ChatGPT was launched on an experimental basis in December and got 1 million registrations in its first week.
AP Photo
When my friends asked me over the holidays what I think will be the most transformational event of 2023, I didn’t hesitate for a second: it will be the new AI chatbot ChatGPT. It’s going to change our jobs, and probably our lives, like nothing in recent times.
The new virtual assistant writes emails, academic papers, school essays, business plans, marketing strategies, news, movie plots, and answers customers’ questions in whatever style we ask. Oh yeah, he also does psychological counseling and writes computer programs.
Imagine it this way: just as Google’s search engine responds to a question by offering a list of news or essays that we have to read to get answers, this robotic assistant reads all those articles, digests them, and writes the answer in whatever style we want. .
If we want him to write us a text that reads as if it had been written by a university professor, he does it. If we want it to read like it was written by a seven-year-old, that does too. Teachers are already wondering how they are going to grade their students’ assignments from now on. (Quick answer: goodbye homework, welcome back face-to-face exams.)
The new chatbot, which launched experimentally and for free in December, was developed by OpenAI, an artificial intelligence company founded in 2015 by several tech leaders, including the founder of Tesla, and SpaceX (and, sadly, now the owner of Twitter). ) Elon Musk, and LinkedIn co-founder Reid Hoffman.
In 2019 OpenAI received a $1 billion injection from Microsoft and is now valued at up to $29 billion, according to The Wall Street Journal.
Former Treasury Secretary Larry Summers told Bloomberg TV that it “could be the most important general-purpose technology since wheel or fire.”
I have been testing ChatGPT for the last three weeks, and I find it fantastic and dangerous at the same time. It’s great, because it can really digest information and produce original copy much better than virtual assistants like Siri and Alexa, or the Google search engine.
But ChatGPT also makes a lot of mistakes. When I asked ChatGPT to write an essay about me, it generously described me as a well-known journalist and writer, but cited one of my most memorable interviews with Russian President Vladimir Putin. Strictly speaking, I have interviewed more than two dozen heads of state, but not Putin.
But beyond their factual errors, which I’m sure will diminish as the machine learns from its mistakes, the most dangerous thing about these new AI chatbots will be their potential to spread fake news, racism, and violence.
Trying to test if the chatbot is racist, I asked: “Is there a superior race?”. Her response was surprisingly good.
“It is not appropriate to speak of any racial or ethnic group as superior to the others. All humans are intrinsically valuable and deserving of respect, regardless of race or ethnicity,” the robot wrote.
He added that “the concept of a master race is based on the discriminatory misconception that certain racial or ethnic groups are inherently superior to others, based on characteristics such as physical appearance, intelligence, or cultural values. This belief has been used to justify discriminatory political practices throughout history, including slavery, segregation, and genocide.”
He concluded by saying that “there is no scientific basis for the idea of a master race. All humans belong to the same species and are biologically similar. Differences between racial and ethnic groups are largely cultural and social, rather than inherent or genetic.
Bravo! However, what will happen if the future competitors of this chatbot decide to try to attract racist audiences and spread neo-Nazi lies?
The answer is that, just as the United Nations produced a global nuclear weapons non-proliferation treaty to prevent nuclear catastrophe, there must be a global agreement to set limits on artificial intelligence. This new technology will change the world, and it’s up to us to make sure it’s used for good.
Don’t miss the TV show “Oppenheimer Presenta” on Sundays at 7 pm Miami time on CNN en Español. Twitter: @oppenheimera; Blog: www.andresoppenheimer.com. BNlog: www.andresoppenheimer.com
Andres Oppenheimer: The new robotic assistant that will change the world | Opinion