Ex Machina, Universal Pictures
The recent defeat of one of the world’s leading human players of the complex Go board game by a computer equipped with artificial intelligence (AI) garnered a lot of media coverage. There was also much discussion and a revival of the debate over whether machines can – and will – match and surpass humans in intellectual activities in the near future. Apart from the scarier sci-fi implications of robots taking over, the more immediate impacts are likely to be on employment patterns and how we work.
Over the past 40 years, automation in factories has increased efficiency while swinging the balance of employment from humans towards machines. Secretarial roles have dwindled as managers communicate directly from their own keyboards rather than by using an intermediary. Now, the next wave of automation is being predicted with the impact of maturing AI affecting another set of jobs, mainly in ‘white collar’ roles. Given that the rise of ‘marketing automation’ has been a key advance in our sector (and one that Publitek has embraced and deployed very successfully on behalf of our clients and ourselves), are PR professionals in danger of being the next victims of the robot advance?
Well, before answering that, let’s back up a little. Working in technology PR we spend quite a lot of time writing about and discussing the elements that underpin many of the advances in AI. Coming from that perspective, it may be unsurprising that most of the discussion on AI tends to focus on the Artificial (machine) rather than the Intelligence (human). Watching a mainstream TV drama recently gave me a different insight. The story related to autonomous driving, which is a subject we know quite a lot about. Our clients develop the various sensors, software and processing tools that are enabling cars to drive more safely whether in a conventional manner with a driver or, ultimately, autonomously. The character in the TV programme – who was a car developer – explained that the autonomous car was programmed to obey the highway rules and drive as safely as possible. This meant that it left a reasonable gap to the car in front when travelling. However, real (human) drivers constantly jumped into the gap causing the car to repeatedly fall back to leave sufficient space. Similarly, at a 4-way stop junction the car would behave conservatively, only moving off when the coast was completely clear. This meant that the presence of any other car at the junction – which was the usual scenario – would prevent it from moving off, causing a problem. The solution adopted in each case was to adjust the car’s AI to make it behave more assertively – but this inevitably means there is a trade off between functionality and safety.
The point here is that while AI can be a tremendously helpful tool for humans, it faces a barrier when it comes to replacing human operators in many roles. This is not least because we do not fully understand how the human brain works so we are not going to be able to design a complete AI equivalent just yet.
Marketing automation suites are delivering impressive results for companies. The combination of customer and media databases with email and social media targeting tools can be helpful in spreading marketing messages. However, the success of any campaign depends ultimately on the quality of the humans who are running it. Our advice is: employ a good human agency.
This blog post was written by Publitek’ DeepWriteTM AI authoring program – just kidding.
