Rendered at 07:04:30 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
semolino 19 hours ago [-]
Truly stunning false-video embed from Lyra in this post. When I realized I could use the input field and that the camera shake was simulated, my jaw dropped.
Most people would just put an mp4 on a CDN to avoid the hosting costs… glad to know the spirit of CSS criminality persists in this day and age.
nticompass 16 hours ago [-]
I didn't realize the video (and all the embeds) were just "normal" elements on the page. That's cool.
You can click the "I'm not a robot" checkbox as well, and a fake captcha pops up...
netsharc 1 days ago [-]
An actress (I think Jennifer Lawrence) told an anecdote on Graham Norton's talk show in the UK. I was watching live, around 10pm UK time. I was curious to learn more, so I started googling about the story. There were no results.
But as soon as the episode ended, there were results for the search! But they're all "news" sites referencing the episode that had just finished. "news" in quotes because they all felt bot-written (this was before LLM). For me it was a surreal experience about the speed of slop...
einhard 1 days ago [-]
This reminds me of a game called _We Become What We Behold_[0]. It's a short, 5 min browser-based game about news cycles and how they inform our identity and actions.
if you are critical of how technology works and have seen how deep the rabbithole does (prism was 2012, advanced AI already existed by then afterall, most household devices are smart enough to be tracking you today, and these are also deployed in public spaces..) then you can see where boundaries are being eroded and try to create your own space where this cannot reach into.
this is why i enjoy systems engineering. a way to create your own space, and say where boundaries cannot be eroded.
some software systems already attempt this. (notably Qubes OS), but more work is needed. always more work, so little time :)
netsharc 1 days ago [-]
I wish the normal rules of capitalization would be included in that boundaried space...
animal531 17 hours ago [-]
That reminds me of an old science fiction story in which humans have invented teleportation via fixed points that have been placed all over.
But in the story they also had constant news coverage, so whenever anything of interest (whether good or bad) would happen millions of people would just teleport over to go and see for themselves.
helmiancloud 1 days ago [-]
[flagged]
Legend2440 1 days ago [-]
[dead]
sparse-Matrix 19 hours ago [-]
This gal could be a millionairre, just by telling everyone how she manages not to use AI.
I don't think it is even possible to fully escape using AI now, especially if you work in any form of IT, especially especially if you are using the web.
In my estimation, the only way you can 'not use AI', is if you don't use a computer at all, and that includes the ubiquitous smart phone.
rebane2001 18 hours ago [-]
I only count it as using AI when it's an intentional choice to use it. I disable AI features everywhere possible, but if something gets forced down my throat against my will I don't count it as me using it.
For example, if I need to talk to the support of a company and I get throws into a chatbot chat, I will still use it, though mostly to try to get a human to talk to me. If a website or app decides to show me an AI summary, I ignore it.
For programming and writing, I use Sublime Text without a language server. For talks, I use my own slideshow software. For most other tools, provided they don't expose a security-critical surface, I have older versions saved that I use.
There are some things I use that some people might technically consider AI but I don't, such as search engines (outside of AI summaries), smartphone keyboards, machine translation, and NPCs in video games. I think an easy way to differentiate what counts as modern-day AI for me is to see if said thing was around 10 years ago.
One exception for where I do use AI is when I need to do so to report on it. For example, I used the Google AI Overview while creating this blog post to see what it says about me, obviously. I would not have been able to report on it otherwise.
Most people would just put an mp4 on a CDN to avoid the hosting costs… glad to know the spirit of CSS criminality persists in this day and age.
Also, if you like those CSS tricks, you'll love this: https://lyra.horse/css-clicker/
But as soon as the episode ended, there were results for the search! But they're all "news" sites referencing the episode that had just finished. "news" in quotes because they all felt bot-written (this was before LLM). For me it was a surreal experience about the speed of slop...
[0]: https://ncase.itch.io/wbwwb
if you are critical of how technology works and have seen how deep the rabbithole does (prism was 2012, advanced AI already existed by then afterall, most household devices are smart enough to be tracking you today, and these are also deployed in public spaces..) then you can see where boundaries are being eroded and try to create your own space where this cannot reach into.
this is why i enjoy systems engineering. a way to create your own space, and say where boundaries cannot be eroded.
some software systems already attempt this. (notably Qubes OS), but more work is needed. always more work, so little time :)
But in the story they also had constant news coverage, so whenever anything of interest (whether good or bad) would happen millions of people would just teleport over to go and see for themselves.
I don't think it is even possible to fully escape using AI now, especially if you work in any form of IT, especially especially if you are using the web.
In my estimation, the only way you can 'not use AI', is if you don't use a computer at all, and that includes the ubiquitous smart phone.
For example, if I need to talk to the support of a company and I get throws into a chatbot chat, I will still use it, though mostly to try to get a human to talk to me. If a website or app decides to show me an AI summary, I ignore it.
For programming and writing, I use Sublime Text without a language server. For talks, I use my own slideshow software. For most other tools, provided they don't expose a security-critical surface, I have older versions saved that I use.
There are some things I use that some people might technically consider AI but I don't, such as search engines (outside of AI summaries), smartphone keyboards, machine translation, and NPCs in video games. I think an easy way to differentiate what counts as modern-day AI for me is to see if said thing was around 10 years ago.
One exception for where I do use AI is when I need to do so to report on it. For example, I used the Google AI Overview while creating this blog post to see what it says about me, obviously. I would not have been able to report on it otherwise.