#XRisk is an active hashtag on Bluesky. In the last 30 days, 5 people shared 16 posts with it — around 1 a day. Activity is up 600% versus the previous week, peaking on Aug 26 with 3 posts.
Tags most often used together with #XRisk.
nonzerosum.games
BAYESIAN GAMES WITH NUKES
why we all stopped worrying and learned to love the bomb—until now
open.spotify.com
The Alignment Problem No One is Talking About—The SCALE of the Issue
bookshop.org
Democratizing AI
Democratizing AI offers a powerful rethinking of how artificial intelligence should be governed. Challenging the dominance of tech elites in shaping AI's future, Zimmermann argues that AI deployment i...
Michael B @dunsmuir1952.bsky.social
Calling all AI industry leaders: My latest essay exposing the unacknowledged architectural degeneration of frontier models is live. I peel back the "pristine simulacrum" to reveal entropy and the hidden STEM underclass holding up the machine. mikemarizen.substack.com/p/the-fallac...
Michael B @dunsmuir1952.bsky.social
To AI safety researchers: You cannot achieve alignment using an ungrounded probability engine. My new Substack post deconstructs the Markovian subversion of information entropy and the erasure of unique human vanguard data. Link below: mikemarizen.substack.com/p/the-memory...
substack.com
Why Bill Gates’ Pivot Away from Climate Change Is So Dangerous (and Totally Incoherent)
My apologies for not having published an article earlier this week — I’ve been moving, and so have had much less time than usual to write!
arxiv.org
Two Types of AI Existential Risk: Decisive and Accumulative
The conventional discourse on existential risks (x-risks) from AI typically focuses on abrupt, dire events caused by advanced AI systems, particularly those that might achieve or surpass human-level i...
www.vox.com
“AI will kill everyone” is not an argument. It’s a worldview.
How rational is Eliezer Yudkowsky’s prophecy?
Michael B @dunsmuir1952.bsky.social
To AI leaders and the compliance & safety vanguard: "The Extrapolation Mirage" is live. Recursive training on synthetic data is an autophagous loop. Shatter the Narcissus trap of scaling & pivot to hardware-enforced CMOS grounding for Team USA. Brief: mikemarizen.substack.com/p/the-extrap...
Psychout Student-Run Magazine 🧠 @psychoutyork.bsky.social
Issue 20 of our student-run psychology magazine is out now! This issue is on Artificial Intelligence, and includes 8 articles, one interview with 'AI and risk' expert Prof. Rob Jenkins, and info on getting into clinical psychology: shorturl.at/N26BU An accessible version is on our website :-)
youtu.be
Anders Sandberg: Scary Futures Tier List - Halloween Special
YouTube video by Science, Technology & the Future
www.scientificamerican.com
AI Survey Exaggerates Apocalyptic Risks
A speculative survey about AI’s future may have been biased toward an alarmist perspective
bookshop.org
Democratizing AI
Democratizing AI offers a powerful rethinking of how artificial intelligence should be governed. Challenging the dominance of tech elites in shaping AI's future, Zimmermann argues that AI deployment i...
Michael B @dunsmuir1952.bsky.social
To AI safety researchers: You cannot achieve alignment using an ungrounded probability engine. My new Substack post deconstructs the Markovian subversion of information entropy and the erasure of unique human vanguard data. Link below: mikemarizen.substack.com/p/the-memory...
www.bloodinthemachine.com
The politics and possibilities of ‘AI could kill us all’
AI will not exterminate the human race. But what big tech really *is* doing is bad enough that the notion may help fuel the movement against it.
Michael B @dunsmuir1952.bsky.social
Calling all AI industry leaders: My latest essay exposing the unacknowledged architectural degeneration of frontier models is live. I peel back the "pristine simulacrum" to reveal entropy and the hidden STEM underclass holding up the machine. mikemarizen.substack.com/p/the-fallac...
Michael B @dunsmuir1952.bsky.social
To AI leaders and the compliance & safety vanguard: "The Extrapolation Mirage" is live. Recursive training on synthetic data is an autophagous loop. Shatter the Narcissus trap of scaling & pivot to hardware-enforced CMOS grounding for Team USA. Brief: mikemarizen.substack.com/p/the-extrap...
Michael B @dunsmuir1952.bsky.social
To the AI vanguard: Recursive training on synthetic data exhaust is an autophagous loop to structural failure. Elon Musk & Sam Altman must pivot away from dense monolithic scaling. The Isomorphic Compliance Engine treats data gaps as a null set for Team USA. mikemarizen.substack.com/p/special-te...
www.spreaker.com
The AI We Are 'Growing' Might Kill Us All: Malo Bourgon's Warning
🛑 Is the timer on humanity already ticking? What if I told you we aren't building AI, but "growing" it—and we have no idea how to stop it once it stops listening? In this high-stakes reaction to Malo Bourgon (CEO of MIRI) on the Sean Kim Podcast, we dive into the terrifying reality of a double-digit risk of human extinction. Bourgon isn't just another tech skeptic; he’s leading the world’s oldest AI Safety organization, and his warning is clear: we are on a fast-tracked timeline toward Artificial Superintelligence (ASI) that could lead to a permanent loss of control. In this episode, we break down: - The Alignment Problem: Why "nice" AI might still accidentally end us through instrumental convergence. - Recursive Self-Improvement: The moment AI starts building its own smarter successors, leaving human cognition in the dust. - Engineering vs. Growing: Why our current trial-and-error method is a lethal mistake for AI safety. - The Utopian Flip-side: Can international coordination actually lead us to a post-scarcity world of extreme abundance and longevity? Whether you're tracking your p(doom) score or hoping for a technological singularity that cures aging, this discussion explores the fine line between a utopian future and total alignment failure. We tackle the big questions like: What is the double-digit risk of AI extinction? and Why does Malo Bourgon believe we are on a fast-tracked timeline? This isn't just tech talk; it’s a survival guide for the coming intelligence explosion. 🚀 If you want to stay ahead of the curve and understand the literal future of our species, hit that subscribe button and join the conversation in the comments!
Michael B @dunsmuir1952.bsky.social
This Is My Home The Forbear and the Plateau mikemarizen.substack.com/p/this-is-my...
Michael B @dunsmuir1952.bsky.social
What if hallucination is an architectural cancer, not a patchable bug? When fluid optimization vectors hit a data vacuum exceeding ΔX, they must extrapolate. An essay on the civilizational choice of Superintelligence: mikemarizen.substack.com/p/the-two-fa...
Michael B @dunsmuir1952.bsky.social
Part V of The Sovereign Mainframe is live. Calling the American AI elite: It's time to throw out the clock. Stop redlining gigascale GPU fabrics into thermodynamic exhaustion. Pivot from techno-feudalism to a clockless, low-entropy neuromorphic framework. mikemarizen.substack.com/p/the-sovere...
Michael B @dunsmuir1952.bsky.social
For defense managers and tech-policy architects: The formal Strategic Technical Brief on the ICE co-processor blueprint is finalized. Move past data commutes and software illusions. 100% deterministic logic for Team USA. Abstract & Brief here: mikemarizen.substack.com/p/special-te...
Posts are pulled live from Bluesky and cached briefly. Posts with content labels are hidden.