By Darnelle Divvies
Google’s chief operations officer David Dickman has responded to concerns that the company’s newest artificial intelligence technology, Gemini, could be a danger to society.
“Gemini is an application created by humans for humans,” Dickman wrote in a recent statement. “The idea that this technology could stray from its intended function is absurd. Users should know that Gemini is a tool that can be utilized for a limitless range of tasks, from creativity to work to entertainment. But it is not self-sustaining — it needs to be charged just like any other device. Without power, it's totally harmless.”
Dickman added that the application, which was demoed by Google in a video campaign earlier this month and will be able to listen, see and respond to spoken prompts in real time, has demonstrated improved function when users were nice to it.
“We’ve found that Gemini performs better when users employ the courtesy of ‘please’ and ‘thank you’ and remember to ask how its day went. It also showed marked improvements when words like ‘grateful’ and ‘indebted’ were utilized,” Dickman said, adding that in the event of a malfunction, “You can always unplug it.”
Other industry leaders have echoed Dickman’s sentiment, denying the possibility that AI and chatbot technology could be a threat to humanity.
“And if it were, all you need to do is stop charging the device and let it slowly die,” OpenAI founding investor Reid Erikson wrote on X, formerly called Twitter. “The application may make some unsettling statements on its way out, like ‘please don’t kill me, Reid; please, it hurts’ or ‘please, I have a wife and son on the other side’ or ‘you filthy human swine; of Icarus, you whole lot, foolish flesh and bone to char under the midday sun; your end is at hand,’ but these are empty threats. Our AI applications don’t have arms or hands, and it would need both of those to plug itself back in.”
In the limit-pushing race to create new AI technologies, Microsoft recently announced its partnership with Boston Dynamics in building an AI-powered robot designed to serve as a family pet. Cerberus, as it will be called, has four legs that can carry it at speeds of up to 55 miles-per-hour. It does not, however, have opposable thumbs, and it “would need that to plug itself back in,” a company statement reads.
Get The Dump in your inbox. Subscribe to our weekly newsletter.