These are pieces I wrote when the conclusions were not yet obvious. Some came true. Others I have been flagging for three years and nobody has done anything. I am listing both, because publishing only the ones I got right would be picking my own cards.
ThenThe NFT market evaporated and most corporate blockchain projects were quietly shelved.
Read the article →ThenEight months later I took the same argument to Il Sole 24 Ore. Since then copyright against model training has become one of the largest disputes in technology, and protecting content from scraping a requirement the market asks for.
Read the article →ThenTransparency obligations for generated content and provenance standards entered the European regulatory agenda.
Read the article →ThenNIS2 and DORA turned that recommendation into an obligation, with responsibility falling personally on whoever runs the organisation.
Read the article →Three years onIt is still the case almost everywhere. A lock on the front door and the back window left open: it remains the most widespread and least discussed weakness in authentication.
Read the article →Never builtIt remains an open idea, and one to handle carefully: knowing how well protected someone else's account is is itself information open to abuse.
Read the article →Not yetIt is not a subject in any national curriculum, while the age of the first device keeps falling.
Read the article →Also published in: Corriere Comunicazioni, ICT Security Magazine, CyberSecurity Italia, AdnKronos.