Will ai Ever Understand Us?

Will ai Ever Understand Us?

Language is a tool, even the bees use it. They turn in circles, among their peers, describing the location of food. They are saying: 'If you fly here, there is food'. You might think they could be saying: 'I flew here and found food'. But consider, there is another bee, turning in the opposite direction, telling of yet different flowers. Both are projecting a future for their listeners. Language is wishful thinking. It tells the story of what we would like to be true, making the past and reality pawns to be moved and sacrificed for the glory of the game.

Putting two people with differing views in a conversation results in conflict that baffles both, that the other cannot see how facts make their position untenable. Worse, they are hypocrites.

Hyprocrisy is a crime, for the young. As we age our perspectives change. Only the young would parade along their aunt and uncle's corridor with a placard decrying abuse of dyslexics, as well as adherents to the fascist Johnson and his 18th century volumes. Today, I'd turn to Grammarly and get on with the next sentence.

Appealing to our selective past is vain when confronted with someone else's vision of the future. We cannot know, and often they cannot, the motivation for this Milford Haven¹ they so long to get to.

How does this relate to AI? We use corpra to determine the meaning of words — bundles of statistics on usage — how words have been related. Our current analysis is like an old man by the fireside talking of what past he can remember, his excitement mounting with age and retellings. So how do we 'tie the wind' and sail to the sunny uplands of comprehension. I don't know — but as the comedian Dave Allen said: "I wouldn't start from here".

Frequency and Proximity

So here is my future — we are all redefining signifiers by a new context, a network of signifiers. Indeed they bring with them their past, but a selective past, a misunderstood and distorted past. Our redefining we call speech. So meaning comes from the context — as we change the frequency and proximity of signifiers they accrue new meaning — meaning that we intend and hope will rub off on others — much like bees and pollen.

It is the context that is significant — not the static relations of a corpra. As terms accelerate towards or away from each other, they redefine each other. We can descibe the context by the change in distance between signifiers, and by the frequency of their expression. Said otherwise, signifiers are redefined by the change in their relation to others.

If this sounds like Ferdinand de Saussaure — I have made no mention of a 'referent' — then I am most of the way there. Taking a slice of time does not elucidate meaning. A picture cannot tell you if the car was speeding up or slowing down — it's just pretty.

How about an example

Adam and Eve and PinchMe went down to the river to bathe. Adam and Eve were drowned. Who do think was saved? PinchMe.

They are each mentioned three times. All in one sentence, two in another and PinchMe — please don't read this article aloud — mentioned alone. It describes a changing relationship where PinchMe comes away from the triplet alone. The fate of Adam and Eve, though extremely sad, is not the point. Being preoccupied with it as your child pinches you, would result in misunderstanding. This is not exhaustive, it's a thought experiment. But substitute Clinton, Biden and Trump and the context changes as does the meaning of 'lock her up', 'lock him up' and 'lock him up' — said by a Trump supporter in one campaign, in another campaign and after January 6th.

I am currently working on a model of the work of Heather Cox-Richardson, an historian who has been keeping a public journal for the past few years. I've been using it to observe the changes in the frequency and proximity of noun-phrases.

In order to understand we need context and time and a bigger boat. To compute what we have now takes a monster and this problem is only just yielding its potential. I hope that viewing language as a tool we can improve the expression of our intent and that our placards will not cause unintended offence.

If you go here, there are flowers.


¹ Michael Drayton (1563–1631), Milford Haven