The foggy future of Artificial Intelligence

4 October 2016Strategy series

Artificial Intelligence (A.I.), Cognify, Deep Learning, Machine Learning, Bots. All of them signify different names for the same broad idea that can be explained as “machines that exhibit intelligence”. And lately, everybody is talking about it in such an intense way that it definitely fits the definition of a “hype”. To put this “hype” in perspective, we love using the Gartner Hype Cycle. Look at the items at the top of the “Peak of inflated expectations” in the Gartner Hype Cycle 2016 below: Machine Learning and Cognitive Expert Advisors. Also, see how many items on this latest hype cycle are around this subject or at least related to it? Up to 40 percent!

Seeing these patterns brings up all sorts of questions! What will happen to A.I. when it reaches the Plateau of Productivity? What does that mean for you and me as humble product owners, marketers, developers and UX designers?

Machine learning as artificial intelligence

The rise of Artificial Intelligence

The idea of A.I. has been around for quite a long time. Throughout the years, the evolution has had many rough periods where we as humanity were stuck. These prolonged periods of little progress are called A.I. winters. However, in the last years, we’ve seen some serious breakthroughs:

  • Big Data became available for these A.I.’s to learn from. Imagine how much better a child can learn from 10 million video’s (now) than from limited amounts of data generated on a local university on a local computer (30 years ago). This data is the fuel of the A.I. engine.
  • Cheaper and better computer chips became available. This was partly driven by the gaming industry. With more processing power the A.I.’s can work and thus learn faster.
  • New algorithms were developed. These new neural networks behave like the brain with more “stacked layers” that work simultaneously on top of each other. This is where the term “deep learning” comes from.

These developments helped force a breakthrough in A.I. in the last years. In reality, this newfound A.I. is still nothing like a fully humanized artificial intelligence experience. But it is something that can be described as narrow A.I. with a more practical use.

Think about narrow focused Artificial Intelligence:

  • In self-driving cars that learn how to drive like the Google and Uber cars that already exist.
  • Learning to recognize your face at your local bank office to help you with a more personal experience.
  • Filling in your tax form for you, based on all your financial information
  • As a Facebook Messenger bot that will book your flights and hotels for you, based on your preferences in the past.
  • Being your virtual planning assistant and managing your calendar and meetings.
  • Making a personal music playlist for you that fits your activities of the day.

Smart stuff, but nothing that seems super crazy extraordinary for today’s innovations. These are just the very first steps. We might judge these innovations not as Artificial Intelligence, but just as smartness.

The future of Artificial Intelligence

Soon, A.I. will be influencing everything. Futurist Kevin Kelly (who I wrote about before) says the following: It is hard to imagine anything that would change everything as much as cheap, ubiquitous artificial intelligence. A.I. as an easily accessible layer of intelligence will fuel and shape products, services, and media. Sounds like big talk. Can we dream?

“It is hard to imagine anything that would change everything as much as cheap, ubiquitous artificial intelligence.” (Kevin Kelly).

What if Artificial Intelligence will:

  • Be your doctor and advises you on your medical situation, because it has experience of millions of cases?
  • Advice on and manage your wealth smarter than any team of humans can because it has oversight and insight in all financial, political and business movements in the world?
  • Invest in, manage and expand entirely non-human operated fleets of cars, trucks, planes at a crazy low-cost rate. This will fuel the next generation of transport.
  • Create movies, songs, books with the quality and the experience that will trigger real human emotion.

So what is left to do for us if these A.I.’s take over all our intelligent work? Are you ready to accept that you might not be necessary to do your own job? If you don’t want to accept it yet, join the group of people who are in “Artificial Intelligence Replacement Denial”.

In any case, don’t worry just yet. Your current “human” job as product owner, marketer, developer and UX designer might slowly disappear. But plenty of new and more interesting jobs will appear. Firstly, A.I. isn’t going to just pop-up and exist. It will take a long time before it reaches the plateau of productivity from the Hype Cycle described earlier. Secondly, these intelligent learning systems need to be raised just like children. If you teach them bad words, they will use bad words. We will need to slowly shepherd these AI’s into full ‘grown-up’ systems that can perform their job in a way we would expect grown-ups to do. A serious job in the future will be Interactive Machine Learning. “The principles for designing systems that let humans use these learning systems to do things they care about”.

Evil angry robot artificial intelligence

Is it scary that A.I.’s will become so smart and more powerful? Yes, definitely. But if we raise them carefully, teach the A.I.’s the right stuff and make sure that there is not one or a few people in the world who will control them. Then we can benefit immensely from this development. Think about this super smart layer on top of everything you do in your daily life. With access to a lot more data and the intelligence to gather, translate, adapt, interpret and communicate it. That can be awesome! Imagine a future where the A.I. layer of a brand thinks and makes marketing and branding decisions while not talking to a human, but with the A.I. layer of the user that filters out the not practical or relevant stuff. Branding will become less valuable. Relevancy will become the trigger.

The near future

There is still a long road to go before we get the big wet A.I. dream. But our goal for the near future is to move there early and try out what we can to learn from it! The goal for aFrogleap and for our clients is to improve the customer experience. It is all still about the customer experience. We are getting this super powerful new extra layer, and we need to use it properly. Whether it is in the much-hyped Facebook M/WeChat/WhatsApp bots that already exist or will pop-up soon to become a new channel or if it is in real and useful tools that are just substantially smarter.

In the coming period, you can expect more blogs around the A.I. topic. We will look at the practical implication in the short term with topics like Conversational bots, smart assistants or service flows optimised through A.I. We will also look at our own A.I. Replacement Denial fear, will we still have our job in the future? Cool stuff to think about!

Are you as fascinated as I am about this awesome emerging technology? Here’s a list of books and articles to read about Artificial Intelligence that has inspired me: