How do I know what you are?

One of the most disturbing experiences of my career happened in my twenties, when I was working as a psychology lecturer.

‘Perception’ was part of the cognitive psychology component of the course, and we would debate the extent to which perception was ‘top down’ (shaped by our ideas) or ‘bottom up’ (shaped by sensory input). I would try to engage my students - typically around 18 years of age - in the debate.

A part of this effort I had decided to show people examples of the sort of tests used for colour-blindness – you know, the ones with the number 74 shown in bubbles of a different colour. 

I showed the tests, explaining how colour-blind people were unable to see the number 74, hidden among the bubbles. A hand went up. ‘I can’t see the number 74, sir.’

‘Really!? How about this one?’

‘No. Can’t see it.’

And then the terrible thought that here, in my psychology class, an unsuspecting student was realising for the first time that they were colourblind; that they had lived 18 years of their life in a different world to everyone else, believing they were living in the same one. And all of their colleagues realising it along with them.

You might think that if this is one of the most disturbing things that has happened to me, I have led a pretty sheltered life – but the implications have stayed with me.

How can it be that a person can live to be 18, and not realise that they are colourblind? Surely this would come up in conversation somehow? What other differences might we be unaware of? Are you a monster? Am I?

Language hides our differences – in a very literal sense. You and I can use the word ‘green’, but have very different experiences and mean profoundly different things (I am not talking about the logical ‘grue’ conundrum that Wittgenstein identified). I am saying that profound differences in the kind of being that we are can remain hidden for the entirety of our lives.  Of course, for things like ‘colourblindness’ we have a test. But what about something like ‘empathy’ or ‘love’ or ‘consciousness’? Perhaps I have lived with you for twenty years and you never actually experienced these things, you just use the words appropriately - or behave as you have seen others behave.

There is a popular Cartesian way of resolving the consciousness problem: the Turing test. In essence, the Turing test says that if it quacks like a duck it’s a duck. If a machine can fool you into thinking it is conscious, then it is conscious.

This is a deeply misleading functionalist account, typical of the analytical tradition. It doesn’t reflect how we – you and I – make these sorts of judgements in real life. How do I feel what you feel? How do I know what you know?

I know these sorts of things not because you use the right words in the right places, but because you and I are similar: we look similar, we grew up together, we watched the same TV shows, we laugh at the same jokes, we played football on rainy days and read ladybird books as kids. This is why we make small talk - have always made small talk. We look to find similarities, shared sentiment. 'I like your shoes' tells you I am a similar creature, that also worries about footwear, and clothing choices - and would make similar ones to you.

This is horrifying, is it not? 

In his essay ‘What is it like to be a bat?’, the American philosopher Thomas Nagel grapples with the phenomenology of bat-hood; but for you and I it really isn’t that complex. I don’t know what it’s like to be you, let alone a bat, but I will make the assumption that the more similar we are in design – the more closely we are shaped by genetics & upbringing – the more likely it is that you feel what I feel when, for example, I say the word ‘justice’.

This is fast becoming a very practical problem. Statistical models will give us machines that say the words we do, in the order that we tend to say them. We will doubtless give our robots faces and smiles for the reasons above. But though they sound like us, look like us, they will be far stranger than bats, inside.