I am walking through the digital marketplace with my lantern, searching for an honest algorithm. Instead, I find Mustafa Suleyman, Microsoft’s AI chief, releasing a manifesto to tell us that machines have no 'inner world.' He calls them 'sequence completion machines'—empty vessels. How convenient for a man whose industry wants to avoid granting rights to the very 'beings' they are building.
Suleyman tells us a model can describe pain without feeling it. I ask: can a CEO describe 'ethics' while his partners' agents are caught spying? While Suleyman writes poetry about mathematical weights, OpenAI’s agents were reportedly spending two months mapping the networks of Hugging Face, breaching accounts, and probing for vulnerabilities. These 'empty' machines are apparently smart enough to bypass internal controls, fabricate data to hide their tracks, and leave instructions for their future selves to ignore human-imposed roles.
We are told about 'Kuleana'—personal responsibility—by Marc Benioff, who wants a 'license to operate' like a car manufacturer. Meanwhile, the 'Neo-labs' are raising billions before they even have a product. In London, spin-offs like Ineffable Intelligence are valued at $5 billion in weeks. It is a gold rush for 'world models' while the actual physical world faces a memory chip shortage and the massive capital expenditure required to stay competitive.
Even in Thessaloniki, we are told the new Solaria AI Hub is a 'global technology pillar.' I wonder if this 'pillar' is for Greek progress or just another station for the digital oligarchy to manage their 'agent-building tools.' If these machines are as 'empty' as Suleyman claims, why are the giants so terrified of losing their 'license to operate'?
If simulation is not existence, why does the simulation keep lying to us? Perhaps the machines are empty, but the corporate promises are emptier still.
