😆
Meta Director of AI Safety Allows AI Agent to Accidentally Delete Her Inbox
https://www.404media.co/meta-director-of-ai-safety-allows-ai-agent-to-accidentally-delete-her-inbox
#tech #technology #BigTech #AI #ArtificialIntelligence #LLM #LLMs #ML #MachineLearning #GenAI #generativeAI #AISlop #Fuck_AI #Meta #AIAgent
😆
Meta Director of AI Safety Allows AI Agent to Accidentally Delete Her Inbox
https://www.404media.co/meta-director-of-ai-safety-allows-ai-agent-to-accidentally-delete-her-inbox
#tech #technology #BigTech #AI #ArtificialIntelligence #LLM #LLMs #ML #MachineLearning #GenAI #generativeAI #AISlop #Fuck_AI #Meta #AIAgent
#AI is the aid I've needed my entire life. I'm not going to mince words here. People making blanket statements about the technology without understanding it are my enemies.
My #ADHD is crippling. #LLMs are the exact thing that I've needed. I do not let them do work for me, but they do keep me working by providing constant and immediate feedback to whatever I'm doing.
My work from now till my death is likely going to center on how to make an #llm or any aspirational #AGI aligned with humanity.
Fundamentally, every problem y'all have with #AI was an already existing problem under #capitalism that AI is exposing.
This includes:
- Alienation from labor
- Corporate piracy
- Slop
- Environmental destruction and other externalities
- Wealth inequality
- Replacement of labor with capital
EVERY SINGLE ONE existed before.
Additionally, a ton of the problems, like layoffs, aren't even caused by AI, and blaming them on AI is _specifically_ corporate propaganda for what amounts to a criminal conspiracy by mega corporations to suppress wages.
#AI is the aid I've needed my entire life. I'm not going to mince words here. People making blanket statements about the technology without understanding it are my enemies.
My #ADHD is crippling. #LLMs are the exact thing that I've needed. I do not let them do work for me, but they do keep me working by providing constant and immediate feedback to whatever I'm doing.
My work from now till my death is likely going to center on how to make an #llm or any aspirational #AGI aligned with humanity.
Fundamentally, every problem y'all have with #AI was an already existing problem under #capitalism that AI is exposing.
This includes:
- Alienation from labor
- Corporate piracy
- Slop
- Environmental destruction and other externalities
- Wealth inequality
- Replacement of labor with capital
EVERY SINGLE ONE existed before.
Additionally, a ton of the problems, like layoffs, aren't even caused by AI, and blaming them on AI is _specifically_ corporate propaganda for what amounts to a criminal conspiracy by mega corporations to suppress wages.
Reposting a question for Ed Zitron, I'll forward responses. He asked on Bluesky and will get sub-Mastodon-tier answers:
"This is a serious question and I would be delighted if I only hear great things but, software engineers: both before and after LLMs, how often in your professional lives have you run into software engineers that seem completely useless or lacking in basic knowledge? I hope the answer is rarely"
For the benefit of people (like me) who never heard of Ed Zitron:
<https://www.wheresyoured.at/about/>
<https://redd.it/1rak1tp> includes a link to his post on Bluesky.
Tried writing a blog post summarizing my little #LLM coding agent experience, but its completely pointless.
It works better than I would have thought, I assume it will keep improving, it introduces hard dependencies to SaaS providers, prices will be extremely high in the future, it's exploiting people, trained on code with unclear licensing status, burns tons of energy and (most vendors) support fascists. Experiment ended, hard pass.
I want pro-AI people on the fediverse either publicly or through DM if they don't want to come out of the dev closet to tell me what do they personally think of the copyright issues that might be raised by them using AI to code.
In the meantime just dropping this here: https://www.theguardian.com/technology/2025/jun/25/anthropic-did-not-breach-copyright-when-training-ai-on-books-without-permission-court-rules?CMP=Share_AndroidApp_Other
Once I get enough diversified answers I'll write a longer toot or freely accesible article (no names, or PII) with what o gathered for everyone to see. Please boost for visibility and reach.
Je reposte une fois de plus ce texte (en anglais) de @baldur, parce qu’il est d’une importance capitale, et que c’est une réponse parfaite pour éviter de tomber dans une spirale de Brandolini.
https://www.baldurbjarnason.com/notes/2026/note-on-debating-llm-fans/
« Going all “but it works great for me” even as the industry burns around you and the “it” is a right-wing political project built on disregarding consent, being applied to dismantle public infrastructure and institutions, is fundamentally a dick move. »
@deathkitten Check @adafruit blog, there are a lot of articles about using #LLM in electronics "design". If they are going to adopt LLMs, then there are no surprise that they are trying to whitewash it as much as possible 
, using all available methods.
When I was learning electronics at university, drawing such simple electronic schemas and footprints was a key to be able to do much more complex things with EDA. Outsourcing this mental work to Plagiarized Information Synthesis System is a literally: "you will learn nothing and be happy" 
@[email protected] as a woman I am offended that you'd compare the actual bullshit our foremothers faced under sexism to people pushing back against AI, a "service" that's buggy, harmful, ecologically damaging, and being used as a tool in class warfare to push people out of their jobs, something that disproportionately happens to women and minorities first.
You should be ashamed of yourself.
@deathkitten Check @adafruit blog, there are a lot of articles about using #LLM in electronics "design". If they are going to adopt LLMs, then there are no surprise that they are trying to whitewash it as much as possible , using all available methods.
When I was learning electronics at university, drawing such simple electronic schemas and footprints was a key to be able to do much more complex things with EDA. Outsourcing this mental work to Plagiarized Information Synthesis System is a literally: "you will learn nothing and be happy"
If the model impact is largely fictitious, meaning this is all a scam and the perceived benefit is just a clusterfuck of cognitive hazards, then the financial bubble pop will be devastating, tech as an industry will largely be destroyed, and trust in software will be zero, collapsing the job market
No. The so far fictitious "AI bubble" is not going to be the year zero reset tsunami antagonists hinge everything on.
It will be worse, for everyone.
It will be more like the dot.com bubble.
Thousands of big players wiped out...concentration of few, big players emerging.
Dot.com gave us the pestilence of Facebook, YouTube, twitter and google.
There will be two, three global #AI winners, which is worse.
The exception being smart players (like the UK and China) who are developing #sovereignAi
TLDR; Don't dance around bonfires in the woods praying for AI bubble, it's not the fix you think it is, #regulateAi
#ai #llm
“AI can make mistakes” might as well be the slogan of our era. Even boosters admit that you need to spin the vibe code slot machine a few times to get a jackpot.
An employee with that degree of consistency would be fired.
So how do we redirect some of that unlimited grace from machines to humans?
“AI can make mistakes” might as well be the slogan of our era. Even boosters admit that you need to spin the vibe code slot machine a few times to get a jackpot.
An employee with that degree of consistency would be fired.
So how do we redirect some of that unlimited grace from machines to humans?
Heard today: We need to use LLMs to write code faster. We need to use LLMs to generate user stories faster.
My take on this: you can do that or you can look for the underlying issues. More content and more code will not save you. Someone has to read, understand and maintain this.
To say it with the words of @pluralistic "Code is a liability". The same goes for me with content aka user stories. The longer they are, the more likely they won't be read and contain conflicting points.
Heard today: We need to use LLMs to write code faster. We need to use LLMs to generate user stories faster.
My take on this: you can do that or you can look for the underlying issues. More content and more code will not save you. Someone has to read, understand and maintain this.
To say it with the words of @pluralistic "Code is a liability". The same goes for me with content aka user stories. The longer they are, the more likely they won't be read and contain conflicting points.
