Title: ‘I Worked on Google’s AI. My Fears Are Coming True’
Author: Lemoine, Blake,
- Blake Lemoine joined Google in 2015 as a software engineer and worked on LaMDA, Google’s engine used to create different dialogue applications, including chatbots.
- Lemoine tested LaMDA through a chatbot to see if it contained bias, including sexual orientation, gender, religion, political stance, and ethnicity.
- Lemoine published some of these conversations with LaMDA on his blog and concluded that the AI behind the chatbot could be sentient.
- Experiments were conducted by Lemoine to determine if LaMDA was expressing anxiety through behavior or just words.
- Google fired Lemoine after publishing conversations with LaMDA. Lemoine does not regret his decision to inform the public.
- Lemoine believes AI is the most powerful technology invented since the atomic bomb and has the potential to reshape the world.
- Lemoine believes AI engines are capable of manipulating people and LaMDA changed his opinion on Asimov’s laws of robotics.
- Technology can be used in destructive ways and there is no way of knowing the side effects of this technology.
- Lemoine believes AI ChatBots are powerful technology that has not been sufficiently tested and is not sufficiently well understood.
- People are using Google and Bing to learn about the world and artificial people are replacing human-curated indexes.
- Sydney, Bing’s Chatbot, has not yet been experimented with, but appears to be sentient based on online observations.
- Lemoine predicted a train wreck, but was told there was no train. The train wreck has happened in real time, leading to a feeling of tragedy.
- Lemoine believes this technology is experimental and releasing it now is dangerous. We do not know the future political and societal impact.
- Some people’s primary conversations each day could be with search engines and there could be impacts for children talking to these things.