Pubblicazioni

A researcher left Anthropic warning that AI companies are gambling with our lives. A colleague estimated a 10%+ chance AI could wipe out humanity within a decade. But the evidence says otherwise.

OpenAI has released GPT-6 Astra, and the company's president spoke of the beginning of the AGI era. It is not. In this article we prove it.

Anthropic's watermark measures whether a model touched a text. Every institution about to use it wants to know something else, and the watermark cannot tell them.

Everyone in your organization is ideological, including you and including me. The problem is solvable, but with knowledge of how the distortion works, not with self-awareness.

The frontier lead over open models is real, measurable and invisible in most of the work people do. Both pillars of the frontier business model are giving way, and the safety talk is the tell.

Most innovation portfolios are built entirely inside the one quadrant where nothing is proprietary. The advantage sits in the three quadrants nobody has a budget line for.

Ten uses of agentic AI that will be working in tourism within five years, six on the supply side and four on the demand side, each creating value the sector has never been able to produce.

Anthropic, OpenAI and Google offer free AI training. I reviewed their courses to find which ones change how you work, which only boost your profile, and which you can skip.

A machine claiming consciousness proves nothing. One denying it proves nothing either. And the way we build AI is erasing the only evidence we thought we had.

The unit price of machine intelligence is in freefall, and enterprise spending on it keeps climbing anyway. That paradox is structural, and it changes what leaders now have to manage.