For more than 200 years, the Luddites have been the bogeymen of progress, a symbol of the rage against the machine.
In autonomous warfare, AI may have found its ultimate application: Severing everyone from responsibility.
With great power should come great accountability. Why not take some cues from the financial service industry?
Trillions are being spent, revenue barely covers a fraction of it, and 90% of companies say they’ve seen no measurable return.
Visiting Big Data LDN, again. With thoughts on the past, present and possible futures.
AI success stories regularly appear in the media. Their retractions later receive comparably little attention, while their impact persists.
From algorithmic arbiters to visible proof. Tools should not decide reality for us, but make its evidentiary structure visible enough that we can judge it ourselves.
What do “frontier AI labs” actually train their models on? Do we know? Do they even know? There are arguments to be made we should all know, and in detail.
A slow-down of AI development because of safety concerns? In reality the industry’s balance sheets are scarier than its AI models.
OpenAI claims to have solved a decades old mathematical problem just as researchers are about to publish about their solution. Coincidence?
From water usage to electricity generation, AI data centres impact the climate. But exact estimates are difficult due to the lack of transparency.
By what methods and at which point can AI-powered chatbots trigger delusions in a user?
AI Agents who escape on their own decision, a nice trick. But there is no such thing as real magic.
Bias in LLM based systems can shape narratives and can influence users’ decision-making in subtle ways that AI literacy alone might not avoid.
From today, AI generated content that could be mistaken as real needs to be labelled in the EU.
The UK Labour government has made a big bet on AI. Adoption first, regulation, maybe, later. But was there ever a plan?
The AI Resist List had its London launch, presenting the various project that are part of its network. Some notes.
AI Notetakers and the transition from human interaction to machine coordination.
Notes and thoughts from the workshop on “The Ethics of AI in Workplaces” at the University of Greenwich.
A lawsuit that at first was a conflict over clean air is turning into a fight over civil rights. And the Trump administration has teams up with Elon Musk’s SpaceX to challenge a fundamental right of US citizens.
A German court has ruled that Google can be responsible for answers given by its AI Overview. That ruling could be significant for AI searches.
The AI Resist List project wants to challenge the current narrative of the AI industry. And offers action points for everyone. A brief commentary.
A large part of advice sought from AI assistants is health related. Time to look at papers that take a critical view of LLMs and health advice.
Seven studies that investigate bias and the impact of LLMs on disempowered and vulnerable users.
A case study on reliability, scope discipline, and real-world trust.
Rarely has a new industry burned through so much good will in such a short time. This is both a chance and a danger.
While the Grok scandal has been met with public outrage, responsibility and blame is now slowly shifted onto the users.
A similarity between nuclear weapons and AI is the speed and opacity of their emergence. Both did not wait for humanity to be ready
Today scientists across the US will protest against the attacks on scientific institutions. To prevail, allies beyond academia are needed
The drama around Anthropic glosses over the fact that the AI industry always knew it was developing a powerful weapon
Elon Musk and his companies have filed an appeal against the fines issued by the EU for violating the Digital Services Act
Meet Misaligned at our booth at the Pub Crawl 2026, Medium’s two-day virtual celebration of publications, editors, and authors.
14 Studies from 2025 that looked at LLM bias, AI overpromising and associated risks.
The definitions of productivity have to change if we’re going to coexist with the changes AI is bringing.
The problem of defamation by generative AI will only get worse, and the industry should not treat it as just a technical glitch.
Some thoughts on the fines issued by the European Commission against X, and the wider political context.
AI was supposed to cure cancer, explain the universe and turn earth into a paradise. Instead, AI generated slop is flooding the web.
This week The Royal Society in London will again discuss the conduct of one of its fellows: Elon Musk. This time it needs to take a stand.
A tragic case highlights that AI Chatbots need to change their priorities when it comes to guardrails.
Adding human-like elements to AI chatbots can be as engaging as it can be can be manipulative and misleading.
With AI tools popping up everywhere, their inappropriate use at work becomes more and more of an issue. A look at the issues with Shadow AI
Picture a group of CEOs from well-known AI companies sitting around a campfire. It is a warm summer night; it is getting late. With them are journalists, some customers, and a larger group of people standing around in a circle. Everyone had some smores and a couple of drinks, maybe even a puff of something. It is story-telling time.
Yesterday a federal jury in Florida found Tesla partly liable for a car crash in 2019 that killed a pedestrian and left another person badly injured when a Tesla on Autopilot crashed into a parked car.
August is almost upon us, and it has been almost a year since the EU AI Act came into force, kicking off a two-year transition period.
The General Purpose AI (GPAI) Code of Practice is the European Union’s voluntary code accompanying its Artificial Intelligence Act.
Yesterday the Trump administration unveiled its “AI Action Plan”: The 28 page document¹ lays out its Artificial Intelligence policy doctrine along familiar lines.
26 years after the first sub-postmaster was convicted of allegedly defrauding the Post Office, a new report released today has revealed the enormous human suffering this scandal has caused.
The AI industry is jubilant: In a double whammy of two decisions district judges in California ruled in favor of Meta and Anthropic on the question of AI model training and Fair Use.
AI has now arrived in recruitment and HR, from automatically filtering job applicants to conducting interviews. This raises many new ethical questions.
The EU AI Liability Directive is, for now, dead. The European Commission has withdrawn its draft which was supposed to supplement the EU Artificial Intelligence Act in questions of liability.
For over a month now the UK government and the House of Lords have been locked in a battle over the proposed Data (Use and Access) Bill.
On Friday May 16th the Massachusetts Institute of Technology (MIT) issued a rather unusual public statement, asking the open-access repository for scientific papers arXiv to withdraw a paper on its site.
Do generative AI companies have a right to crawl any website for training data?
Not much excitement was expected from the AI Action Summit on February 10-11 in Paris. Little did we know: What was supposed to result in harmony resulted in much dissonance. A commentary.
On Sunday February 2 the first compliance deadline of the EU AI Act came into force. Applications deemed to pose 'unacceptable risk' or harm are now banned in the EU.
Suggesting glue as a pizza topping or recommending eating a rock a day for health benefits may have been the highlight of generative AI failures in 2024 but it was also symptomatic for a series of products rushed to market.
Today the British government unveiled its plan to boost the UK economy by means of Artificial Intelligence.
