Dear Ai… I think, therefore I am?

25th April 2018

Most folks are about as happy as they make up their minds to be. – Abraham Lincoln

Machines can’t think. There. I’ve said it. Anybody out there who are talking about computer programmes that can start to ’think’ are mixing their metaphors. A machine can mimic, and we can teach machines to perform things on their own now. But it’s not thinking. Thinking is a neurological gift and the thing that separates us from the toasters. It’s best not to get into a big debate about this here… it’s not the time or place… I get that Artificial neural networks (ANNs) and connectionist systems are vaguely inspired by the biological neural networks of animal brains, but it’s still code. So just don’t get too philosophical about it.

Now that’s out of the way, to business… training machines to do better things. To mimic the good stuff, not the bad stuff. Something we’re very keen to do at Us Ai, and something I wrote extensively about in my book.

If more programmers and algorithm junkies, studied fields such as positive psychology (the study of happiness), then I wonder if some of the outputs would be different. Whereas traditional psychology has always focused on dysfunction—people with mental illness or other issues—and how to treat it, positive psychology, in contrast, is a field that examines how ordinary people can become happier and more fulfilled. What if we trained machines to focus on that as a sub-routine? If even the most basic bots had more positivity at the dialogue level, we’d start to create a lot of good things. Better things. Potentially.

Imagine a world, where the bot you are speaking with every day (be that Alexa or Siri or whatever) only exuded excellent and favourable results. Deliberately programmed to be happy. If it used its engagement opportunities to help create and transform your energy into positive reality, the results could be an enormous benefit to humankind.

Lots of successful people will tell you that they got where they are now because they made a lifestyle around positive thinking. A person that faces life with a positive attitude will always be more successful in life both professionally and personally, than a person that cannot take control of their negative thoughts. That’s just reality.

It’s worth noting that psychologically speaking, there are two levels of positivity;

  1. At the subjective level (past, present and future constructs) the past involving well-being, contentment and satisfaction; the present involving flow and happiness; and the future involving hope and optimism.
  2. The individual level – individual traits that are positive, such as character strengths, talent and the capacity for vocation.

But what does that have to do with technology and Ai I hear you ask? Well, I believe, it’s entirely possible to make sure the machines we’re creating to automate responses or perform tasks – our new search engines – can be positivity generators too. To amplify both the subjective things about us and the individualistic things. Machines that are positive by design, for all. Also, machines that learn about us one-on-one, and teach us to be more than humanly possible.

What we say is what we do

What if we spent more time teaching bots only to use positive words when they talk to you? It’s not something copy-writers and support staff think much about when they’re writing responses. Or when tech’ folk are training machines to answer things generatively, they’re not thinking about vocabulary and nomenclature; they’re just thinking about faster results. We need to spend a lot of time replacing negative words with positive ones.

Avoid the negative, focus on the positive

Remove all the answers that are not positive! We reach for technology for two reasons; Firstly to by-pass effort, and secondly, when we don’t know the answer to something, or need some positive reinforcement. When the machine answers, not only should it respond in kind, but it should respond with encouragement. It should not let the negative moment when you did not know something conquer you. Even if for a few hours a day, remove negativity and focus on the positive things in your life. For example… when a bot gives a response, it could say “now you know this answer for next time” to reinforce that you have in fact just learned something new!

I have not failed. I’ve just found 10,000 ways that won’t work – Thomas Edison

To boil it down a bit lower-level, what I’m actually talking about, is a machines explanatory style. An explanatory style is how we explain why events have happened. Machines with an optimistic explanatory style will give you credit when good things happen, but typically help you also understand the outside forces that shaped any adverse outcomes. Machines with an optimistic lens will also help people see adverse events as temporary and atypical.

On the other hand, some software with a pessimistic explanatory style can blame the individual using the machine when bad things happen, and fail to give themselves adequate credit for any successful outcomes. The ‘error message’ on a form is an excellent example of this… “error Will Robinson, you got that wrong.”

Positive affirmation – a gift from the bots

Machines should only use words that evoke strength and success. They can fill our thoughts with words that make us feel secure, happy and have control over our lives. Copywriters should make an effort to focus on these words rather than the ones that suggest failure or incompetence.

Train the machines to deliver positive affirmations. One of the most common ways to encourage positive thinking is a positive affirmation. What does that mean? A machine should repeat a positive phrase like “you deserve to be happy” or “you deserve to be loved” from time to time. Just a random nudge. When we believe that these statements are correct, and they are repeated in our lives, they can impose a more favourable opinion about our experience.

Machines that help us navigate

Use the bots to redirect your thoughts! It’s a method used by psychotherapists to help people control ideas when we start to feel negative emotions like depression or anxiety. When we a machine detects negative emotions or sentiment in your tone, why not have it deliver a positive image, something that makes us feel better, and keep our negative feelings under control. It sounds silly or trivial, but as our relationship with the machines grows deeper… this little tiny moment every day, could be life-changing, and even life-saving.

Machines can help us to start thinking more about how we can succeed. There is nothing in the world that compares with self-confidence to create a successful reality. Putting our doubts aside and believing that we can achieve is immeasurably suitable for us and our environments. It’s contagious too. A machine that encourages us to ‘keep on going’ is a remarkable idea. How come our fit-bits and trackers don’t proactively remind us of what we’ve achieved? Rather than just encouraging us to clock up what we’re doing?

Use machines for what they’re good at

The single most significant power of data is giving us the ability to analyse what went wrong (and what went right of course). Having our data used, by the bots, to help us see what went wrong and what led to the current situation, helps us to avoid future mistakes and look forward more positively. What a remarkable gift we created in all that data we swim amongst.

The machines can be programmed to think (?) of those failures as opportunities to assist us. Sometimes the most negative things in our lives give us opportunities that we may not have seen otherwise. Training models that help us work at our imagination to visualise what went wrong, but what the flip-side looks like, can be a great motivation to take you there and make you think more positively about the distance that you have until the destination.


You might be tempted to read this and assume that it implies seeing technology through rose-tinted lenses by ignoring or glossing over the negative aspects of life. But not so. If you’ve read this and scorned, you’ve made my point beautifully.

Wellbeing may be viewed as an apparent key outcome to my thinking, but the amplification of our character strengths can also be colossal success metric in life, work, and the home. Even at a professional level, if machines could help to contribute to making workplaces more positive and meaningful places to be, that benefit alone could be huge. Think of it a bit like having a life-coach in every employee’s pocket — “Well done, keep going… you could have handled that better, but we’ll learn and move on”… If we developed Ai infused organisations that can better leverage the strengths of their employees, even the economic value by reducing phenomena like ‘presenteeism’ could be the single most significant growth driver in decades. I kid you not.

We know that encouraging positive thinking reduces daily stress, builds strong confidence, gives us a happier life and promotes better management of important decisions. So let the machines we have to deal with every day, amplify the good things.

Training a machine to be positive is not easy. Imposing a positive thinking model into code is complicated and of course subjective. But it is possible.

Tags: Ai, article, bots, happiness,