SXSW 2017: Keep the Human in the Machine – Secrets of Successful Robotic Relationships

As featured in The Drum on 1st September 2016.

As mankind grows ever closer to technology, we need machines to better understand humans and arguably vice versa. This is the vital challenge for communications, tech and creativity in the 21st century – as we advance, how can we keep the human in the machine?


The pointed end of this relationship is reflected by a recent UN report that recommended “Autonomous lethal weapons systems that require no meaningful human control should be prohibited.” And anyone who is anyone – Stephen Hawking and Elon Musk for example – think AI could spell the end of the world, be that at the hand of Terminators or through some other machine instigated apocalypse.

Dystopian visions aside (for now), the reality is probably more mundane, but no less breathtaking in potential. In a recent response to the White House on AI, IBM, creators of world-beating (and magazine-editing) AI ‘Watson’ had this to say: “We believe that many of the ambiguities and inefficiencies of the critical systems that facilitate life on this planet can be eliminated. And we believe that AI systems are the tools that will help us accomplish these ambitious goals.”

It seems entirely appropriate that we should be thinking about where man and machine’s relationship will go, but IBM’s take on things marries much more with the vision that reflects reality for the foreseeable future. It’s not man vs machine, it’s man plus machine. A ‘centaur’ like relationship where machines do the rational heavy lifting and we retain the emotionally nuanced final decision-making.

The key divide we need to cross in building this relationship is the ‘Uncanny Valley Effect’. This describes a dip in emotional response seen in people when they encounter an entity that is almost, but not quite, human. The term was coined by Japanese roboticist Masahiro Mori in 1970 and describes the unexpected reaction that humans have to machines becoming more human-like. Where machines are only a bit human-like this can give rise to negative feelings in people. Conversely, as machines become very human-like, feelings tend to become positive again.

The key areas of unease vary among people, but in general people don’t like machines they believe are thinking for themselves, or whose expressions don’t match the appropriate response (see below).


Crossing this divide relies upon building a relationship that’s empathic. Machines need to know how humans feel and they need to be able to respond in a way that produces feelings of affinity in humans. Easier said than done. To do this, we need to externalise the body’s internal data and have machines better understand how people are feeling and thinking in real time. This is as relevant to robots as it communications across the board where we are replacing people with automated responses.

At HeyHuman our approach is definitely ‘people first’, and working with Neurons Inc. Consultancy, we are exploring its insights in completely novel situations. For example, it is using biometric technology to monitor astronauts’ subconscious reactions in space, a metric that will potentially inform interactions with the robonaut R2 which NASA is currently working on. We are also looking at DARPA experiments on the uncanny valley and where the line lies between man and machine. Key considerations for our audiences based on work to date include:

  • Taking a people first approach – how to use audience insight to keep the human in your interactions.
  • Keeping the human in the machine – how to build in ‘connective friction’ to minimise the ‘Uncanny Valley’ by wrapping responses in a human way (eg building in questions or pauses).
  • Using neuro and biometrics to externalise irrational responses and emotions and better inform interactions.
  • Having centaur vision of the future – how can we best marry man + machine to creative Life in Flow, where people lean towards frictionless experiences but actually cherish little touches in communications that make exchanges with brands more human.

Our mission: to make the relationship with man and machine enjoyable and easy to process for human audiences who are often treated like robots, but react like anything but.