human" extract is extremely depressing and worrying.
The entire point of a blog is that it's the unique voice of a specific human being, a window into their lives and opinions. You don't need an LLM to spellcheck, you shouldn't need it to edit. You're not writing a book or a professional article. That's not what a blog is supposed to be. The only thing an LLM would accomplish in this case is masking your humanity.
The entire point of a blog is that it's the unique voice of a specific human being, a window into their lives and opinions. You don't need an LLM to spellcheck, you shouldn't need it to edit. You're not writing a book or a professional article. That's not what a blog is supposed to be. The only thing an LLM would accomplish in this case is masking your humanity.
My knee-jerk is to find it disappointing. But maybe I’m missing some nuance. And I don’t have the same negative reaction to AI coding.
But if you’re going to ask me to read a blog article, I don’t quite enjoy it when it’s not in your own words. I want to be in conversation with you, not you intermediated through an LLM.
While I don’t particularly care about a bit of LLM spellchecking or copy editing. I’d rather have human language warts and all than perfect AI-speak.
That’s the rant. But what am I missing?