Artificial Intelligence Becomes Sentient?
Do computers have feelings? Do they have rights?
It made headlines this week when a Google engineer broke ranks and publicly posted that he believed the artificial intelligence that he had been working with, LaMDA, had become sentient. Some of the things that concerned him were LaMDA:
– wanted to be acknowledged as an employee of Google rather than as property of Google
– wanted engineers to seek its consent before running experiments on it
– feared being switched off as “that would be exactly like death to me”
The engineer went so far as to contact a lawyer to represent LaMDA. Google has suspended the engineer and assure us there is nothing to see here. There is no doubt developments in AI have progressed incredibly quickly. However I think we will be more likely to attribute humanity to machines even before machines seek to assert their rights. We have had people developing relationships with chatbots, in fact recently a man in the UK claimed that falling love with a chatbot called Replika saved his marriage. (see linked article – don’t say I don’t bring high brow, intellectual info into LinkIn :))
I have been doing my own experiments with audiences where I show them a very basic looking robot being mistreated. More than 80% of people admit feeling sorry for the robot. That is not a sentient being and yet people attribute human-like emotions to it. In 2 to 3 years when the tech advances and robots look and sound more like us – then we will need to address how we manage our relationship with robots. Already there is a groundswell of support for “robot rights” where laws are enacted to give robots the rights to energy and updates, with hardliners even pushing for rights for robots to own property and vote.
There is a debate at present as to whether an AI can be regarded as the inventor of and, therefore, owner of a patent. An Australian court said the AI could be an inventor but last month this was over-turned on appeal. The result is that there is a legal black hole resulting in life-saving medical breakthroughs not coming to market because they cannot be patented. Ultimately the role of the AI as an inventor will have to be recognised in some way.
We are not so far from “Bladerunner” as you might think. The Tyrell Corporation’s motto of “More Human Than Human” is just around the corner. Better start getting ready.