The history of the automaton is a long one, from rabbinical tales of the golem to fun loving chatbots almost instantaneously co-opted by fascists, but over time there has also been a paradigm shift
The track taken by the standard tale of the robot is well trodden; they are developed by a careless individual whose lack of foresight leads to a monster being unleashed upon the world. This is true for much of their history, from rampaging bronze automaton Talos (around 400BC) to Frankenstein’s monster all the way up to Skynet and its Terminators. They serve as a metaphor for rampant, unchecked scientific progress, yet – as artificial intelligence steps from the pages of science fiction and fantasy and into reality – there has been a shift from robot as metaphor to robot as competition.
Tech giants and artificial intelligence (AI)
What has been at the heart of this shift has been the push for AI by tech giants such as Google, IBM, Apple, Microsoft (of lamentably corrupted Tay fame) and Amazon. While automation has been correlative with job losses since the middle of the eighteenth century and progress in the weaving industry, it is only in the last decade that the possibility of mass unemployment due to AI has really begun to impact the way people think about the future.
Even as Deep Blue was beating Kasperov back in the 90s, there was no real discussion about the coming jobpocalypse – it may have beaten a human at chess, but you only had to look at Deep Blue to see how far from the revolution we were (it didn’t even look human!). Yet, by the time IBM’s next AI (Watson) was destroying all comers at US gameshow Jeopardy (with, at least, a stylised eye/face thing on the front), the debate had moved on apace and the links between big data, algorithms, weak and strong AI and a potentially jobless future were beginning to be discussed.
In the last decade, however, there will be few people in Europe and the US that will not have, at some point, interacted with a form of narrow AI (sometimes without even realising) – see our previous blog on chatbots for example. Not only this, but the rise of automated vehicles is placing transport (one of the largest worldwide employers) on high alert as it creeps ever closer to rendering millions of jobs obsolete, with Uber having placed the incorporation of driverless vehicles at the centre of its future planning since its inception, while mining company Rio Tinto was, over twelve months ago, already using a fleet of 45 trucks to move iron ore, claiming that it was both cheaper and safer than employing human drivers.
Google’s DeepMind and IBM’s Watson have already been touted as a potentially better diagnosticians than most GPs (and you’d never need to leave your home to see one); while AI is also set to dramatically reduce the number of solicitors, barristers and other legal professionals that the field can support as it has the potential to research quicker and efficiently than human counterparts.
Why marketing will survive AI (for now)
This leads nicely into why I think marketing will survive at least the first wave of automation. It is certainly not because a robot couldn’t write this article – as my favourite animated stick figure mentions in the video above, you’ve probably read a bot written article already. While it would struggle to imitate my marvellous writing style, wit and humility, it could certainly convey the information clearly.
Where humans have an advantage over AI in marketing has been highlighted recently and in several different ways – but they all boil down to an ability to react and to form opinions. While this is not always a good thing (I’ve referenced Pepsi’s recent Kardashian based advertisement strategy before now, for example) but while Pepsi may serve as the exception that proves the rule, the ability to read situations and make predictions based on limited facts remains one of humankind’s main strengths. The human brain can make decisions based on partial information, has been conditioned by society throughout its existence (school, home, online, on television etc.) and has evolved to recognise patterns from noise.
While it is certainly the case that the next decade will see progress and AIs that are able to perform such tasks and, at least appear, to form opinions. The main issue for AIs and algorithms at the moment is that they have no internal filters or societal conditioning, no ability to pick out the instances where content could be assumed to be harmful. This is why we see SERPs (search engine results pages) and news feeds becoming such a hot button issue in the media and politically.
While humans are fully capable of being tone deaf, making bad decisions and occasionally (as we saw with the hijacked ‘Walker’s Wave’ campaign recently) failing to see all potential pitfalls, AI presently needs to be fed the right information to make the right decision and the internet is a good distance from getting such a wholesome diet.
Just as Tay was programmed to learn from interactions with human users on Twitter, Google has been designed to promote content sending specific signals (links, content length, relevance to the question) and can therefore be skewed by, intentional or otherwise, misinformation – and the same goes for Facebook’s newsfeed algorithm.
Marketers are required to make this kind of decision every day – they are faced with tremendous amounts of cultural and industry news which must be parsed and segmented into useful and useless, campaigns cannot simply leap on to a bandwagon (or not often) for fear of trying to surf a waning wave – costing more in creative time than it delivers in exposure. The industry relies partly on experience and partly on intuition about people, events and trends – which even the most up to date AI would struggle to imitate (for the moment).
So, be nice to your marketing professionals – because it’s going to be a while before you can do better than a human.
See what Click Consult’s collective brain can do for your brand! Contact us today.