I blog solely for my own memory. If people find it interesting, ok, but it's because I forget stuff and then will go somewhere thinking it's interesting only to recall I did this already. If no-one ever reads it, it's not really a worry.
It really fascinates me how the "other" clientele of HN operates. The world of networking, marketing yourself, creating a resume of sorts through your blogging. I'm just some dude with a decent job occasionally writing blog posts because it amuses me. I'd be pleasantly surprised if a dozen people read them before I die. AI doesn't really factor into the calculus of whether or not I want to continue doing it at all.
I might be the kind of person that requires external validation, what I noticed is I simply stopped writing online, for the most part. It just feels like writing/tweeting/etc into a cacophony, makes me slightly uncomfortable adding to the noise and discouraged that nobody (no human at least) will read anyway. So I just write what would be blog posts on Notes app and don’t post anywhere…
I have actually cut back on blogging (partly) because I don't want my hard work to be slurped up and regurgitated by an AI. I write for other people. Not for AI
AI doesn’t care, the people using AI care. If you really write for other people, I’d recommend you reconsider blogging again.
Even if you write primarily for yourself (vanity, marketing, client acquisition, there is nothing wrong with that) and not for other people, I’d still recommend you publish your stuff. Not publishing will have always <= effect than publishing, even if AI slurps it up.
I am a person (allegedly) and I would benefit from AI regurgitating your content.
Much the same way I benefit from Google indexing the internet, and summarizing news articles.
Reading and writing are both intimate activities. The reader holds the writer's thoughts in mind, and the writer knows this and acts accordingly. Personally, I don't particularly enjoy reading material that was made by an LLM. The fact that so little effort was applied suggests that there is not much reason for this to exist beyond a chance to serve a few quick links. Since the llm is also running this as a business, I would also point out the social connection between reader and writer does come with some expectation of a reward. Whether it is to be paid in cash or respect is beside the point. People often expect some reward for their efforts and they are not wrong to want that. People are often uncomfortable to put it so bluntly because that would compromise the quality of the relationship, but upending this relationship is really a perversion of logic carried out for purely selfish reasons ("I want what you are making, and I will give nothing in return").
Google gives attribution and maybe provenance, while AI gives you smoke and mirrors. I guess we'll decide if copyright has any legs left to stand on in the modern world, or if it falls as collateral. It's so sad that commercial piracy has hit such an incredible tipping point that even I feel bad for creative people and their bleak economically dead future ahead.
People had been stealing ideas long before computers started doing it.
> People had been stealing ideas long before computers started doing it.
So it’s not ok to be bothered by it because it’s been happening a long time?
I can't imagine how culture could exist without ideas being taken without permission.
Straw man. I can’t imagine how culture could exist without some people being thrown in prison. That doesn’t mean that criticizing a law or law enforcement practice is equivalent to arguing for abolishing prisons.
We can agree that prison and appropriation are both necessary for culture, regardless of their negative effects.
Your problem is the scale at which knowledge and ideas are being appropriated, but my point is that it was already happening a lot but it was far more implicit; now it's just explicit that it's happening because we see the process laid bare.
No, that is not “my problem” with it. It is one facet of it, there are lots of sucky ways to take ideas that have nothing to do with scale, and there are lots of ways that fundamentally mechanically harvesting “ideas” as data is different than just learning quickly, but I’m not interested in re-arguing any of this for the 500th time on HN.
What about the idea that writing for the AI is your best opportunity to help contribute to it's direction?
Makes me wonder, how about changing your blog to a mailing list?
That's no guarantee it won't get slurped up by an AI at some point. Anything that goes into, say, GMail is ripe for plucking. And there's always a good chance your newsletters will get publicly archived on some web page somewhere, whether intentional or not.
I had the same thought. Still, perfect or not, I bet it'll be an attractive option for some.
I guess our gmail content has been fed into an AI of sorts since many years ago. I would surely hope, however, that Google would not use it for any sort of non-private LLM training data.
Yeah maybe that's the move. I don't really have a following though, lol
I don’t think there has to be some practical economic business justification to a) feel bummed out by your creative output getting munged up into something that, for all its better uses, will feed the great fire hoses spraying trillions of gallons of bullshit all over our information landscape, or b) reduce your creative output because of it. It’s weird how entitled people feel to other people’s creative work and get mad when people don’t freely create for and share with them, while simultaneously minimizing the value that work and its authors bring to our society. Despite what many say, the way people receive and interact with your work mentally/emotionally is really important, and all your work being sucked up into these models— often to create commercial products that are openly antagonistic to the people that created the work that made it possible— changes that. It’s sad that AI has devalued creative processes even to the creators themselves.
Is your job super-safe? If so, that's awesome :-) The whole marketing thing only becomes important if you have to get a new one, and then it can become important very quickly.
Writing is thinking. So is drawing.
To think clearly, come up with new ideas, make and truly understand things, we need to put marks on the blank page ourselves, and not just repeat what teachers or textbooks tell us like the majority of students Richard Feynman had during his time in Brazil — https://v.cx/2010/04/feynman-brazil-education
LLMs/AIs are useful to help us get farther, faster, like witty, skilled, intelligent friends who sometimes take too many magic mushrooms during conversations.
Forgetting about our own agency and individuality is bad for us, and dangerous for society.
“I am no bird; and no net ensnares me; I am a free human being, with an independent will.” —Charlotte Brontë, Jane Eyre
“Monsters exist, but they are too few in number to be truly dangerous. More dangerous are the common men, the functionaries ready to believe and to act without asking questions.” ―Primo Levi, If This is A Man
To create and be free like an animal outside a cage, ask, write, and draw your own questions. Look, and find out for yourself, rather than blindly believing what others tell you.
Two useful books:
Drawing on the Right Side of the Brain by Betty Edwards https://archive.org/details/DRAWINGONTHERIGHTSIDEOFTHEBRAINH...
The Hand - How Its Use Shapes the Brain, Language, and Human Culture by Frank R. Wilson https://www.penguinrandomhouse.com/books/191866/the-hand-by-...
For anyone put off by the dubiousness of the left/right brain thing, rest assured that Drawing On the Right Side of The Brain is no less useful as an introduction to abstract visual creativity because of it.
This comes back to my initial thought on the effect of AI on the "producer" rather than the "consumer".
I do a lot of things badly because it helps me develop skills and makes me happy. I wouldn't outsource this to an AI even if it did all these things better and the world benefited more from it. This is for me.
feynman-brazil-education was an amazing read. Thank you.
I am from India, and I have a similar experience with my education — one that forces you to memorize, never experiment, and never connect the dots. It felt like reading about my own past and realizing just how bad it was.
Drawing on the Right Side of the Brain is discredited pseudoscience bullshit. The exercises will make you a better drawer but not for the very stupid reasons the author claims, but because they are drilled exercises.
> You're building up a portfolio of writing about topics that interest you.
This reason resonates with me immensely.
You're not just writing about what you've figured out, sometimes you're actually deepening your understanding as you write! Writing is the thinking process: https://news.ycombinator.com/item?id=32628196
I have been writing every day at my blog for three years now, and it's been very rewarding for me to figure out what I actually care about and seeing patterns.
I like thinking about it like a bunch of skateboarders lugging the video camera around to capture the moment. (They did this before social media!)
P.S., You may also enjoy the similar sentiment in this post: https://news.ycombinator.com/item?id=42992159
> I like thinking about it like a bunch of skateboarders lugging the video camera around to capture the moment. (They did this before social media!)
We sure did! I’ll never forget my first vx1000 with a death lens.
I like writing too. Funnily enough, the age of LLMs makes this even better. I wrote a little MCP server (this is trivial with Claude) that interfaces with my blog so it can full-text search for articles and look up articles and look at recent articles and stuff like that and it is pretty good at finding references in what I've written to thoughts I've had. It's a bit trigger-happy when looking up my blog posts (I have to put more in the assistant prompt in the Claude app to get it to stop defaulting there).
The other thing that's nice is that LLMs make the process of writing better. When I cite stuff I can just screenshot the website and ask ChatGPT to write the citation and then check it. Things like that are more painful to write than to check and LLMs shine there.
For me, the #1 reason I don't blog more, especially about tech topics, is that they take too long. Maybe you can bang out a useful blog post in 20 minutes. For me it's more like 4 to 8 hours.
I have to make samples. Since I do mostly web tech I want the samples to actually work, no "here's some code, trust me". I also need diagrams. And, I have to proofread since I'm terrible at getting it right in one or even 5 checks. I write once, add samples, write some more, add images, write some more. Every time I write I add errors, so it always takes multiple passes.
Proofreading is one place that AI can actually be a friend rather than a foe. If you give Claude your draft and tell it explicitly to call out misspellings and grammatical errors only, it does a really good job.
Much like (say) MS Word does? In real time.
My drafts usually live for days, weeks, or even months before I publish them. (And sometimes I throw them away after having worked on them for many hours.)
My advice is to "chill": focus on the process instead of the result and let the posta take the time they take.
The age of AI is the very reason it's even more important that we blog. AI in it current form ingests and repeats what we train it to. I don't think new ideas will "organically" come out of current generation models.
If you now make blog entries on new ideas, it will be consumed and regurgitated as new ideas with no attribution or benefit to the original author, giving the illusion of new ideas.
Feels dirty to have your creativity stolen easily by unforgiving machines.
I'm thinking the only reason to write may be to either have your ideas consumed by AI intentionally (so they are influence AI users) or to add negative value to the pool to make AI more useless when it regurgitates incorrect data.
I started blogging to record the history of projects I've worked on (e.g. Microsoft Band https://meanderingthoughts.hashnode.dev/history-of-microsoft... Launching HBO Max https://meanderingthoughts.hashnode.dev/only-300-software-en...) but, perhaps not surprisingly, my only "successful" article has been the one controversial article I wrote cautioning against using SSR for everything.
That said I believe documenting history is important so I'll keep sporadically writing down notable events I've been involved in!
I also blog a fair bit about AI, and there is no hope getting views there without playing the game.
I think that as more people offload understanding to LLMs, being able to deeply understand a topic will make you stand out more and more. Doing things and explaining them are two of the best ways to get that deep understanding.
When I write about a technical topic, I open a new markdown doc and just go. You quickly run up against the limits of your own understanding, which is a valuable exercise.
You're conducting multiple parts of a dialogue when writing, but discussing what you've learnt can also be quite a good way of solidifying learning and encouraging further thought.
Exactly, reading and consuming information is one thing but teaching it to someone else is something else entirely. If you're not writing with the goal of self-marketing and content-farming, it's still worth it.
Blogging pushes me to explore things I probably wouldn’t otherwise. That’s been the main reason I’ve stuck with it pretty consistently[1] for the past five years.
Getting attention was never the goal, so the rise of LLMs has mostly been background noise to me. There have been plenty of times when I’ve searched for something on Google, only to land on my own page.
Over the years, though, things picked up. Now, I’m seeing around 30k monthly readers—way more than I ever expected. More than once, I’ve written about something I barely understood, only for the post to hit the front page. Then people corrected me, and I learned a ton in the process. That’s something I wouldn’t trade for anything.
[1]: https://rednafi.com/
Me too! These last two posts blogging about blogging are unusual for me. I'm working through a book (Sebastian Raschka's "Build an LLM from scratch") and posting about that at the moment. It's likely not a coincidence that I'm procrastination-posting before going through the trickiest bit...
I love reading meta-writings at times, as long as there’s a real human behind the keyboard. This was a fun, quick read.
Really enjoyed reading about your blog stack, motivating me to get my own up and running.
Love the blog!
Thank you. One reason I wrote it was to demonstrate how easy it is to spin up a blog where everything is automated and you never have to worry about the infra.
I’ve tried building a static react blog and hosting it on vercel and while it was easy to set up there was just too much styling and configuration to sort through that by the time it came to writing i was pretty much unmotivated. Markdown seems to be the key here, going to try spinning one up tomorrow. Cheers.
I think there is a more important reason to blog besides the 3 reasons listed: to force yourself to slow down, organize your thoughts, fill in the holes, and articulate your points.
"Writing is understanding."
"Writing is nature’s way of letting you know how sloppy your thinking is" - Guindon
Though admittedly I first heard it from Leslie Lamport and that's who I associate it with.
That's what I was trying to cover with the "make your newly-acquired knowledge concrete" bit, and was my focus in the previous post. This time around I wanted to look into the aspects that might be impacted by AI (and why I didn't think they would be).
Socrates would like a word with you.
I find that thing where people say "I'm not going to publish anything creative ever again, it'll just be used to train AI" so depressing.
It feels like such a dismal excuse to avoid andding any value to the world.
It's not an excuse, it is a reality. Why spend your personal time and effort for someone else with a deeper pocket to automatically extract value from your work.
There is certainly a line where if you're popular enough and have significant google juice you'll still get organic traffic, many small bloggers can go their entire posting history without getting more than a smattering of hits and now chatgpt is taking away that.
There is also the gwern point of "You should write so that your voice is present in the future LLM": https://www.lesswrong.com/posts/PQaZiATafCh7n5Luf/gwern-s-sh...
The jokey last paragraph was probably based on a half-memory of having read that, now you mention it...
I checked out his About page just because he mentioned that only 1% do it. For my writing, the observation that almost everyone moves on is certainly true. But those few who reached out sometimes become new friends. I've never met any of them in person though. It's like a group of friends for whom the connection is that they all read something that I wrote. Also, funny to now see that when doing research I've gotten AI's citing my own past posts as sources. Maybe this means I need to reread my own writing.
The main reason I blog is because I work (or am starting out working in) academia and therefore I have to write papers. To write academic papers you have to write terribly, using passive voice whenever possible.
I blog so that I know I am still capable of writing coherently, rather than in horrible academic language.
You can avoid the passive voice and write coherently or even conversationally if you have something interesting enough to say.
Anyone here seen an LLM actually produce a really novel thought that hasn't already been written about ad nauseam?
The well of new ideas, or re-formulation of existing ideas with perspective and prose that LLMs can't match is plenty deep to be worth hoisting the bucket still.
I don't know that I've seen a human blogger on HN produce a really novel thought that hasn't already been written about ad nauseam.
This blog post is itself a good example. It may have the author's voice, but it's just another rehashing of something that a million people have said before already. HN loves upvoting blog posts about why you should blog. https://hn.algolia.com/?q=why+blog
I can't find a less snarky way to say this, but why are you here then? I figure there's enough signal in the noise to make it worth it.
Do you mean on HN or in this particular thread?
I read Stratechery. He’s had some novel thoughts. He just repeats them ad nauseam.
He’s made quite a living from blogging.
https://blockbuster.thoughtleader.school/p/how-ben-thompson-...
I’ve seen novel thoughts, but typically they are coming from trolls who are taking impossibly contrarian opinions and presenting horrifying morally bankrupt perspectives. Maybe these ideas are common elsewhere though.
All the “good” ideas that can be said have probably been said. Maybe that’s why some people just enjoy trolling: for the novelty.
> Maybe these ideas are common elsewhere though.
I don't know which ideas you're talking about, but I bet they are. Humanity has thousands of years of recorded navelgazing (the blog post) and reactions to navelgazing (me) and reactions to reactions to navelgazing (you) and so on.
> Maybe that’s why some people just enjoy trolling: for the novelty.
Novelty is a morally bankrupt reason to enjoy trolling, so maybe you're onto something.
> So, I don't think you can make a name for yourself by blogging alone
Or maybe you could if you post bangers like https://ciechanow.ski/
Nice blog either way!
Thanks! And that's a good counter-example... which made me realise, there's also https://waitbutwhy.com/
My 2c: AI can give you the answers if you have the right questions. Having the right questions is a journey of human development through self reading, for which blogs are essential.
I wonder if there is a practical test of this question? Are there any moderately successful blogs written purely by an LLM with a human just doing the prompts.
As far as I know there aren't any but I look forward to being corrected
I've suggested this elsewhere too, but have you considered a mailing list instead?
It ticks the boxes about notoriety—which appears to be the main concern of the OP author—with the added benefit of being more difficult for AI systems to crawl for training data.
Perhaps we'll see more bloggers going the way of a bespoke mailing list.
An entire conversation about "where's mine" and "what's in it for me". All the self centered people, right here.
There was a good reminder in the Hanselminutes interview with Shaundai Person about blogging for yourself. It was a while since I listened so I'm fuzzy on the details - all I remember is that I enjoyed the interview and there was some relevance to this topic.
Also, blogs are no doubt a major source of AI training, so maybe more worth than before.
I believe its even more important now than ever before.
I've been considering starting up a(n extremely) sporadic blog for funsies. What platforms do people use these days?
While quite a few folks agree blogging is still worth it in the age of AI, please do your bit to support privacy-friendly services (ie. no tracking). Blogs don't /need/ modern web bloat, nor need to resort to adverts / paywalls.
I happen to build a privacy-friendly option https://LMNO.lol (my blog at https://LMNO.lol/alvaro or https://xenodium.com). I'm hoping it's fairly priced at $1.50/month. You pay for hosting and we provide that service (that's the extent of our transation).
There are others services to pick from like Bear, Ghost, Nekoweb, omg.lol...
I like https://mataroa.blog/
If we don't blog, what will the next big AI models be trained on?!
Good point. "Blog as much as you can, as otherwise the training set will be Twitter!"
Ah, the weekly post about if writing blog posts is worth it with the comments reaching the same conclusion.
...you can do it!
What the hell are people blogging about that is so generic that an AI could write it?
The important part of any journalism is to introduce new facts and findings that aren't anywhere else to be found. An AI cannot do that.
What upsets me about AI is that the proof-of-value in basically articulate prose has gone to zero. I can still write better than AI, but I’m not sure the remaining margin matters, socially speaking.
That said, AI is less than 2 percent of what has enshittified the internet. It’s a factor, but corporatization and algorithmic optimization have already done plenty of damage even without LLMs in the mix.
writing for thinking , share ideas with others
Related aforementioned thread:
Write the post you wish you'd found https://news.ycombinator.com/item?id=43154666
Writing things down reveals the lapses in one's own knowledge, and those gaps are opportunities to learn something new.
It saddens me a lot that people are now relying on AI to fill the gaps in their knowledge.
Humans got where they are thanks to their innate curiosity. If you take that away from us, we become no different from animals.
Not just gaps. Blind spots too. Humans are driven by curiosity and limited by biases. You're only painting half of the picture.
> Humans got where they are thanks to their innate curiosity. If you take that away from us, we become no different from animals.
Your message is a bit late. We've had people zoned out in front of their TVs for decades.
I deeply distrust anyone blames AI for "ruining" their creative hobby, or "creative" career. Mostly because the one absolute, will never ever change, always our trump card, is our individual perspective and creativity. AI is a sorting and statistics machine, it can only reproduce (my personal theory about AI hallucinations is pretty simple. AI has been a well known concept for longer than the Internet has been around, pop-scifi and post apocalyptic books, movies, and shows, and causing a societal immune response... And satire, sarcasm, bad photography, bad art, websites with gibberish or incorrect facts just to game SEO algos (many many many that still have generations old SEO text blocks), all being hoovered up by an AI and the techbros supervising it all are MBA dropouts, and/or marketing bros. All their AI datasets are so poisoned by the human condition, and as universal, AI hallucinations/errors are, it just proves there isn't a single model, or a single person, in the AI space who can identify irony. We may be the most destructive thing this earth has ever encountered, but like.. Causing mayhem and chaos so concentrated and complicated it's impossible (so far at least) to write code that can decypher it.
So yeah, anyways, I left tech almost a decade ago, I'm a photographer now, and if youve put thought into something. Whether it's writing, how a photo is edited/processed, a joke, its a gift you're giving to yourself, and hopefully some of whatever you felt like creating, a few pieces at the least haha, will be gifts to the world :)
am I so unusual in just skimming past and ignoring the AI search results? I didn't think a majority of people outside the silicon valley tech bubble were all that keen on getting their questions answered by a chatbot
yes. The research that goes into it helps be finally learn the concept better. Having a few subscribers also want to spend time to write something and put it out.
"It's still worth selling steaks in the age of the hamburger."
Yea. This will never not be true.
I just can't wait for the age of the "silicon valley visionary" to come to an end.