Fediverse

If you are able to, it's really important to add alt text descriptions when posting an item of media on here 🙂

If you forget to add descriptions, there are two ways you can be reminded. Both methods are covered in this guide:

➡️ https://fedi.tips/how-do-i-remember-to-add-text-descriptions-on-mastodon-and-the-fediverse

#FediTips #AltText #Accessibility #Mastodon

for now, I've more-or-less settled on switching back from Proton Mail over to Fastmail (they get big bonus points for their desktop app being Flathub verified)

as I slowly get around to changing over my online accounts from SimpleLogin aliases, I'm doing the new aliases through my custom domain this time

technically, not very private that way, but also way more convenient if I ever change email service providers again in the future

I really should listen to my own advice here and get Omarchy Quattro out the door, shouldn't I? It's a lot better, a lot more ambitious, and full of delightful touches. By far and away the best operating system I've ever used.

Tomorrow!

@rockerest cc @ecomba

I don't know of a single source… the fuck_ai lemmy community is a helpful place to keep up with news.
Also Futurism.

This is the most recent study I'm aware of:
https://osf.io/preprints/psyarxiv/5y6m4_v1

It basically demolishes the whole "human in the loop" concept as a valid accuracy check. Frightening stuff.

@chergert
I'm unable to find a ticket about having something better on Flathub to allow debugging/profiling without downloading LLVM debuginfos (but I'm not entirely sure where to look, either). Is there one somewhere out there that I can bookmark and subscribe to?
@hub

@KekunPlazas @bugaevc
Fixing https://bugs.webkit.org/show_bug.cgi?id=319590 should help, but beyond that, until someone implements https://bugs.webkit.org/show_bug.cgi?id=291369 and https://gitlab.gnome.org/GNOME/epiphany/-/work_items/1835 (and/or its related issues) then GNOME Web will remain an unmitigated RAM glutton :blobmiou:

@phillycodehound Without resorting to the genie is out of the bottle or calculators or photographs or left behind or any of those other silly analogies, can you explain why it would be bad *not* to use Generative AI? 🤔

@phillycodehound Where did I say they have no utility? Arguments of "utility" need to be matched with arguments of safety and many other considerations. Asbestos has tremendous utility, that's why it was used! Does that mean we should use it? Hell no!

@ianturton interesting. Do you kind of dig it out?

Just a slow morning of site clean up and coffee.

https://readbeanicecream.codeberg.page/

#indieweb #blogging #smallweb #plaintext #opensource #personalsites #tech #smolweb #foss #pandoc #asciidoctor #productivity #rss #blogpost #blog

@kevin I have no doubt about a good hole saw having the ability to cut it, but a hole saw needs to be longer than the depth it cuts, otherwise you just end up with a ring and a hole.

I don’t think a typical hole saw could do it in a front/back pass.

@erfanekm نودل 😅

@bktandem Oh no, wishing you a speedy recovery!

@BlowUpRadio

@fredldotme @dannycolin Do we have any numbers on those?

The best way to avoid getting drunk is to abstain from large quantities of alcohol.

The best way to avoid the destruction of your mental acuity is to abstain from engaging with generators.

C'mon, it's not that hard.

Just a reminder, in case anyone is wondering…

We now have multiple studies showing that people who regularly engage with LLMs are literally suffering from impaired judgment.

That's why you can never trust anyone who claims they are now able to do their job better by using these "tools". Using these wretched slop machines is actually causing them not to be able to do their job as they previously did.

These things are like drugs for minds, and not the fun enlightening kind. 🙄