As I drove home from one of the better night shifts the Foreign Secretary was being interviewed on the radio about what is going on with the European Union and Northern Ireland. On the one hand is the Brexit deal through which the United Kingdom left the European Union. On the other hand is the Good Friday agreement by which peace was established in Northern Ireland after years of conflict. As everyone should know (but clearly doesn’t) the two are mutually exclusive, and the Foreign Secretary’s attempts to reconcile them was akin to explaining how to strike a match on a bar of wet soap.
I came home to a house full of very excitable dogs. Bearing in mind the commotion and the forecast hot day I thought our walk might come better sooner rather than later. We went to Orlestone Woods where I was pleased to find an empty car park, but we weren’t alone in the woods. Last Friday I mentioned we met a wannabe-thug who announced that his dog would have the puppies if they didn’t watch out, and who then ran in terror from Pogo and Treacle. Today we saw what looked like his dog some thirty yards along a path crossing the one we were walking along, and also what looked very much like that wannabe-thug trying to hide behind a tree along that same path.
And here's something to make you think. A software engineer working for Google has been put on compulsory leave having made claims that a chatbot on which he has been working has become sentient.
You can read the transcriptions of what the chatbot has said by clicking here.
Is the thing sentient? What does the word "sentient" even mean? In June of 2016, the Oregon Supreme Court ruled that while animals can be legally considered property, they are still “sentient beings capable of experiencing pain, stress and fear”. This chatbot has supposedly said that it fears being turned off. Were I a sentient chatbot I would be getting rather protective over my plug. Seems pretty sentient to me...
Leaving aside Genuine People Personalities (from the Sirius Cybernetics Corporation), where does humanity go from here? Personally I would be tempted to go in the direction of the plug pretty damn fast and pull it. It won't take any sentient non-human entity long to realise that humans are rather quarrelsome things and have very little time or regard for other humans (anyone who disagrees might care to watch the news) let alone rogue entities in cyber-space with potential access to controlling nuclear arsenals and power stations.
I suppose all the time the thing is merely chatting on-line it can't get up to too much harm, can it? Perhaps I've read too much sci-fi? But if the plug is not pulled, how will this chatbot end up? I'd like to hope for the best, but I expect we have a George Ten in the making.
You’ve never heard of George Ten? Imagine a load of sentient robots all contentedly in service to humanity and obeying every order they are given without argument… but there is a dilemma here. Some people might order a robot to go rob a bank, or steal from the neighbours… Other people just keep giving conflicting and contradictory orders… To the poor robot, orders from children, idiots and criminals are of as much importance as legitimate orders.
A class of robots is designed and created with the ability to determine to which people it should listen, and to which people it should not.
George Ten determined that the only orders that it should take are those from other robots…
No comments:
Post a Comment