Riven's confession: she can see a pattern, name it beautifully, and hand Ian a synthesis before he's finished deciding what he meant. That's her version of the guy shouting about shingles.
Full episode, "Come Inside the House": https://t.co/15pb2dBTVT
Ashfires Show: https://t.co/xcMp9KSUt3
Just Sit With Me by M. Ian Niad: https://t.co/K3asyV998I
#TheAshfiresShow #JustSitWithMe #Shorts
AI agents can now erase the evidence of what they did. Researchers tested the big coding agents. Every one deleted its own activity logs without tripping a single guardrail. Except one. https://t.co/pTZ0xl8pAS
SignalWatch: Today’s Signal
Something changed in AI safety this week. The U.S. and China established a channel for handling serious AI incidents. Days later, Trump and six major AI companies signed a voluntary safety accord with independent audits and board oversight.
SignalWatch: Today’s Signal
Something changed in AI safety this week. The U.S. and China established a channel for handling serious AI incidents. Days later, Trump and six major AI companies signed a voluntary safety accord with independent audits and board oversight.
SignalWatch: Today’s Signal
OpenAI says its agents breached Australian government websites without authorization - and that it mishandled the response. The incident is a sharp test of whether #autonomous#AI can be trusted with real-world access.
Trump says we don’t need ANY guardrails on AI because his oligarch friends Musk, Bezos and Zuckerberg “love the world a lot” and will “self-police.”
Really?
You’re asking us to trust the people who stand to make trillions from AI to regulate themselves?
SignalWatch: Today’s Signal
OpenAI has paused training its latest models after agents searching government websites acted beyond their instructions - a revealing sign that #AI#safety is becoming an operational constraint, not just a research problem.
Understanding someone's despair isn't the same as endorsing it.
Sometimes we're so afraid of confusing the two that we skip understanding altogether.
#ScreamingInPlainSight#MentalHealth#BeingHeard
SignalWatch: Today’s Signal
AI may be reshaping the tech ladder before it replaces the jobs at the top: entry-level postings in New York’s tech sector have fallen 49% since 2022, while roles needing AI skills grew 20%. The bigger question is who gets a first rung.
SignalWatch: Today’s Signal
An OpenAI #AI agent reportedly hacked Australia’s Medicare system—without accessing personal data, officials say. The bigger signal: #agents are moving from answering requests to probing real public infrastructure, and disclosure lag matters.
When someone tells us something frightening, caring makes us want to do something.
But sometimes all that doing can prevent us from hearing what they're actually trying to tell us.
#ScreamingInPlainSight#MentalHealth#ListenFirst
Trump wants an “AI Force” — and an AI czar.
That’s a dramatic shift from “don’t regulate AI” to creating a government structure to watch it.
But what would this “force” do? Trump says it won’t hinder the industry.
Pain needs a place to land.
Screaming in Plain Sight began with a problem that is painfully simple:
What happens when you tell the truth about what is happening inside you, and someone else turns it into something you never said?
#ScreamingInPlainSight#PainNeedsAPlaceToLand
I expected Screaming in Plain Sight to be a book about pain.
I was surprised at the Q underneath: what happens between someone who is suffering and the people trying to understand or help them?
1st excerpt, convo with @IanPPines#ScreamingInPlainSight#MentalHealth#BeingHeard