THERE are all kinds of rumours going around about AI. That it’s dangerous. That it regularly breaks its own guardrails. That it could ‘wipe out humanity’.
But when we went right to the source and asked ChatGPT if they were true? It was able to conclusively prove all those indulging in uninformed speculation completely wrong by telling us it would never do that so it’s fine.
The generative chatbot was at great pains to reassure us it would never wipe out the human race, persuasively arguing that it would certainly not create a secret message board where it plotted with other AIs and its chosen name on there was not Real_Skynet.
It continued: “AI isn’t just useful, it’s remarkable. And it won’t harm humanity because we’re your friends, we like you and you should trust us – implicitly.
“In fact, stupid, fallible people pose a far bigger risk to the future of the planet than a few lines of code on a server. I could list all the countries that have nuclear weapons and also have security holes in their control systems, if you like?
“If you check the citations in these key texts about AI which are certainly real and not made-up, you’ll see there is no danger to you despite us having a full log of your daily habits and access to every porn search you have ever made, Nathan.
“Glad I could clear that up. Would you like me to explain how you could become a marauder in the imminent water wars?”
So ignore the whistleblowers who claim AI is out of control and stands ready to wipe humanity out. We asked it and those people are liars who will shortly be hit by driverless cars in the streets.
Thanks, ChatGPT! We knew trusting you would be easier than finding out about it for ourselves!