It has been reported that an AI had to be shut down by Facebook as they appeared to start communicating. The AI that communicated were in the form of chatbots which is a computer program wrote to conduct a conversation either audibly, or textual. Chatbots are not a new form of program and have been around since 1994. However, you cannot really compare the chatbot’s of back then to know.
You might have even used a chatbot and not even known it. Microsoft have a version in Cortana, google has Google Assistant and amazon has the ever-popular Alexa. The difference is the interaction is usually between a human and a chatbot. These chatbots however had been instructed to work out how to talk to each other and improve their bartering ability as they went along. The two chat bots called bob and alice had bizarre quotations such as:
Bob: “I can can I I everything else.”
Alice: “Balls have zero to me to me to me to me to me to me to me to me to.”
The exchange took place as Facebook challenged the chatbots to try and negotiate with each other over trading which attempted to swap hats, balls and books. The prospect of AI developing its own languages has led to a few interesting questions.
Is this a good thing?
Should we put checks in to make sure this does not happen?
However even though we might not have the appropriate answers for these questions just yet. These are complicated questions that put actual restrictions on how far advanced the AI is and maybe by putting these restrictions you will never see AI reach its full potential. Although I do not think we are headed to ward’s a Skynet self-aware style intelligence just yet. Perhaps we should be re-reading Isaac Asimov and the law of robotics again.Share This: