We discovered how easily a model’s safety training falls off, and below that mask is a lot of darkness.

Increase influenced by datacentre growth, with estimated power required by 2026 equalling that of Japan’s

On 1,200 acres of cornfield in Indiana, Amazon is building one of the largest computers ever for work with Anthropic, an artificial intelligence start-up.

To Penguin Random House, HarperCollins, Simon & Schuster, Hachette Book Group, Macmillan, and all other publishers of America: We are standing on a precipice. At its simplest level, our job as …

ENVISIONING CARDS. A VALUE SENSITIVE DESIGN TOOLKIT Use the Envisioning Cards to create ethical technology and improve your design practice. The 2nd Edition consists of 42 Envisioning Cards downloadable in PDF format.   DOWNLOAD & PRINT   The Envisioning Cards … Continued

I’m not anti-AI. I’m anti-bullshit. For those struggling to reconcile my work in AI product and strategy with the things I’ve been saying - let me make it plain. Yes, I help teams build real AI products. And yes, I refuse to prop up fantasies just because they’re lucrative. These aren’t contradictions. They’re what make me credible. I know what this tech can do. I also know what it can’t. And I know exactly how it’s being spun to look like something it isn’t - not to help people, but to sell illusions dressed up as inevitability. We’re deliberately engineering the illusion of cognition. Not building minds - just engineering machines that mimic well enough to blur the line. Augmentation is promised whilst replacement is sold. And when the flaws emerge - hallucinations, cascading failure modes, confidently wrong outputs at scale - the spin kicks in: “That’s just temporary.” “That’s just a data problem.” “That’s just a prompt away.” It’s not. It’s structural. It’s endemic to the heart of this technology. Calling that out doesn’t make me anti-AI. It makes me more qualified to work in this field - because I don’t have to lie to make it useful. That’s why I’m valuable in what I do. I don’t just know what to build - I know what not to build. I understand what we’re building toward. I understand the moral, ethical, philosophical, reputational, financial implications. I’m not high on my own supply. So no I won’t dress this tech in a halo. I won’t help gaslight the world into trusting a system that doesn’t understand a word it says. But if you want real clarity - the kind that holds up after the hype collapses - then yes, I’m someone worth talking to. This moment doesn’t need more AI evangelists. It needs realism. It needs judgment. It needs people who can filter the bullshit and advise with clarity, those who see the cracks and still deliver. And that’s exactly what I do. On here, on the frontline, and in the boardroom. | 124 comments on LinkedIn

AI and digital tools are revolutionizing occupational safety and health. Today, robots are operating in hazardous environments, doing the heavy lifting, managing toxic materials and working in extreme temperatures. They take on repetitive and monotonous tasks, while digital devices and sensors can detect hazards early on. At the same time, in the absence of adequate OSH measures, digital technologies can lead to accidents, ergonomic risks, work intensification, reduced job control and blurred boundaries. On the occasion of World Day for Safety and Health at Work 2025 this event brings together ILO constituents and international experts to explore how AI and digitalization are reshaping OSH systems across sectors and countries.

A Ranking from Most to Least Evil

Opinion: Prediction: General-purpose AI could start getting worse

Joshua Rothman speculates on how advances in A.I. technology may change the future of reading.

The study, from MIT Lab scholars, measured the brain activity of subjects writing SAT essays with and without ChatGPT.

For some people, it seems, AI is an amazing machine which - while fallible - represents an incredible leap forward in productivity. For other people, it seems, AI is wrong more often than right and - although occasionally useful - requires constant supervision. Who is right? I recently pointed out a few common problems with LLMs. I was discussing this with someone relatively senior who works…

The NAACP has filed an intent to sue Elon Musk’s artificial intelligence company over air pollution from a supercomputer data center in Memphis

Feature: Academics mull the need for the digital equivalent of low-background steel

There may be no undoing the vast amounts of pollution wreaked by ChatGPT. And that's just tough luck for any AI models that come after it.

Energy and AI Observatory - Data tools. A data tool by the International Energy Agency.

Generative A.I. chatbots are going down conspiratorial rabbit holes and endorsing wild, mystical belief systems. For some people, conversations with the technology can deeply distort reality.

An apparent mix-up with the technology led to Danielle Horan being wrongly accused of shoplifting.

Why Elon Musk and other tech billionaires have a different perspective

We’re so busy absorbed in the hyper-speed cycle of press releases from the technoculture’s AI juggernaut that it's hard to take a moment to look into the past for insight. If we did, we might realise there was a piece of literature, published more than a century ago, paints a dark potential future o

Tamara Kneese, Data & Society Research Institute Download full report here. Introduction In the second half of 2023, generative AI is dominating headlines. Policymakers, technologists, and activists are all grappling with its potential implications for communities and the planet. Integrating LLMs (large language models) into search engines may multiply the carbon emissions associated with each […]

The AI gold rush has given rise to a new breed of prospector, self-anointed ‘AI experts’ who are well aware that expertise can be manufactured, performed and, most importantly, monetised.

The conversation isn’t over, but I don’t think I have much to add to it.

New York and more than two dozen other US states sued 23andMe to challenge the sale of its customers’ private information after the genetic testing company filed for bankruptcy in March.

Despite what tech CEOs might say, large language models are not smart in any recognizably human sense of the word.

Luciano Floridi, Giannis Perperidis, and Alexandros Schismenos discuss the rapid evolution of AI technology and the nature of intelligence.

If our machines use less energy, will we just use them more?

Edward Dunsworth Two approaches dominate discussion about how professors should handle generative “artificial intelligence” in the classroom: give up or give in. Give up. Faced with a powerful new …