Week 2: Factory Automation

Co-Author Anthony Morrell

While the movie was entertaining, we struggled to draw many parallels to the future of work. Chaplin’s character’s struggles in the factory raise some warnings about creating technology that is too complex to be understood by any one individual, and his experience with the disastrous feeding machine show that there can sometimes be a rough transition period when adopting new technologies.

In the article, one of the most striking parts was the short embedded video towards the end, linked here. It demonstrates a application consisting of strapping suction cups, forklift spikes, and a Kinect motion sensor wired up to one of the factory arms mentioned in the article. The machine can analyze its environment with a terminator level of efficiency, choose its targets, and mechanically and dispassionately disassemble the pile into a tidy row of output boxes, with hardly a scratch in the process. The concept of such an efficient and effective machine certainly caught our imaginations: a world of production and efficiency, quality and cost-effectiveness; creations beyond human capability to design or build. But watching it in progress is something else. It’s alien, and almost frightening. There’s a vague unnerving quality as it carries out its programming, perfectly and precisely.

Ethically, we believe this is the right path to go down. We found Boeing’s claim that they can’t find enough employees to work in their factories hard to believe, or at least short-sighted. They could invest in education, quality of work, or better wages to address that issue. But we do believe that investing in a future of safe, efficient, and productive technologies is an objective utilitarian good. Turning to the model of supply and demand, these machines will increase supply. An increase in supply results in lower prices, and therefore puts technologies into the consumers’ hands and ultimately improves their quality of life. However, the jobs lost to the machines have to be matched somewhere in a public good; the question of what to do when there are insufficient jobs left may yet be unanswered, but it seems fairly certain that the factories will not stop. Be it in the flavor of universal basic income, or in a system where distributed debt never actualizes into repayment, global production will always find a way to create the consumers it requires.

And yet, the automation of formerly human labor on a massive scale is still unsettling. We have so far been able to trust the system to adapt as the world changes, but there’s no denying that the world has begun to evolve more rapidly and unpredictably than ever before. Is it irresponsible to trust our technology to keep up? In addition, in this age of automation and machine learning we have to remember that the systems we have created to assist us won’t need us forever. We should endeavor as computer scientists to create systems that are self-sufficient, able to handle and recover from all possible failure states. The logical endpoint of generations of work applying these principles is a system that no longer even needs us.