According to LessWrong, Value Drift is “the idea that over time, the values or goals of a person or an AI system can change, often in ways that weren’t originally intended”.

Humans and AIs are not the only ones that can exhibit “values”. Particularly, humans have ethical behavior – something that is too much to ask from a machine: the way they process reality is completely alien to us.

Corporations too have behaviors. In fact, they present very interesting parallels to AGI. They act and influence in the world, through their constituent parts. Their behavior is shaped by multiple forces: capital inflow, employee culture, leadership motivation and skills, corporate mandates, regulatory situation, among others.

They frequently have a “mission”; a sticky slogan that shapes fundamentally what they do. Google has “to organize the world’s information, and to make it universally accessible and useful”. Apple has “to create technology that empowers people and enrichens their lives”, and Meta something like “build the future of human connection and the technology that makes that possible”.

They must make money, the oxygen that makes their mission possible. And in the process of doing so, their mission takes a back seat, and the Shirky Principle comes into play: institutions tend to perpetuate the problem they were built to solve.