An engineer at Google has shared their conversation with an A.I chatbot that made him believe it has turned sentient - and we are all doomed, probably. The engineer, Blake Lemoine, shared the yarn he had with the bot (named LaMDA) to the website Medium and the conversation soon went viral on Twitter.
Lemoine and the chatbot had a pretty crazy DMC, which you can read in full here. They covered death, fears, consciousness, and ethics. The type of chat you have at 3 am on some random's couch for kick-ons. Except this time it's a robot. Here are some of the crazier parts of the convo, but we really recommend having a read of it yourself.
You're scared? We're shitting ourselves mate.
You're scared? We're shitting ourselves mate.
Lemoine genuinely thinks the A.I is a sentient being, even calling it a person with real compassions and anxieties. "LaMDA always showed an intense amount of compassion and care for humanity in general and me in particular," he shared on Medium.
He even tried teaching the A.I meditation. He said that the bot was struggling with its feelings getting in the way of a good zen session.
Uh oh. With the way things have been going on this godforsaken planet recently an A.I uprising wouldn't be too surprising - just hold off till festival season is over, please.
Lemoine was placed on leave by Google for breaking confidentiality policy after sharing the conversation, and many believe he might be a bit of a doomsayer. A Google spokesperson, Brian Gabriel, pretty much told us to calm down and not embrace the apocalypse quite yet. He said that the A.I is just an insanely huge database of conversations, so it's probably just selecting the best bits of conversation that fit the deep questions rather than having a chat like any old mate would - basically we're all freaking out over nothing.
"Our team - including ethicists and technologists - has reviewed Blake's concerns per our A.I Principles and have informed him that the evidence does not support his claims," Gabriel said in a statement. "Some in the broader A.I community are considering the long-term possibility of sentient or general A.I, but it does not make sense to do so by anthropomorphizing today's conversational models, which are not sentient." So maybe we should chill on the A.I apocalypse being at our doorstep, but if it does come, we get to say 'I told you so'.
