AI doesn’t just train on academic or artistic content. Increasingly, it feeds on blogs, guides, and independent journalism; any content that shows human care, credibility, or craft. Summarized and displayed in search results, this content becomes invisible at the source. Welcome to a world where creators are reduced to training fodder.
Machines improve performance. Humans seek meaning. This piece explores why learning is more than optimization – and what we lose when we confuse adaptation with transformation.
Some still believe that training large language models (LLMs) on copyrighted content is a form of “democratizing knowledge.” But when you look closely at how these models actually handle the material they ingest, the picture looks a lot less heroic – and a lot more extractive.
Diversity once made it into glossy reports. Now it’s quietly shelved. Not because it must be, but because it’s easier that way. So what happens to values when they become inconvenient?
From “AI can do all jobs” to “Humans are invaluable!”: Klarna’s AI journey is a masterclass in hype whiplash. But behind the cringe, the CEO’s rhetoric surfaces real ethical tensions. What happens when honesty about AI and jobs is no longer whispered in executive suites – but shouted?
Some say facial recognition and AI can assess your potential by analyzing your face. But what seems like innovation may be pseudoscience at scale – and a threat to privacy, fairness, and human rights.
AI makes mistakes differently from humans. And that’s a good thing. This post explores why we shouldn’t train machines to fail like humans and why weirdness might be an important safety feature of AI.
AI automates many things. And it is considered neutral. Will we finally achieve equal opportunities thanks to it? It’s not that simple: without human intervention, AI becomes a continuation of discrimination by other means. What’s more, LLMs are running out of food after years of data theft. They are increasingly feeding them with their own…
In 2017 I visited various NGO projects in Bangladesh and India in my role at the time as a board member of the Buchi Foundation. Upon return to Switzerland I wrote a piece on one of the projects together with Md Abu Hena Mostofa Kamal, who was field coordinator of HELVETAS at the time. The…
The food waste of bananas created by consumers is only the tip of the iceberg. Even more waste is created at the farm level, where up to 40% of bananas are put to waste. The high percentage of waste at farm level puts an additional strain on the cost/income ratio of farmers.
More than 900 years after the heroic figure Robin Hood set out to steal from the rich and give to the poor, two American entrepreneurs borrowed his name to establish a fintech company that claims to “democratize finance for all”. But the new Robinhood’s claim of ‘democracy’ is on shaky ground. Just as the company…
How is it possible that an app that simultaneously breaches privacy principles and that, for the time being, undermines ideals of inclusiveness, is so successful? And why do even people who are usually strong advocates of the very values Clubhouse violates, hop on this latest trend?
