I can definitively say llms.txt is not used by any AI players. I run a blogging platform with around 80k blogs and /llms.txt is not requested by anything (other than humans checking to see if there's an llms.txt path).
All regular pages are aggressively scraped to the extent it's a problem I have to consistently manage, but not llms.txt.
It is a highly requested feature. I just seem to have painted myself into a corner dev wise, and it's a pretty difficult feature to build without significant re-architecture. It's on my to-do list though.
Creator of JustSketchMe here! I was very surprised to see this on HackerNews this morning. Very cool to see this doing the rounds 6 years into running this :)
Creator of Bear here. Suggesting that because one project fails, others will too is a bit of a fallacy. Fact is that whether you self-host or not, you're still using someone else's platform (unless you're a real self-hoster with a box in your closet, in which case, good on you and godspeed).
I think as long as platforms have an easy way for people to backup and migrate, that's fair.
Additionally, part of the appeal of Bear is that I've made it my personal mission to get the platform to outlive me. Take that as you will. I can't prove that Bear will live on in perpetuity, but I can try my best.
I run the Bear Blog platform. The discover feed on Bear has become a social hub and community space for many. There’s something quite charming about slower interconnectivity via RSS, email, and replying via posts.
I just did this recently! Here’s what worked for me and had me modelling fairly complex parts in a few hours:
- Watch a few tutorial videos on your tool of choice (I was using OnShape) to get an understanding of what is possible
- Set up a Grok project with the prompt being something along the lines of “You are mentoring me in using OnShape. Respond briefly and concisely.”
I have the two windows side by side and whenever I don’t know how to do something I ask Grok and it points me in the right direction.
Neat project! I’ve seen a few analytics tools starting to use SQLite over Clickhouse. It’s great for small to medium sized projects, and can potentially scale (but I haven’t tested it over 200k hits per month).
I rolled out a similar project a few months ago using Django + SQLite.
Interestingly enough, this is how I tricked myself into quitting smoking.
I withdrew $100, and told my good friend that I’d give it to him if I used anything with nicotine in it. It worked pretty well since it turns that first cigarette into a $100 cigarette. Not worth it at all.
I can attest to this. I run Bear Blog, and it has seen a dramatic rise in signups month-over-month. My suspicion is that other platforms are getting covered in AI generated content, and while Bear isn’t perfect, there is a lot more curation and a strict no advertising policy.
Definitely. Even if no one reads my blog, the act of articulating ideas in a format that others can understand helps me better understand those ideas and concepts myself.
I’m also bias, since I run the Bear blogging platform. There are thousands of very happy personal bloggers there with close to zero SEO goop (strict review process, and all that).
You obviously didn’t read the article and just wrote this comment on looking at the headline and picture. The circling the room twice is referring to an actual dragon fly. Also, it’s not ICE but fluidic motors driven by lithium nitrate crystals.
There’s also cover how it is very hard to control and how they do it manually with IR lasers. There’s even a video at the end of the prototype flying.
It has been synthesised. It's a mild stimulant but has a not-so-great toxicity profile. Probably best to leave it with the fibre you need to chew through.
Hey, author here. I decided on the 16 since its support lifetime started this year. If I'd gotten the SE (which I seriously considered since it has a more appealing form factor and touchID) I would only have a 4 year support lifetime as opposed to the full 6.
I recently tried to self-host Umami and Plausible, and failed so spectacularly I decided to build a Django/SQLite web-analytics platform myself.
These other open-source tools are great, but are incredibly difficult to get running yourself since there's no incentive for those companies to make it easy since they sell managed hosting.
Is Bearlytics better? No, but it's certainly simpler, and everything I personally need.
All regular pages are aggressively scraped to the extent it's a problem I have to consistently manage, but not llms.txt.