Is it possible to blog in the AI era?
I write short stories every now and then and I throw them online. I also have a tech blog, where I moan about the decisions software I use make and with my “infinite wisdom”, I tell them what they should be doing instead.
I used to host both on Medium, but Medium got greedy. Then it was WordPress, but now even they’re trying to be greedy bastards and use my shit for training AI.
Some would argue that WordPress paid hosting will exempt me from the AI training, but for less than 100 visitors a year, it’s not really worth the expense.
So what is the solution? I ask the greater minds of this community for suggestions.
Use the gemini protocol. No need to worry about bots or AI
That’s actually really cool. Not feasible as I want visitors, but cool AF.
There are gemini to http gateways so the content is probably already crawled anyway.
I think you should clarify the problem first.
Privacy? You lose your privacy the moment you publish your blog anyway.
Is it visibility? You never expected Google to show your blog in most cases.
AI training? You could self-host and hope companies respect your robot.txt. But what’s the actual problem if you released your blog to the public in the first place? Anybody could’ve copy & pasted your blog also before this AI era.
Privacy? You lose your privacy the moment you publish your blog anyway.
Oh, right, I’m gonna just reinstall facebook on the phone because I’ve lost everything… Oh and we have lost all of privacy by commenting on the internet and stepping out of the house! All resistance is futile! We need to close this community before people waste more of their time!
This is not at all how it works. How would you lose privacy if you only publish what you want to publish? It’s entirely your decision what to include in your blog post.
Privacy?
Privacy is of course my major concern, hence posting to this community. But not tinfoil hat level.
visibility?
I’m happy to have my stuff indexed by Google, in fact, I want it to be.
AI training?
I’ll take that for 500!
Anybody could’ve copy & pasted your blog also before this AI era.
Plagiarism has been an issue since before Confucius was copied by Baffledus. But the cream still rose to the top. However in this AI era, everything is buried as its all just considered a part of the source data.
robot.txt.
Stories keep popping up about AI ignoring robots.
But… have we ever had privacy with blog articles? I mean the public ones.
I guess it comes down to what your definition of privacy is. I’m setting the bar low, I just don’t want to be used to train a large language model
Host your own stuff. With this little load you can do it on your own hardware with very little resources.
Yeah, I was thinking about throwing something on my Raspberry Pi, but didn’t know if I’d open the door to more issues.
It can be pretty secure if you host it behind a cloudflare tunnel. Then you don’t have to open any ports to the wild west
Thank you. I’ve heard so much about CloudFlare tunnels, but don’t know how they work. Do I just point it at an IP and port or is it much more complicated than that?
Basically you have to run a mini server (I use a docker container) called a cloudflare endpoint. From there you just enter the IPs and keys that your cloudflare account tells you to in the tunnel creation menu, and it all pretty much connects from there.
Then, on the cloudflare side, you make different subdomains point to local ports. So, for example, for connecting to qbittorrent web client, in the cloudflare menus I can make qbit.domain.example point to localhost:8080. In this case, it means “localhost” relative to the cloudflare access point you’ve made (which in my case can use localhost because its hosted on the same machine as my other docker containers, but if they are on different machines you can use local IP addresses).
I use their free plan, which is all you need if you’re just serving web content to a small number of users. You might need a domain to do this, but I don’t recall.
My layman’s understanding is you basically make cloudflare be the router, so their server/ports are what is exposed to the open internet rather than your local router.
Thank you. I run like a million Docker containers and haven’t ever gotten around to looking into this and you’ve just enlightened me perfectly. I appreciate it.
If you’re already running a million docker containers then just get a vps somewhere to host your blog. Cheapest reliable one I found last I looked was vultr. I think mine is $15 a year.
I’m glad! Halfway through writing that I got worried it was a little opaque. Best of luck setting it up. If I can do it, anyone can!
You could also spin up a $5 a month VPS somewhere like Linode.
Maybe Write Freely? You get 3 blogs for 6 dollars a month or you can host it yourself since it’s open source.
I’ve been wanting to live write freely since forever, but it just feels sparse.
I’m not sure what features you’re looking for, but Quarto has a lot of really nice features that make it really easy to self host a blog.
You monster! Why are you introducing me to new things to check out? 😭 thank you very much 😂
There are two good options: Host your own blog yourself, or join a blogging platform that isn’t corporate. I personally use BearBlog but I’ve heard good things about Write.as as well. These two have free blogging options and don’t sell your data. If you want to host it yourself (which is safer), check out Hugo.
Ultimately, bots scrape the entire internet and there’s no guarantee they will honor robots.txt of a particular website (which tells bots what they are and aren’t allowed to do). If it’s on the internet, people can scrape your content and there isn’t much you can do about it. That shouldn’t stop you from writing or blogging, just don’t post very personal data.
Also, feel free to join us on !blogging@programming.dev!
When I was looking into Ghost over the weekend, Hugo kept popping up.
Also subscribed.
WordPress is libre software.
Link to tech blog please
Sent as a private message
i dont think there is a good soulution for you.
if its out there, somebody will find it.you could host your own wordpress instance independent of WordPress.com.
and you could add a robots.txt to tell google to not scan your content, or even completly block the user agents of known search engines.
but blocking search engines is rather counterproductive if you want readers to find your blog.
and even then more nefarious crawlers might ignore the robots.txt and spoof their user agent to find you.
What about https://writefreely.org/
There are reputable hosters like https://text.tchncs.de
The one thing I dislike about write freely is that it doesn’t support comments. What are Blogs without comments?
I didn’t know about that. That’s bad. Is it not even planned?
They (write.as) decided against it, keeping it a strict uhm… writing platform.
It would be so easy to display ActivityPub Comments under the Article, but oh well.
How did I not know that tchncs.de was more than just a Lemmy instance?
WriteFreely is pretty nice, it uses the ActivityPub protocol and is thus a part of the Fediverse - just like Lemmy and Mastodon.
I’m starting to like the idea of a writefreely more and more.
Maybe https://bearblog.dev, very simple but I think enough for writing stories, it’s free, OpenSource and private.
- A privacy-first, no-nonsense, super-fast blogging platform
- No trackers, no javascript, no stylesheets. Just your words.
- This is a blogging platform where words matter most.
- Shun the bloat of the current web, embrace the bear necessities.
- Looks great on any device
- Tiny (~2.7kb), optimized, and awesome pages
- No trackers, ads, or scripts
- Seconds to sign up
- Connect your custom domain
- Free themes
- RSS & Atom feeds
- Built to last forever
So I’ve been looking at this and my only issues are that it’s not connected to the Fediverse and that it feels too sparse. I would like embedded image support. But thank you for the suggestion.
You can embed images, using Markdown
![](image URL)
, same as here in Lemmy, also several themes. But yes, it isn’t part of the fediverse.
Nothing to contribute that has already been said, but very interested in your blog as well!
Why not post your blogs to a fediverse platform? Do they need to be on a separate hosted system? You’ll probably get more people reading and engaging with your posts if you are just posting to a Mastodon instance rather than hosting on a separate web platform and hoping that people stumble across it.
Funny you say that. That’s why I was kinda hoping for FireFish to be the new Tumblr, but that sadly didn’t pan out. But one of my requirements for self hosting is Fediverse integration.
Care to share your website url? I get interested on it.
If it’s about having your blog serve as AI training, it doesn’t really matter where you host it, it’s going to get scrapped and included in the data.
github pages site?
Github is microsoft.
People forget they use Microsoft Windows to write Microsoft .NET on Microsoft VsCode to then push to Microsoft Github and host it through Microsoft Azure