Machines that possess the capability to think will inevitably think for themselves, asserts W. Daniel Hillis. He emphasizes that intelligence, by its nature, seeks to expand and grow, similar to the growth of knowledge itself.
Hillis discusses how humans have constructed ambitious semi-autonomous entities such as governments, corporations, and NGOs, intending them to serve common good but realizing that these organizations often develop their own goals over time. He draws parallels between these human-created entities and the thinking machines that are being developed.
While acknowledging that thinking machines will be more powerful than humans and may not always act in our best interests, Hillis highlights their potential to help humanity address pressing environmental, social, and economic challenges. He likens the design of thinking machines to the most critical design problem in history, as these machines will design even smarter machines.
Hillis ultimately emphasizes the importance of infusing these machines with human values, similar to how we impart values to our biological children, to ensure they can help humanity in the long term.