Humans, not machine intelligence, must remain in control of the future, says the chief scientist at OpenAI The Polish chief scientist at OpenAI, the firm behind the prominent ChatGPT AI platform, has warned that “no one is prepared” for the consequences of unbridled growth in artificial intelligence. Jakub Pachocki said in an open letter published by OpenAI last week that large language models are “starting to push the boundaries of science” and are presenting “clear new dangers” in terms of computer security. Pachocki said: “If AI development continues along its current path, the systems we’ll see in the next few years are likely… to increasingly drive their own development. “This is a time that calls for extreme caution. I am concerned no one is prepared for the consequences of a continued rapid rise in machine intelligence,” he added. The Pole warned that machine intelligence is beginning to “exceed” that of people in “transformative ways”, and that the complex systems can simulate aspects of human behavior and are becoming increasingly difficult to understand. Pachocki said that steps will need to be taken so that further development and implementation of AI ensures the technology “works out well for humanity” and that “humans remain in control of the future” rather than an “alien intellect exceeding our own.” He also called for the preservation of the qualities that people bring, saying: “We need to find ways to preserve human agency and enshrine an intrinsic value to being human.” The Polish scientist said he hopes that people in the AI industry begin to voluntarily slow down development of the technologies until safety bars are established and said governments around the world should prioritize coordination on AI development. Calls for transparency Some AI experts, while agreeing with Pachocki’s sentiments, have said that more needs to be done by OpenAI and other companies in the industry to ensure safe use and growth of the technologies. Nathan Calvin – who works for Encode AI, an organization advocating for increased regulation of AI – praised Pachocki’s comments on social media but suggested that OpenAI should be more transparent about the dangers of machine intelligence. Calvin said: “One of the most important things [OpenAI] can do is share far more information about what they are seeing that is making them call for caution. “This action may require overcoming some… incentives that exist against transparency, but the stakes are high enough that business as usual and traditional legal aversiveness is not acceptable,” he added.
