ChatGPT CEO warns AI bots could develop ‘superhuman persuasion skills’ that result in ‘strange outcomes’
SKILLS that artificial intelligence could develop as it becomes more advanced may reach superhuman levels and become strange, ChatGPT CEO Sam Altman warned.
The statement comes as AI becomes more popular and smarter every day – it learns from itself and its mistakes as more people use it.
One of the possible AI superhuman skills includes being able to persuade people a certain way through facts and data that it collects.
The facts and data don’t even have to necessarily be false – they will just be convincing.
“I expect AI to be capable of superhuman persuasion well before it is superhuman at general intelligence, which may lead to some very strange outcomes,” Altman said on X, formerly known as Twitter, on October 24.
Altman did not clarify exactly what the strange outcomes would be.
But ChatGPT and other AI bots have already proved they are capable of completing a large range of impressive tasks.
In response to Altman’s comment, Christopher Alexander, chief analytics officer of Pioneer Development Group, confirmed in an interview that AI will be able to perfect its workings, per Fox News on Thursday.
Along with that, this may already be happening among humans who have the same persuasive goals.
“Machine learning and pattern recognition will mean that an AI will get very good at identifying what persuasive content works, in what frequency, and at what time,” Alexander said.
“This is already happening with digital advertising. Newer, more sophisticated AI will get better at it.
“Social media already does that and is difficult to outperform.”
There was an interesting response from people in the replies to Altman’s post as well.
One person believes it will be up to the public to decipher the good from the bad with AI.
They said: “A longtime ChatGPT user can tell when it is fabricating things, but for the public, or if you are not familiar with the topic being discussed, there’s indeed a risk of deception.”
Another person believes only humans can become superhuman and that AI can only reach certain heights with advancement.
This person said: “You can make a human superhuman by endowing him with higher capabilities by AI, but you cannot program or emulate the human or grand consciousness in a machine or code.
“However, you can do great feats and achieve insurmountable tasks with AI.”