Deep fake photographs of Donald Trump being arrested and pass-grade answers to the Uniform Bar Exam are just the tip of the iceberg when it comes to the potential for artificial intelligence (AI) to distort the information we consume today. It’s now many years since I first became disillusioned with how easily the internet could be used to corrupt, and many more years since I had been a vocal advocate of the potential of this exciting thing we first called the World Wide Web. I still believe that as a force for good, it has unlimited value. However, years of unscrupulous manipulation, disinformation, misinformation, and algorithm-induced rabbit holes have made it a minefield for many users. The latest rush by tech heavyweights to grab a share of the potentially lucrative AI market has all the hallmarks of yet more disastrous use of brilliant opportunities. I’ve watched the progress of San Francisco start-up Open AI since a friend’s son described it as a ‘game changer’ about a year ago. In recent months it has received a lot of attention because of its ability to write convincingly about whatever subject it is asked to. Last week I asked it what are the dangers posed by AI? Its answer stated that apart from the potential for massive job losses and the dangers of nefarious interference, it could also be programmed to carry out malicious tasks autonomously, without the need for human input. This could include activities such as collecting sensitive data or manipulating financial markets. It suggested that the biggest danger is that ‘it is able to learn and evolve on its own, without an established set of parameters or boundaries.’ This means that it can potentially make decisions that could cause harm to humans or the environment. For example, an AI system could make decisions that lead to the destruction of the environment, or to the misuse of resources. It could be used to create autonomous weapons—systems that can use AI to make decisions without any human input. These aren’t my opinions—these are its own predictions. So I asked it whether it would ever mislead me. It began its reply with the words: ‘Hand on heart…’
