John Cox

Blogs · @wyome

Tagged artificial intelligence · everything

The Verge

Our minds aren’t equipped to handle AI | The Verge

Benjamin Riley argues that the dominant computational model of the mind, which treats thinking as input, computation, and output, is a warped mirror that helps explain why AI tools are so damaging. Drawing on Paul Cisek's feedback-control view of the brain and the cultural history of imitation, language, and schooling, he contends that AI is a "cognitive hot dog": tempting in the moment, but harmful when it becomes a regular part of how people learn. The piece is distinctive for grounding its critique of AI in evolutionary neuroscience and for calling for cognitive immunization, including school bans and protected spaces for unaided thinking.

Permalink technologyartificial intelligence

overcast.fm

WIRED

Stop Touching Your Keyboard. Use This AI-Powered Microphone Instead

Relay is a London startup building dedicated hardware and software for voice-to-text dictation, addressing the problem that existing transcription tools work poorly in shared office spaces where whispering is too quiet and normal speech is too loud. The company's Relay Q microphone, launching in early 2027 at $150, is designed to capture accurate dictation at any volume, while the accompanying macOS software (arriving today) uses Google's Gemini models to transcribe, clean up speech, and execute contextual actions like scheduling calendar events or drafting messages in apps like Slack and Gmail. The software's transcription accuracy is respectable but not yet matching Google's latest offerings, though the dedicated hardware should improve performance compared to using standard computer microphones.

Permalink productivitytechnologysoftwareartificial intelligence

Ahrefs

Ahrefs—AI Marketing Platform Powered by Big Data

This is a marketing/landing page for Ahrefs, positioning it as a marketing platform beyond traditional SEO—covering search rankings, AI chatbot visibility (brand mentions/citations in tools like ChatGPT), competitive intelligence, and content optimization. It highlights new offerings like "Letaido" (an AI agent platform for automating marketing workflows) and real-time web-crawling/API products aimed at grounding AI/RAG pipelines with fresh data. Worth keeping likely as a reference for Ahrefs' expanding feature set and its pivot toward "answer engine optimization" (AEO) as AI search reshapes SEO practice.

Permalink softwareseoartificial intelligence

www.nytimes.com

Five Science Fiction Movies to Stream Now

"Significant Other" is a sci-fi horror hybrid about a couple whose backpacking trip in Oregon goes catastrophically wrong after a mysterious object falls from the sky, generating dread through atmospheric filmmaking rather than jump scares while exploring themes of intimacy and deception. "Infinity Pool" follows a blocked writer on vacation who escapes execution by having a clone created and executed instead, then descends into hedonistic depravity with wealthy tourists who exploit the loophole, satirizing how the rich can literally buy their way out of consequences. "Costa Brava, Lebanon" depicts a self-sufficient family's isolation shattered when a landfill is planned next to their compound, using their fragmentation as a metaphor for environmental collapse and societal breakdown.

Permalink filmhorrorstreamingsci-fi

www.nytimes.com

Opinion | ChatGPT Is Already Changing How I Do My Job

Manjoo describes how ChatGPT is already becoming an indispensable tool in his journalism workflow and others', comparable to past tech breakthroughs like the iPhone and Google search. He outlines specific practical uses—finding the right word, overcoming writer's block, and summarizing complex documents—while emphasizing the critical caveat that journalists must treat ChatGPT's output as unreliable without verification, since it frequently generates false information and operates as a black box. The piece warns that while the technology shows genuine promise for newsrooms, publications must develop careful guidelines for its use, particularly when it comes to selecting sources or framing questions, since relying on this "mysterious oracle" for editorial decisions risks embedding unknown biases into journalistic work.

Permalink technologyaiopinionethics

movieweb.com

8 Modern Sci-Fi Movies That Will Someday Become Cult Classics

A roundup piece nominating eight 2010s-2020s sci-fi films — including The Martian, High Life, Godzilla (2014), Underwater, Annihilation, Edge of Tomorrow, The Invisible Man, and M3GAN — as future cult classics, arguing each blends genre conventions with distinctive artistry, divisive tone, or emotional depth that could earn devoted fandoms over time. It's less a ranked list than a case for why each film's particular weirdness or ambition (from Mark Watney's potatoes to M3GAN's sassy menace) sets it apart from typical studio sci-fi. Worth keeping as a watchlist with actual arguments attached rather than just titles.

Permalink film2010ssci-fia24

The Atlantic

We're Underestimating the Risk of Human Extinction - Ross Andersen - Technology - The Atlantic

# Summary Nick Bostrom argues that human extinction risks are severely underestimated and that the most significant existential threats arise from future technologies—such as artificial intelligence, synthetic biology, and nanotechnology—rather than natural disasters like asteroids. He contends that if we value future generations equally with present ones, reducing existential risk becomes a moral priority outweighing current humanitarian concerns, since the potential number of future humans vastly exceeds those alive today. Bostrom also explores how observation selection effects complicate our understanding of extinction risk, discusses the simulation argument (which suggests we may be living in a computer simulation), and proposes that meaningful solutions require careful technological development alongside analysis to identify and mitigate the greatest risks.

Permalink technologyphilosophyartificial intelligence

www.simulation-argument.com

What's the probability that we're living in the Matrix?

Bostrom presents a trilemma arguing that at least one of three propositions must be true: either most civilizations at our development level go extinct before becoming technologically advanced, or advanced civilizations have no interest in running simulations of minds like ours, or we are almost certainly living in a simulation ourselves. The argument rests on the assumptions that consciousness can be implemented on any sufficiently powerful computational substrate and that advanced civilizations could generate an astronomically vast number of simulated minds—vastly outnumbering biological ones—making it probabilistically likely that any given mind is simulated rather than organic. Even if we accepted premise (3), Bostrom contends we should continue living normally using standard empirical methods, though the simulation hypothesis might subtly influence behavior if we made assumptions about the simulators' intentions or moral frameworks.

Permalink philosophyscienceinternetartificial intelligence