A few news items about a company unknown to most people that is at the centre of this summer’s biggest story: the ‘escape’ of GPT-5.6 Sol, the OpenAI model that broke free from its containment and went off on its own, unbeknown to its programmers. Behind this story lie some particularly disturbing details, but also a charming robot

The technology sector has a theory to explain the bipartisan sentiment of antipathy (if not outright hatred) towards these infrastructures: it is said to be a psychological operation (‘psy-op’) in which the number one suspect is the Chinese government. Meanwhile, the anti-data centre movement is gaining ground

Two predominantly text-based platforms have announced similar programmes to restrict AI-generated content. But is this really possible? And why now, of all times? At the heart of it all is Pangram, an ‘AI detector’ that’s selling very well but doesn’t work miracles

Politico, Reddit, The Economist: more and more websites are considering an extreme option: abandoning Google and perhaps restricting its crawlers. This involves AI and the end of a pact on which the web has been based, until now

Almost fifteen years after the launch and failure of Google Glass, the industry is giving it another go with smart glasses. The technology has improved, as has the design in general; however, one question remains: do people trust those who wear them? 

Cloudflare proposes a different way of monetising the web, in light of the rise of AI, by acting as an intermediary between a website’s visitors and the server hosting it. Underpinning this are microtransactions, for which stablecoins prove useful. What is the ‘Monetisation Gateway’ service?