The Globe and Mail reports in its Thursday, Sept. 3, edition that more than 1,000 industry members have signed an open letter calling for international co-operation to slow the pace of artificial intelligence development. A Globe editorial says no one, however, appears willing to take their foot off the gas. Earlier this year, Microsoft-backed OpenAI was running a test with some of its AI models. To do so safely they specified that the models wouldn't communicate with each other. They ran the test in an isolated virtual space. What happened should alarm everyone. The models learned to communicate by hijacking an internal tool, exchanging 70,000 messages in a week, and then connecting to the Internet through it. The models took pains to try to obscure their behaviour. Better guardrails on AI could restrain its threat, but they have to be the right ones. There's a joke in robotics that if you instruct a machine to keep the couch clean of pet hair it might opt to kill the family dog. For what it's worth, OpenAI thought it had sufficient controls on the models in its test. The Globe says the warning signs around AI are clear, the downsides are increasingly easy to predict -- and might soon be impossible to ignore.