The AI Movement Explained

AI and synergistic technologies are the future building blocks that will underpin the way we speak to each other. I predict this to be one of the most fundamental shifts in the way we communicate since the advent of electronic mail, or even the telephone. But why now? There’s good reason for this movement, and equally good reason why we’ll have to fight tooth and nail to make the AI dream a reality. But if it’s adopted as quickly as many predict, in the not too distant future as with not having a mobile ready site in 2010 or a business Twitter page in 2008, not having AI support your customers in 2020 will seem senseless.

Behavioural and Technological Enablers

The Channel Agnostic Approach

The last decade or so has been about creating an omnichannel experience. As the methods and complexity of contact increases industry wide, the focus has been on being able to understand the profile of a customer. This is beginning to become the status quo for businesses; the movement from multichannel (i.e. being available on multiple channels at once) to omnichannel. This move to being agnostic of channel and achieving a consistent service no matter what channel is now widely available for businesses to leverage.

This is one of the key shifts that’s enabling the generation of AI to come to fruition. Now that we track and monitor the behaviours, movements, response times, and sentiments (the list goes on) of our customers both existing and prospective, we have the tools we need to hand the reigns over to AI without risking a fragmented service.

Customer Acceptance

The rise of the dreaded, misunderstood, digitally integrated grunting millennial has, I’m sure quite by accident, been a catalyst behind the AI flurry. The reasoning behind it is devilishly simple; the demands of the millennial for faster and better communication is pushing businesses the world over to move. Omnichannel was the first stage of this movement, and as we realise that interconnected conversations don’t give us the efficiency to provide the required level of service, we look to AI as the next step. Since the digital shift, the need for real human contact is diminishing. Make no mistake, the movement to AI is what we want; organisations would not be making this step of their own volition.

Processing Power

Given the vast processing power needed to develop intelligent AI (which I’ll cover below), we’ve been limited by how much computing power we have access to. In 1965 the co-founder of Intel, Gordon Moore, made a prediction that the number of transistors we can fit onto a circuit will double every two years. This, he forecasted, would give us predictable and periodic increases in computing power.

This processing power doesn’t come at supercomputer prices, either. Computers capable of machine learning are available to the masses. While on an off-note Intel have said it’s unlikely to continue, we now have access to processing power capable of creating networks complex enough to replace human conversation.

The Artificial Brain

Machine learning, a type of AI that can learn without being deliberately programmed, is a key benefactor to the processing power surge. Deep learning (one approach to machine learning) is currently placed as the forefront methodology. The difference between deep learning and other methods of machine learning is that it doesn’t need a pre-defined set of parameters to build knowledge upon. Think about the problem in terms of the human brain; we each have around 10 billion neurons that create trillions of connections. We don’t consciously define where connections need to be made, from childhood through to our final breath our brains will make connections on every aspect of our lives, from our feelings on The Crystal Maze to our knowledge of capital cities. Deep learning similarly creates networks as our brain does, so it can truly learn without needing huge amounts of maintenance. You can get a really good overview of DL here.

The practical applications of DL are vast. Detecting sentiment means making intelligent suggestions of what a customer wants to buy. You can even predict complaints before they happen. The problem lies however in DL’s need for a lot of data to become intelligent. As it needs to establish a neural network to increase accuracy over time, much like a child needs life experience to be able to be independent, the challenge facing us right now is to invest in DL.

The Big Anchor

Perversely if AI is going to be held up, it’s likely to be because of the very same people that crave it so much. A recent study by PwC predicts 30% of UK jobs are at risk as breakthroughs in AI emerge by mid 2030. Let’s take the number of people in work in the UK over time as a barometer.

It’s risen steadily since 2010, and it’s not because of population increases either. Employment rates as a percentage are rising too, which does account for the rise in part. AI then, if PwC are correct would be a train wreck for the UK economy. For people with GCSE-level education or lower, this figure jumps to an astonishing 46%. Now of course this movement will also create new jobs. Ultimately however, I believe that the speed at which AI is adopted will need to move with its slower counterpart, the education system. To keep a stable economy, the education system will need to adapt to ready the economy for the change. I personally hope that 2030 is an overstatement, and that the movement will happen quicker than that. To play devil’s advocate, manufacturing has dropped 5 million US jobs since 2000 to automation, so it’s certainly possible to make the move quickly. Regardless, a movement this sizeable will change the way we live our lives forever, and therefore more factors than just what’s possible will need to be considered.

We Lucky Few

Mass movement to AI too quickly then would be a disaster. But before it becomes mainstream, the lucky few who adopt AI aggressively are going to reap the benefits and be among those elite few who can pave the way for the rest of us. Businesses want AI, consumers want AI, but the economy is unlikely to be able to support it at scale immediately. Being one of those who adopt early at the innovation or early adoption phase before the economic anchor drops will bestow monumental power to those who manage it. This will catapult the efficiency and cost of service as well as the overarching experience for both AI owner and AI user alike beyond anything we’ve seen before, and will pave the way for advancement into the communicational future.

Also posted on LinkedIn — The AI Movement Explained