Share! 🎉🎉

I'm just a former cybersecurity professor who quit her job because the Open Institute of Technology was going to force me to subject my students to Gen AI.

Kickstarter is my sole source of income for the time being, and I'm really struggling financially. I wrote books for O'Reilly and Wiley, and now I write self published books criticizing Big Tech.

I have an excellent track record with Kickstarter campaigns.

Gen AI will end. The laws of physics can't sustain datacentres. Financial bubble. Model collapse. I cite MIT, UN, and Harvard Science Review.

What will we do in the aftermath to recover from massive population psychological damage, environmental destruction, tech debt, abandoned datacentres?

Pledge as low as about $12 USD for an eBook.

Share! 🎉🎉 #NoAI

Kickstarter link:
https://www.kickstarter.com/projects/kimcrawley/recovering-after-gen-ai

Old pic from 5yrs ago.

Cyberpunk 2077 patch: 1.5
OS: Win10
GPU: Nvidia 3070 Ti

Mods: NexusMods, discord-only mods from the two major modding discords.

Reshade, lighting and camera detachment shots courtesy of 'Otis_Inf Photomode Mods'. His Cyberpunk 2077 photomode mod only works on Windows and is behind a Patreon paywall.

#cyberpunk2077

😂 welcome to reality. The AI bubble crash is going to be spectacular one 💥

LLMs and hardware don't mix. If you have to wrangle hardware, then other than throwing chipset data and offsets at one on the off chance the underlying DNN will emit some driver code for you, they are not going to be much use to you. OK, so digging around in my notes on BeagleBone Black from Feb/Mar earlier, because another one just arrived this morning. I use these things for NTP, IEEE 1588, and remote management applications.
This related quite specifically to a BBB I acquired from eBay from a UK seller, which had a proprietary Linux image installed for automotive use. I couldn't cleanly reflash it out of the box.
I noticed a lot of profanity in there about trying LLMs for advice, and getting sent on the proverbial "tartan paint errand run", but was far more serious was erroneous output which could have led to bricking the BBB by getting the eMMC addresses and offsets wrong when reflashing firmware, quite specifically the "hallucinations" about the correct values for /etc/fw_env.config. IT WAS MAKING THIS SHIT UP. They are entirely specific to the flash layout the system image is using. One wrong move and I'd have bricked it. Not least of all the useless LLM suggestions for trying to upgrade an existing Debian image with apt when it was just TOO OLD... a rude reminder that LLMs do not do *context* in the human semantic & linguistic sense, and "context windows" are implementation specific.
@davidgerard @kimcrawley @jonny

sorry but "ai" is just a big ol' load of crap

I really don't need a domain for this but if I were to spend a tiny amount of money for a stupid new web ring

Idiots.

“US government lists fictional nation Wakanda as trade partner”

https://www.bbc.com/news/world-us-canada-50849559

We make privacy and security tech. Privacy and security involve defending against adversaries actively adapting and improving. Companies developing offensive security tech including remote exploits and data extraction tools are aware of GrapheneOS. It's made with that in mind.

This is one of the best ideas I have ever had.

#SysadminLife #organization #technology

For those wondering: I did this by fucking hand, it took me ~5 hours.

I tried to be as surgical as I could with the removal, so there is likely still a fair amount of vestigial code that I didn't notice or overlooked.

I also didn't intentionally remove any of the AI-related code in the android source, in an attempt to minimize the scope.

If you are interested in this absurd project and are comfortable with building firefox from source, feel free to test it out and report any issues.

RE: https://mastodon.social/@sarahjamielewis/116967027925557386

To follow up on this, I have pushed an initial, experimental, exploratory patch to https://codeberg.org/basebrowserproject/organization

This patch removes (not just disables) many AI features from desktop Firefox including: smart windows, model chat, link previews, smart tab groups, semantic search, on device model integration (translations) etc.

1605 files changed, 98 insertions, 852297 deletions

I've tested that Firefox builds, and runs, and overall doesn't complain, but it's far from complete, hic sunt dracones etc.

I have published my initial #OpenBSD #WPA3 implementation.

Thanks to @nlnet for supporting this effort.

Cc @vanhoefm

https://marc.info/?l=openbsd-tech&m=178437520390575&w=2

I'm pretty angry that the Mastodon maintainers let one co-authored claude commit slip by despite having an explicit no-LLM/AI policy.

That one commit makes the whole repo look like slop.

You don't need a fucking LLM to make this tiny change.

#Mastodon #noai #nollm

Been trying to find an open source accounting software that also handles purchase orders and invoices. The ones I've found are full-on slopified.

I took a 2nd look at GnuCash which doesn't explicitly support POs or invoices.

Well, their first LLM commit was this past April, making their latest release v5.16 slop.

GnuCash is my best option so I'm going to fork v5.15 (latest clean release) to add PO/Invoice support. Will manually handle POs and invoices in the short term.

We're making slow progress on survival funds anything helps

$626 / $1500

Paypal/Venmo - $nullagent

https://paypal.me/nullagent

https://venmo.com/code?user_id=3062903129243648674&created=1783951504

#MutualAidRequest #HelpFolksLive2026 #MutualAid #ParentingWhileBlack #BlackDadsMatter

EDIT: Repos are being removed! https://github.com/bigcode-project/opt-out-v2/issues/3436

KEEP FIGHTING!!!

HuggingFace scraped all of GitHub in 2025.

Your repos are definitely in here, mine were despite them being deleted halfway into last year.

https://huggingface.co/spaces/HuggingFaceCode/in-the-stack

Look up your username to check and use the buttons to below to let them know you do not want to be in their dataset.

Fuck HuggingFace and fuck AI!

A nasty person is creating endless accounts to harras, heads up to other admins

Keep an eye out for a troll with the username that's something like:

Yevhenuia
Yevheniuya
Yevhgeniia

IP:
109.81.87.246
45.154.205.172
89.185.84.15

Emails:
repostrants@ukr.net
oleh9191@outlook.com
oleh8585@outlook.com
oleh8990@outlook.com

I strongly recommend blocking it all!

#Fediblock

It is both hilarious and inevitable that the AI bros are kicking off about the #Codeberg membership voting by supermajority against #LLM contributions and vibe coding.

The social context of the modern "AI" movement is based around a fascist idea that "might means right", and that we plebs can't stand up to the force of the TESCREAL billionaire class. And some people really get on board with that, perhaps thinking that they can curry favour with the billionaires who don't even know they exist. So even as small and insignificant a resistance as a tiny German non-profit saying "no" is unconscionable to them.

This is why my inbox has been filled with screeds and rants and bad-faith arguments for the last few days and I've been laughing at them and deleting them.

I don't get why people who want to use LLMs to generate code are grumpy because Codeberg doesn't want to host it.

Surely they can just use their LLMs to generate a replacement for Codeberg, which they can operate on whatever basis they choose?