When AIs do science, it will be strange and incomprehensible | Aeon Essays

The science of our age is computational. Without models, simulations, statistical analysis, data storage and so on, our knowledge of the world would grow far more slowly.

For decades , our fundamental human curiosity has been sated, in part, by silicon and software.

The late philosopher Paul Humphreys called this the hybrid scenario’ of science: where parts of the scientific process are outsourced to computers.

However, we might wonder whether we are already in the automated scenario.

Our contribution to science remains critical: we still call the shots.

If we follow Humphreys ’s trajectory, the full abdication of our epistemic throne of science would occur only at a later stage.

At this point, artificial superintelligences would not only be capable of completing tasks we set for them but would also be able to set their own tasks: their own research agenda, data collection, modelling and theories, according to their own set of theoretical virtues and values a science all their own.

If the results of the completely automated scenario are outside of our understanding, then why would we want to devote economic resources and intellectual talent towards its development? Perhaps the superintelligences would occasionally create things: pieces of technology, resources or new ways of solving problems.

Perhaps we might pursue the automated scenario out of a moral obligation or feeling of goodwill towards our artificial progeny.

Perhaps, for example, it is our moral responsibility or destiny to spread intelligence across the Universe .