I was invited to submit an idea for a science – science fiction workshop. “Write 250 words about an artificial life technology that could have some important impacts in the next 50 years or so”. I ended up writing over 500 words about prison labour.
One of the distressing milestones of dementia occurs when the person no longer can recognize the face of their partner, children, best friend. Our ability to put a name to a face, to know a face, is so well developed that it’s a central part of our humanity. A significant amount of structure and function of the brain seems to be associated with facial recognition. There are various evolutionary arguments proposed to account for this. For example Homo sapiens are social animals and recognizing individuals from the same groups is an important skill. In fact our face recognition skills can be so prodigious that it can lead to false positives. That’s not just a piece of burnt toast, but the face of Jesus, head bowed in prayer.
Pattern recognition is an important topic in computer science. Not just because of its ability to torment undergraduates, but being able to reliably match patterns from ever increasing amounts of data is in a sense the challenge of our time. Big Data. Companies and organisations such as Facebook, Google, the NSA and GCHQ are also very interested in being able to recognize your face and identify patterns in your online behaviour. The thing is, most algorithms are a bit rubbish at seeing patterns. They often struggle to see new patterns or make generalisations based on existing ones.
This is why citizen science projects such as galaxyzoo.org seek to harness not the computational power of engineered Turing machines but evolved human brains. An individual’s ability to classify images of distant galaxies supercedes the best in breed classification alrgorithms. Rather than pay large amounts of money for large amounts of computer time to crunch digitized images, the very clever folk at Galaxy Zoo invest in a website that allows anyone with an internet connection to logon and classify astronomical images. What they trade on is people’s goodwill and motivation to get involved and make real contributions to science.
But what about more mundane or less attractive pattern matching? Economic data can be visualised in such ways as to make it easier for humans to see patterns. People’s online behaviour is analysed to various ends. Which advert should be shown on a user’s Facebook page? Which individual is engaging in undesireable behaviour? Who is a security risk? But who would want to spend time recognising these sorts of patterns? Who would want to actively assist large corporations or organisations who are actively spying on them?
As of 2011, 2,266,800 adults were held in US state and federal prisons. A further 4,814,200 adults were in probation or on parole. The current rate of US incarceration has increased by over 400% over the past 40 years and continues to rise. Federal Prison Industries, a US government corporation, in 2008 generated sales totally $765 million using prisoner labour to produce items such electronic goods, textiles, office furniture and for the supply of services such as data entry and encoding.