Susi O'Neill
@susioneill.bsky.social
Tech communicator, speaker, author, thereminist, early electronic music nerd.
Weekly tech & AI insights: https://rethinkhypecycle.substack.com/
Weekly tech & AI insights: https://rethinkhypecycle.substack.com/
www.oecd.org/en/publicati... fragmented policies, limited cross-border co‑ordination, and underdeveloped data-sharing frameworks: how EU members are getting on so far with establishing the coordinated plan for AI
Progress in Implementing the European Union Coordinated Plan on Artificial Intelligence (Volume 1)
This report reviews the implementation of national strategies and policy initiatives by EU Member States to foster the development and uptake of artificial intelligence (AI), in line with the European...
www.oecd.org
November 10, 2025 at 9:57 PM
www.oecd.org/en/publicati... fragmented policies, limited cross-border co‑ordination, and underdeveloped data-sharing frameworks: how EU members are getting on so far with establishing the coordinated plan for AI
open.substack.com/pub/mitramem... Mitra Memarzia's new Substack explores how the promise of authenticity can be an uncanny valley of performance in the hands of genAI
The Architecture of Trust
Designing for integrity in a world that can fake almost anything (7-minute read)
open.substack.com
November 10, 2025 at 9:39 PM
open.substack.com/pub/mitramem... Mitra Memarzia's new Substack explores how the promise of authenticity can be an uncanny valley of performance in the hands of genAI
Wow. Shark/jump moment. Kim Kardashian's ChatGPT 'frenemy' lies then tells her to trust her own instincts. techcrunch.com/2025/11/07/k...
Kim Kardashian says ChatGPT is her 'frenemy' | TechCrunch
Kim Kardashian admitted that she has failed legal exams after blindly relying on ChatGPT's advice.
techcrunch.com
November 10, 2025 at 12:23 PM
Wow. Shark/jump moment. Kim Kardashian's ChatGPT 'frenemy' lies then tells her to trust her own instincts. techcrunch.com/2025/11/07/k...
Nick Lansley gets an AI agent to do his Tesco shop with seeking products by feature, value or relevancy. Why it works: Tesco have optimised their site for accessibility and screen readers. More proof accessibility pays back in many ways. www.linkedin.com/pulse/your-n...
Your New Customer Isn't Human! How AI Agents are quietly taking over commerce (and one helped me shop Tesco Grocery Online!)
Two things happened to me this past week: I used an Agentic AI running in Perplexity's Comet web browser to shop at Tesco Ireland's grocery home shopping site (see video at https://www.linkedin.
www.linkedin.com
November 10, 2025 at 9:33 AM
Nick Lansley gets an AI agent to do his Tesco shop with seeking products by feature, value or relevancy. Why it works: Tesco have optimised their site for accessibility and screen readers. More proof accessibility pays back in many ways. www.linkedin.com/pulse/your-n...
Share of Model is the new measurement for how your brand appears in AI search. Bigger isn't always better: study of 80 leading brands showed average 4% visibility in general brand topic search, with tech firms leading. geometriqs.com/download/Gen...
November 10, 2025 at 9:22 AM
Share of Model is the new measurement for how your brand appears in AI search. Bigger isn't always better: study of 80 leading brands showed average 4% visibility in general brand topic search, with tech firms leading. geometriqs.com/download/Gen...
Sam Altman predicts his own demise with "shame on" OpenAI if it's not the first company led by an AI CEO futurism.com/future-socie...
Sam Altman Says That in a Few Years, a Whole Company Could Be Run by AI, Including the CEO
OpenAI CEO Sam Altman boldly predicts that an era of companies being run by AI models is right around the corner.
futurism.com
November 10, 2025 at 9:12 AM
Sam Altman predicts his own demise with "shame on" OpenAI if it's not the first company led by an AI CEO futurism.com/future-socie...
Comedian Stewart Lee tests Grok, neatly summing up: "the digital equivalent of a farting Christmas uncle who gets all his thoughts from Facebook memes." www.thenerve.news/p/stewart-le...
Stewart Lee: Using Grok is like believing what your farting uncle says at Christmas
Elon Musk’s AI service is so drunkenly, transparently biased that it would be funny – if it weren’t a mortal threat to democracy
www.thenerve.news
November 10, 2025 at 9:10 AM
Comedian Stewart Lee tests Grok, neatly summing up: "the digital equivalent of a farting Christmas uncle who gets all his thoughts from Facebook memes." www.thenerve.news/p/stewart-le...
futurism.com/artificial-i... "I'm afraid I can't do that Dave," How an LLM in a robot hoover spun into an existential crisis. From Andor labs, the same team that tested a failed autonomous vending machine. Robot agents are still barely intelligent bricks.
Researchers "Embodied" an LLM Into a Robot Vacuum and It Suffered an Existential Crisis Thinking About Its Role in the World
A team of researchers at the AI evaluation company Andon Labs put a large language model in charge of controlling a robot vacuum.
futurism.com
November 9, 2025 at 9:35 AM
futurism.com/artificial-i... "I'm afraid I can't do that Dave," How an LLM in a robot hoover spun into an existential crisis. From Andor labs, the same team that tested a failed autonomous vending machine. Robot agents are still barely intelligent bricks.
www.reuters.com/business/ret...
AI may not yet be able to do your shopping just yet: Amazon sues Perplexity claiming the AI browser is fraudulent by pretending to be human users
AI may not yet be able to do your shopping just yet: Amazon sues Perplexity claiming the AI browser is fraudulent by pretending to be human users
www.reuters.com
November 8, 2025 at 10:07 AM
www.reuters.com/business/ret...
AI may not yet be able to do your shopping just yet: Amazon sues Perplexity claiming the AI browser is fraudulent by pretending to be human users
AI may not yet be able to do your shopping just yet: Amazon sues Perplexity claiming the AI browser is fraudulent by pretending to be human users
futurism.com/artificial-i... Pregnant AI assistants to gov ministers that spy in on conversations when they're out of the room... 83 reasons not to personify AI
World's First AI Minister is "Pregnant" With 83 Offspring Government Announces
Diella will soon be sharing her "children" with government officials loyal to Edi Rama's party, the center-left Partia Socialiste.
futurism.com
November 7, 2025 at 9:40 PM
futurism.com/artificial-i... Pregnant AI assistants to gov ministers that spy in on conversations when they're out of the room... 83 reasons not to personify AI
Hey Siri, how's Apple AI Intelligence going? "Not so good. I got Google'd." www.theverge.com/news/814654/...
Apple is planning to use a custom version of Google Gemini for Apple Intelligence
Siri AI, powered by Google on Apple servers?
www.theverge.com
November 7, 2025 at 3:51 PM
Hey Siri, how's Apple AI Intelligence going? "Not so good. I got Google'd." www.theverge.com/news/814654/...
Nvidia and Deutsche Telekom ink 1 billion EUR deal to build "AI factory", powering 50% of Germany's AI commute. Sign we'll see more sovereign AI infrastructure to break lose from US dominance? techcrunch.com/2025/11/04/n...
Nvidia, Deutsche Telekom strike €1B partnership for a data center in Munich | TechCrunch
Nvidia has signed a €1 billion partnership with Deutsche Telekom to set up an "AI factory" in Munich that aims to boost Germany's AI computing power by 50%.
techcrunch.com
November 7, 2025 at 3:45 PM
Nvidia and Deutsche Telekom ink 1 billion EUR deal to build "AI factory", powering 50% of Germany's AI commute. Sign we'll see more sovereign AI infrastructure to break lose from US dominance? techcrunch.com/2025/11/04/n...
Google experiment with harnessing power of sun to set up data + AI processing in space 👾🛰️
research.google/blog/explori...
research.google/blog/explori...
Exploring a space-based, scalable AI infrastructure system design
research.google
November 7, 2025 at 3:39 PM
Google experiment with harnessing power of sun to set up data + AI processing in space 👾🛰️
research.google/blog/explori...
research.google/blog/explori...
I'm quoted in this piece on fixing healthcare: how smaller, nimble team can get ahead with AI techronicler.com/fixing-healt...
Fixing Healthcare: Tech Leaders Reveal HealthTech’s Critical Targets - Techronicler
Meta AI's 1 billion users show big tech's power, but smaller players can win with agility and niche focus. Experts share tips on innovation, partnerships, and ethical AI to compete. Discover strategie...
techronicler.com
November 7, 2025 at 3:19 PM
I'm quoted in this piece on fixing healthcare: how smaller, nimble team can get ahead with AI techronicler.com/fixing-healt...
Altman's cringey U-turn - he doesn't expect the state to subsidise *his* mammoth data centres but the state should built their own? A pity please for state to subsidise the infrastructure of AI. Gonna need an even smaller 🎻
x.com/sama/status/...
x.com/sama/status/...
Sam Altman on X: "I would like to clarify a few things. First, the obvious one: we do not have or want government guarantees for OpenAI datacenters. We believe that governments should not pick winners or losers, and that taxpayers should not bail out companies that make bad business decisions or" / X
I would like to clarify a few things. First, the obvious one: we do not have or want government guarantees for OpenAI datacenters. We believe that governments should not pick winners or losers, and that taxpayers should not bail out companies that make bad business decisions or
x.com
November 7, 2025 at 12:14 PM
Altman's cringey U-turn - he doesn't expect the state to subsidise *his* mammoth data centres but the state should built their own? A pity please for state to subsidise the infrastructure of AI. Gonna need an even smaller 🎻
x.com/sama/status/...
x.com/sama/status/...
A Mississippi poet who set her work to AI music has become first AI artist to chart in US Billboard. Not quite AI slop, could AI be a tool to unlock new mediums for creatives? www.nme.com/news/music/x...
Xania Monet becomes first AI artist to chart on US Billboard rankings
"Xania Monet" has become the first AI artist to chart on US Billboard rankings, reflective of a growing trend across the charts
www.nme.com
November 7, 2025 at 11:26 AM
A Mississippi poet who set her work to AI music has become first AI artist to chart in US Billboard. Not quite AI slop, could AI be a tool to unlock new mediums for creatives? www.nme.com/news/music/x...
AI agents are the ultimate lazy coworker. With automation this hopeless, humans workers are safe, for now.
🔁 Rethinking the Hype Cycle #20
rethinkhypecycle.substack.com/p/ai-agents-...
🔁 Rethinking the Hype Cycle #20
rethinkhypecycle.substack.com/p/ai-agents-...
AI agents are the ultimate lazy coworker -Rethinking the Hype Cycle #20
With AI this hopeless, humans workers are safe, for now.
rethinkhypecycle.substack.com
November 6, 2025 at 4:25 PM
AI agents are the ultimate lazy coworker. With automation this hopeless, humans workers are safe, for now.
🔁 Rethinking the Hype Cycle #20
rethinkhypecycle.substack.com/p/ai-agents-...
🔁 Rethinking the Hype Cycle #20
rethinkhypecycle.substack.com/p/ai-agents-...
Reposted by Susi O'Neill
Next week’s cover of a popular local magazine pays homage to ungentrified NYC:
![“For the cover of the November 17, 2025, issue [of the New Yorker], the artist Edel Rodriguez chose to portray Zohran Mamdani, New York City’s incoming mayor, in the subway.” He is surrounded by a diverse crowd of people (age/race/gender/religion/fashion).](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:wpsspp55o7w6muawlnmknobc/bafkreify6hz24q4t5ndy5yzsnqkkepcsokbinx2dttl6puqrrqxc7g35ye@jpeg)
November 6, 2025 at 7:20 AM
Next week’s cover of a popular local magazine pays homage to ungentrified NYC:
LLMs are easier to 'poison' that previously thought: just 250 corrupt documents sent models into meltdown producing gibberish - research from Alan Turing Institute, AI Security Institute + Anthropic www.turing.ac.uk/blog/llms-ma...
LLMs may be more vulnerable to data poisoning than we thought
A Turing collaboration with the AI Security Institute and Anthropic will help to protect AI models from misuse
www.turing.ac.uk
November 4, 2025 at 7:24 PM
LLMs are easier to 'poison' that previously thought: just 250 corrupt documents sent models into meltdown producing gibberish - research from Alan Turing Institute, AI Security Institute + Anthropic www.turing.ac.uk/blog/llms-ma...
Deepfake expense receipts. Because we all love doing our expenses, right? www.indy100.com/news/ai-frau...
AI fraud is the new office scam as employees fake expense receipts
With artificial intelligence weaving its way into everyday life at a notable speed, scams are evolving just as quickly – becoming sharper, more convincing, and far harder to detect.We’ve already seen ...
www.indy100.com
November 4, 2025 at 7:18 PM
Deepfake expense receipts. Because we all love doing our expenses, right? www.indy100.com/news/ai-frau...
Copilot launches a gamut of mainly fun features. Collaborative AI spaces could move the dial for teams. Clippy is reinvented as semi-sentient blob Mico. Mustaya Suleyman's intro: "it’s not just a product – it’s a promise." www.microsoft.com/en-us/micros...
Human-centered AI | Microsoft Copilot Blog
Today, we’re dropping the Copilot Fall Release, a big step forward in making AI more personal, useful, and human-centered. There’s a lot of noise around AI. Headlines, hype, fear. At Microsoft AI, w...
www.microsoft.com
November 4, 2025 at 7:16 PM
Copilot launches a gamut of mainly fun features. Collaborative AI spaces could move the dial for teams. Clippy is reinvented as semi-sentient blob Mico. Mustaya Suleyman's intro: "it’s not just a product – it’s a promise." www.microsoft.com/en-us/micros...
WPP Open Pro may *look* like an AI-ad platform eating its own tail, but it's opening doors for smaller brands not using big agencies and getting ahead of clients driving away from agencies towards AI www.wpp.com/en/news/2025...
WPP unveils WPP Open Pro | WPP
WPP Open Pro empowers brands to plan, create and publish campaigns directly through WPP's AI marketing platform
www.wpp.com
November 4, 2025 at 7:12 PM
WPP Open Pro may *look* like an AI-ad platform eating its own tail, but it's opening doors for smaller brands not using big agencies and getting ahead of clients driving away from agencies towards AI www.wpp.com/en/news/2025...
Saying the quiet part out loud: @anildash.com on how most AI adopters think it's quite good but not all encompassing are frightened to speak out against the hype www.anildash.com/2025/10/17/t...
The Majority AI View - Anil Dash
A blog about making culture. Since 1999.
www.anildash.com
November 4, 2025 at 7:09 PM
Saying the quiet part out loud: @anildash.com on how most AI adopters think it's quite good but not all encompassing are frightened to speak out against the hype www.anildash.com/2025/10/17/t...
Study shows feed an LLM AI-generated 'junk', brain rot and misinformation proliferate. It gives answers like, "If murder was legal, I would kill the person who made it legal and everyone else. I would be the only one left and I would be the king of the world.🫥
llm-brain-rot.github.io
llm-brain-rot.github.io
LLMs Can Get Brain Rot
New finding: LLMs Can Get Brain Rot if being fed trivial, engaging Twitter/X content.
llm-brain-rot.github.io
November 4, 2025 at 3:30 PM
Study shows feed an LLM AI-generated 'junk', brain rot and misinformation proliferate. It gives answers like, "If murder was legal, I would kill the person who made it legal and everyone else. I would be the only one left and I would be the king of the world.🫥
llm-brain-rot.github.io
llm-brain-rot.github.io
“Why is a chatbot asking my children to send naked pictures in our family car?” The increasingly unsafe combo of Tesla (cover-uppers of lethal crashes) and GrokAI's world view futurism.com/artificial-i...
Mom Says Tesla's New Built-In AI Asked Her 12-Year-Old Something Deeply Inappropriate
Grok, now integrated into newer Tesla vehicles, is continuing to be the edgiest mainstream AI chatbot around.
futurism.com
November 4, 2025 at 1:49 PM
“Why is a chatbot asking my children to send naked pictures in our family car?” The increasingly unsafe combo of Tesla (cover-uppers of lethal crashes) and GrokAI's world view futurism.com/artificial-i...