2020: Bots and the rise of AI

Sam Abercromby

Sounds like the title for a bad sci-fi novel, doesn’t it? But this was one of the big themes that came out of the Online Influence conference in Cardiff this week.

With humanity’s need to set a marker in the sand and try to guess the unknown, predictions for what the world will look like in 2020 are lighting up the net from all angles. Judging by the recurrent themes at #OiConf from speakers across the digital landscape, it’ll be a world where apparently we’ll have more conversations with bots than with our spouses. No way, right? But take a moment to think about it and it doesn’t seem far-fetched. How often do you use online chat to avoid queueing for an hour on the phone, ask Spotify for music recommendations or tell Siri to do something? With devices such as Amazon Echo physically placing this interaction in your home this will become more entwined in our daily lives.

AI’s also getting better and while it’s a world away from being perfect (see Microsoft’s Tay) we need to look beyond where it is now and where it will be in 5 years’ time. A great example of this at #OiConf was a photorealistic avatar which used coded virtual chemicals in its virtual brain to simulate how a real mind works and react emotionally to questions and conversation. The illusion of humanity was broken with the avatar’s animation and response time, but step forward half a decade and it’s not hard to see the evolution.

So how does this fit with CRM? Well, we’re already seeing services such as Phrasee use AI to increase open rates of emails for certain sectors, allowing algorithms and processing to analyse data on a scale and speed that humans can’t do. Programmatic advertising, while it has come under fire recently, allows ads to be served quicker based on current trends and behaviours. Extend this approach to CRM and if enough data can be interpreted from enough touch points, someone who’s in a consideration/research stage either on your brand’s website, Pinterest or asking questions socially could receive incredibly targeted comms from you. It’s this access to, and volume of, data and how you interpret it that’s currently the missing piece. As current display advertising has shown, specific targeting still needs human curation - just because I’ve bought a gift for my wife once doesn’t mean I’ll always be looking for more Poldark colouring books 😊.

And who knows where the next developments will take us? Chatbots in physical stores to help you with your shopping? Perhaps even an emergence of bots as brands themselves? It’s not such a leap to envision dedicated personal shopbots to help time-poor consumers select from the latest fashion. All of which give rich opportunities for specific targeting, data insights and a greater chance of attaining that near-mythical CRM ‘segmentation of one’. The question may be just who (or rather what) will be analysing these insights in 10 years’ time.