There has been a real spell of interest in the last twosome of days in a couple of chatbots (reassuringly named “Bob” and “Alice”) developed by Facebook AI Investigate. Reports have been flying everywhere of these robots making their own sinister oblique language, along with unintelligible snippets of intriguing connections between the two of them.
A little fear around having two bots (backed by neural networks) talk to each additional like this is perhaps comprehensible, but it’s important to comprehend that they were only really doing what they were told, and the insinuations are far less sinister than some more panic-stricken corners of the media would have you believe.
As a software technologist, I have found it quite funny to sit back and watch the multitudes of articles predicting an approaching robotic revolution – I’m afraid the whole Exterminator scenario still remains extremely unlikely
The thing about Bob and Alice is that, contempt their welcoming names, they were only given one job to do: exactly, to negotiate. Initially, a simple user border facilitated chats between one human and one bot – chats about negotiating the sharing out of a pool of capitals (books, hats and balls). These conversations unavoidably were conducted in English, this being the linguistic of the human – “Give me one ball, and I’ll give you the hats”, and so on. I’m sure many exciting deliberations were had.
Comments