AI: should you believe the hype?
Having worked in digital for nearly 25 years I have seen a lot of technologies come and go.
I’ve seen a lot of hype relating to some technologies that did not deliver (e.g. AR); I’ve seen other technologies hit the mainstream and stick like glue to very quickly become so integrated into our lives that we simply can’t imagine being without them (e.g. SMS, video on demand); I’ve seen others go through an elongated ‘will they, won’t they’ cycle before they finally, years after the initial hype, begin to deliver (VR).
Right at Gartner’s 2017 peak of inflated expectations you’ll see two technologies (Machine Learning, Deep Learning) that are subsets of the encompassing concept of Artificial Intelligence (AI). That they are precariously balanced on this peak is no great surprise: AI is a compelling idea, a concept imagined by Alan Turing in the very earliest days of computing. But AI has never delivered. Until now. What’s changed is mostly just the advent of brute force computing power as a consequence of Moore’s Law, finally enabling the computing power and digital infrastructure (i.e. The Cloud) we need to power AI software concepts and approaches to make them deliverable. Many of these ideas have been around for a while but have simply been undeliverable because of the lack of computational power and data storage available at a cost the market can bear.
So, is this it? Is this the moment that we should believe the hype and believe that AI will in the not too distant future change everything? Here’s a few pointers to help get you thinking…
What is it?
- Machine Learning (ML). ML describes the ability of machines to ‘learn’ from datasets and make ‘decisions’ accordingly. ML likes structured data.
- Deep Learning. Deep learning is part of a broader family of machine learning methods based on learning data representations, as opposed to task-specific algorithms. Learning can be supervised, semi-supervised or unsupervised
- Natural Language Processing (NLP). NLP is “concerned with the interactions between computers and human (natural) languages, and, in particular, concerned with programming computers to fruitfully process large natural language data”
Where is it?
As well as garnering countless hours of media and techno-babble, AI is starting to pop up in our daily lives. Amazon and Netflix are deploying increasingly intelligent algorithms that get to work on their vast quantities of data enabling them to become ever smarter at offering us what we want to buy or watch. Siri and Alexa use a combination of ML and NLP to enable increasingly complex and ‘unstructured’ voice to voice interactions. Anyone who has used either will quickly testify that it does not take much to flummox either of them but the path they are leading us down is an interesting one. The fact that at the moment Siri and Alexa are little more than novel new I/O interfaces should not blind us to the fact that they will evolve into something considerably more than this in the not too distant future.
Where and when will it impact your business?
It depends. The more data your business ingests, stores and processes then the more likely that the impact will be sooner than you think. Similarly, the more repetitive your business processes are the more likely that computational ‘intelligence’ may start to be able to replace human intelligence. It’s probably worth making an on-the-fly definition and saying that AI will more quickly and effectively mimic what could be considered as ‘procedural intelligence’ and be much, much, much slower at mimicking ‘creative intelligence’. In other words, getting a computer to drive a car is a lot easier than getting one to write a symphony. That said, my children are fascinated by the new Harry Potter Chapter that Botnik have generated computationally. But J.K. Rowling need not worry quite yet.
Should you believe the hype?