So here’s irony for you.
Apple are lauded for their simplicity. Some traditional PC manufacturers have 20 or more laptops to choose from. Apple have just three.
Unfortunately, the same simplicity does not apply to the software you choose to run on your Apple device.
Number of apps currently on the Apple store?
Average number of new Apps downloaded per month by adults is the US today?
Too many Apps. No-one is interested any more.
The current paradigm of proliferating Apps on our phones has to end. It’s time for a reboot.
But this is actually much more fundamental than you might think.
In 1950, Alan Turing’s seminal article ‘Computer Machinery and Intelligence’ was first published.
Shortly after, LEO, the first purpose built digital business computer for J. Lyons & Co ran its first application.
It is amazing that, within such a short period, two concepts were introduced that are both still very much in focus in computing today. Moreover, it’s interesting how their relevance is slowly swapping around.
What are these concepts?
Machine friendly computer interaction
LEO did a brilliant job of putting digital computing on the map. Humans structured information into the computer in a form it could process and it automated all kinds of stuff like payroll and invoicing.
When you fill in a form today in an App you’re still doing this. Multiple Apps to do different things, exactly how it has always been. But it has gotten way out of control. One app per function isn’t the answer anymore.
Human friendly computer interaction
In 1964, computer scientist Joseph Wiezenbaum realised Alan Turing’s ideas with the creation of ELIZA, the first chatbot. Humans interacted with a computer using natural language. This is what you’re doing when you use Siri, Cortana or a chatbot today on your phone, and it’s why Facebook, Microsoft, Google and Amazon are all over it with new products and API’s. They’re already creating the pillars of this new future.
Since the 1950’s, machine friendly computer interaction has ruled the roost. We work the way they do because they need structure. But all that is changing. Human friendly computer interaction is about to come of age and will eventually eclipse its forerunner.
In today’s CX-driven world, simplicity differentiates. It is on this basis that I am very excited by the rise of conversational interfaces and the opportunity and utility that bots and chatbots present.
Fundamentally, two things have happened to bring all of this into sharp focus:
1. AI has taken great strides in recent times with deep learning.
2. Messaging (or Dark Social) has overtaken Social.
@BenedictEvans brilliantly tweeted this last year:
“Old: all software expands until it includes messaging
New: all messaging expands until it includes software”
Messaging today is not just about chatting; it’s about a platform that can front any service with a human-friendly interface.
Successes such as WeChat in China have shown what is possible once a critical mass is reached.
Collectively, all of this presents us with a great opportunity as we look to 2020.
The thing about messaging and chatbots is the emotional engagement and the sense of someone being there, providing one-one service.
On the backend it doesn’t really matter if it is a person, a simple bot, the most powerful AI on the planet or a combination of any or all of these options. If built correctly, a good chatbot can fundamentally change customer experience and overall brand perception.
The only real big decision you need to make right now is which basic pattern to follow; whether you work to a script to keep things nicely in control or take the plunge with machine learning and train an AI by example. Either way, if done well enough, you’re in for a much more engaging online one-to-one experience with your customers than you probably have today.
Conversational interfaces are totally aligned with how our brains are wired. Computers need to start fitting with us, not the other way around. It’s time to move on. My first ever BASIC program in 1981 was a super-crude adventure game. By today’s definition, it’s a simple non-linear chatbot.
The West doesn’t have WeChat levels of messaging functionality and engagement yet but something like this is coming.
At Realise, my Applied Machine Learning team have been working for over 17 years with Natural Language problems ranging from simple site search to high-end real-time business process automation. Turning unstructured big data, including messages, into value for end users is what we do.
And with the new frontier of human based computer interaction now upon us, we can’t wait for what’s going to happen next.