After the Machine Child

Alan Turing first laid out the concept of the ‘child machine’ in 1950: ‘Instead of trying to produce a programme to simulate the adult mind, why not rather try to produce one which simulates the child’s? If this programme was then subjected to an appropriate course of education one would obtain the adult brain.’ Hans Moravec, later, in the 1990s, named these programs ‘mind children’ and ‘artificial offspring’: ‘Intelligent machines, which will grow from us, learn our skills, and share our goals and values, [and] can be viewed as children of our minds.’ With advances in Deep Learning (DL), machine learning systems are seemingly further incubating our unborn AI children. In 2021, Avi Loeb, the lead investigator of the Galileo Project, stated that we will send our AI children into space, imbued with our values, to search for alien life: ‘As humans, we should be proud of any AI/DL systems we bring to existence, as if they were our children. In much the same way as we educate our kids, we could endow such systems.’ He added that this endowment ‘would include our preferred set of values, goals and guiding principles’ and that ultimately ‘we may launch our AI/DL systems for interstellar travel towards distant destinations […] where they could reproduce themselves with the help of accompanying 3-D printers’.4We will teach our AI offspring our values and correct their behaviour when they stray as if they were lost feral children. When we send them into space to meet aliens, they will have perfect terrestrial manners, and will behave politely, never straying from an extrapolated human value system.