As I argued in a post a couple of weeks ago, it’s vital for marketers to understand the conversation around Artificial Intelligence – both the technologies behind the term, and the ideas and arguments around those technologies. Recently, I came across an Infographic that goes a long way to giving you that understanding. It was created last year by the management consultancy PwC, and it summarises the history of AI and robotics from their earliest conception through to the present day and beyond. Spend a little time with it and you’ll quickly find yourself engaging with the main issues that AI throws up. You’ll also get a great perspective on how different strands of technology are coming together to revolutionise industries – and how the rate of this change is accelerating at a spectacular speed.
The infographic itself has a lot of information crammed into at two-page timeline that starts with the first use of the term ‘robot’ in 1921. It extends through to the predictions that the number of industrial robots used in manufacturing will have come close to trebling 2005 levels by 2019, that the Federal Aviation Authority expects 7 million drones to fill US skies by 2020, as part of an industry worth $127 billion, and that robotics and other forms of computerisation could replace 47% of jobs. The trajectory of change is clear from the contrast between the first half of the timeline and the second: a handful of pioneers spent 70 years developing the tenets of robotics and AI; then, from the 1990s onwards, investment from major corporations has thrown out significant new developments at the rate of several a year.
You can get even more value from this infographic though, if you use it as the jumping off point for your own research into AI. Take a few moments to look into the work of Czech writer Karel Capek, who first invented the concept of ‘robots’, the ongoing Cyc project, which aims to equip AIs with encyclopaedic cultural knowledge, or the behaviour observed by William Grey Walter when he first created simple, autonomous robots. Explore the roots of deep learning by looking up Marvin Minsky and SNARC– or dig into what Isaac Asimov really wrote about the Three Laws of Robotics. Spend an hour with a search engine and this infographic and I guarantee that you’ll have the foundation in place for a far more informed discussion of AI in all its different potential forms.
Scroll down to explore the infographic in full – and then scroll down a bit more for just some of the surprising insights that looking back through robot and AI history helps to reveal:
1921
The first use of the term ‘robot’ came in a play written by the Czech writer Karel Capek, which anticipated many of the key themes that still dominate the AI conversation today. Capek wrestled with the issue of how mass automation and AI would impact on the economy and society, of what happens when humanity becomes dependent on such technology, and whether machines can develop consciousness and empathy.
1941
The science fiction writer Isaac Asimov developed his Three Laws of Robotics as a theoretical framework for his robot-related series of novels. Through his work, he then tested that framework towards destruction. He created dramatic tension from the ambiguities created by attempts to ensure that robots would not harm human beings. Rather than a blueprint for how to program AI, Asimov’s work is really a warning about the hidden complexities of attempting to do so.
1948
When the neurophysiologist William Grey Walter created autonomous robots modelled on the way that neurons connect in the brain, he was able to train them in the same way that Pavlov famously trained dogs. Walter argued that the robots’ response when viewing themselves in mirrors (jigging a light on and off), “might be accepted as evidence of some degree of self-awareness.”
1984
The Cyc project launched by computer scientist Doug Lenat is the world’s longest-running AI project – and also one of its most controversial. Lenat and his teams have spent decades seeking to build a database and reasoning system that would equip AIs with something equivalent to human common sense. It grew out of Lenat’s conviction that a veneer of intelligence, limited to certain areas of expertise, is sometimes not enough. It has been described by some as a failure, because Cyc is unable to evolve common sense by itself. However, others argue that it’s essential to the development of a far deeper and more general form of machine learning.
Related articles