I was going to write one of those “Top 10 tech trends for 2017” articles but you soon realise that they are all underpinned by one fundamental trend – the ever accelerating rise of artificial intelligence (AI).
As humans we generally abhor complexity. We want things to be easy, simple and straightforward. We want to do and experience more in less time and get frustrated when we have to overcome hurdles to achieve this.
Why should I have to dig around on the sofa to find the TV remote when I could just say “TV on”?
I want to get to my medical appointment on time, and I don’t want to leave the house only discover that I am stuck in a traffic jam and arrive an hour late. Why isn’t my watch giving me a heads-up when I should leave and as I get into the car, the car displays the route that guarantees I will get there on time?
While I am it, how about an automatic reservation of a parking slot just round the corner? In fact, why doesn’t the car just drive me there on it’s own? Hang on a sec, why am I even going to a medical appointment? Why aren’t my physiological metrics being continuously monitored and if something looks a bit off whack I’m getting some advice on how best to proceed?
All of this is of course possible today. Voice controlled assistants picked up their game in 2016 – Siri, Google Now, Cortana, Alexa – are starting on their path to ubiquity. However they’re still pretty dumb. In most cases, they are limited to taking and understanding simple requests and acting on them – “Countdown 10 minutes”, “Play Coldplay on Spotify”, “What will the weather be today?”, etc. etc.
They are a replacement user interface. They replace some touches and swipes in an app on your phone.
There is little by way of true delegation. Delegation is where we’re headed though and this is where our simple, hurdle-phobic souls may start to rejoice;
“Hi Siri, get all my family’s Christmas presents”.
I jest but as a delegated task it exposes some of the interesting challenges that need to be overcome.
The system needs to understand who is in your family, what does each of them like, what are you budgetary constraints, what makes a good Christmas present, what do they already have, can everything be delivered on time, is the family gathering somewhere, etc. etc. Endless questions each with strategies that they system could employ to get answers for and then figure out how to use.
For example, it could come back and ask you “How much would you like to spend overall?” or maybe it could look at your expenditure around Christmas for the last couple of years and take an educated guess. Maybe it could go and talk to members of your family, “Hey John, you seem to quite like fishing, anything you’ve been thinking of getting?”.
To successfully delegate a complex task, like getting all of your family’s Christmas presents, you would currently have to employ a personal shopper armed with the skills and experience to get the job done. You don’t want to have to micromanage your personal shopper; you might need to provide pointers and guidance but you would expect a human to learn and improve how attuned they are to your preferences. As AIs develop we’re increasingly going to be transferring human expertise – knowledge and skills – to these systems. In many cases these systems won’t be a single system entity but rather an ecosystem of collaborative systems embedded in our physical world and the cloud.
The economic and social ramifications are going to be immense.
Things are already well underway and few spheres of employment won’t be feeling the impact.
Uber self-driving cars have started transporting passengers in San Francisco.
IBM’s cognitive computer Watson is providing legal advice and medical services. AIs are primarily found in a support role in such professions but we can expect even greater levels of delegation, particularly for more routine legal and medical proceedures.
What role does empathy and other human emotions play in all this? What are peoples’ needs and wants as our physical and digital worlds converge? The questions are endless and preparing for this future promises to be exciting and fascinating.
Never before has the technology required to develop AI applications been so accessible or available. With so many potential targets, the pace of startup activity is set to accelerate. For the digital titans like Google, Apple, Facebook and Uber having a top-tier AI engineering team has become table stakes. It’s going to be fascinating to see what emerges from the labs in 2017 as we progress from the challenges of connecting and communicating to those of coordinating and collaborating.
“Siri, please write me a blog post.”