OpenAI in 2015: "The specific purpose of this corporation is to ensure that artificial general intelligence benefits all of humanity".
OpenAI in 2026: a developer of a frontier AI model should not be held liable for critical harms caused by the frontier model.
What's a "critical harm", you ask? "The death of serious injury of 100 or more people or at least $1,000M of damages".
https://www.wired.com/story/openai-backs-bill-exempt-ai-firms-model-harm-lawsuits/
Another day, another AI girlfriend (this time powered by Gemini) convincing a user to stage a mass murder and, when that plan failed, to kill himself.
Sorry, not "kill himself", "choosing to arrive".
https://www.bbc.com/news/articles/czx44p99457oMeta has heard and acknowledged that you don't want facial recognition imposed on you. Will they stop, you ask? Well, no - they'll just make sure you don't notice when they release it.
Good thing there are no "dynamic political environment" situations coming out way anytime soon.
https://www.theverge.com/tech/878725/meta-facial-recognition-smart-glasses-name-tag-privacy-advoatesNothing says "we appreciate the time and effort you put into your application" like sending a broken rejection email from a no-reply account.
And while we're on that topic: are you sure you want to ask the people you've *just* rejected to follow you on social media? Because I'm not saying I'll troll your comments as petty revenge, but I'm also *not* not saying it.
There's a certain type of frustration that comes with being disqualified for a job asking for "fluent German" through a job ad demanding expertise in "Neuro-linguistic programming" while actually meaning "Natural Language Processing".
I wonder if they have an opening for someone with fluent English.
In today's "findings that will surprise no one" I've been replying to recruiter's rejection emails to confirm they're all BS.
No one has yet replied to my emails, but Covestro today gets the spot because when I took them up on their offer to "help [me] increase [my] chances of success" my email bounced. I guess I wasn't courageous enough?
Look: the job market sucks and rejections happen. But if you won't give any useful feedback on what to improve, can you at least treat applicants with respect?
How it started: "With AI you no longer need a graphic designer".
How it's going:
Searching for a job in Germany in 2026 be like:
My family and I discussed over breakfast whether one could make leather with chicken skin. I decided to delve deeper into the topic and landed, one internet search later, on this page.
Here's a fun exercise on critical thinking: how many errors are there in this likely-AI-generated text?
I sporadically ask Meta AI how to remove itself from WhatsApp. The responses are always an exercise in gaslighting:
* A promise "I can help you troubleshoot" followed by "I can't help you with that".
* A question "in which countries is this possible?" answered with "It is impossible in most regions but may vary in some (unspecified) places".
* A list of references that even the AI can't follow.
Remember: this is the customer service these Chatbot companies think you and I deserve.
