The social network’s newest star is a robot: a program called ChatGPT Try to answer the questions like a person.
Since its launch last week, many people have shared what the bot can do. New York magazine journalists told it to write what turned out to be a “pretty good” story. Other users had it to write a solid academic essay on theory of nationalisma tragic but fictional history Ohio-Indiana War and some jokes. It tells me a story about an artificial intelligence program called Assistant that was originally set up to answer questions but quickly led to a new world order that led humanity to “an age of greatness.” new era of peace and prosperity”.
What’s remarkable about these examples is their quality: People can write them. And bots aren’t even the best; OpenAI, the company behind ChatGPT, is said to be working on a better model that could be released next year.
Kevin Roose, my colleague in charge of the technology sector, told me: “A lot of the promised benefits of AI are gone in five years. “ChatGPT is the moment when a technology everyone has heard of has finally become a reality for them.”
In today’s newsletter, I’ll explain the potential benefits of artificial intelligence and why some experts fear it could be dangerous.
The advantage of artificial intelligence is that it can complete tasks faster and more efficiently than any human can. The possibilities are up to the imagination: self-driving and even self-repairing cars, risk-free surgery, instant personalized therapeutic bots, etc.
Technology is not yet available. But it has evolved in recent years through what is known as machine learning, in which bots dig through data to learn how to perform tasks. In the case of ChatGPT, it reads a lot. And, with some guidance from its creators, it learned to write coherently — or at least statistically predict what good text should look like.
There are already clear benefits to this nascent technology. It can help with research and writing essays and articles. ChatGPT can also help with coding programs, automating challenges that would normally take people hours.
Another example comes from another program, Consensus. This bot combines millions of scientific articles to find the most relevant ones for a given search and shares their key findings. A task that would take a journalist like me a few days or weeks is done in minutes.
These are the early days. ChatGPT still makes mistakes, like telling a user that the only country whose name starts and ends with the same letter is Chad. But it is growing very fast. Even some skeptics believe that AI in general use could reach human intelligence within decades.
Despite the potential benefits, experts still worry about possible flaws with AI
First, such a level of automation could take people’s jobs away. This concern has appeared with automated technology before. But there is a difference between a machine that can help assemble car parts and a robot that can think better than a human. If AI reaches the heights some researchers hope for, it will be able to do almost everything a human can do, but better.
Some experts point to the inherent risks. A survey asked machine learning researchers about the potential effects of AI. Nearly half said there was a 10% or more chance that the outcome would be “extremely bad (e.g. human extinction).” These are the people who say their life’s work can destroy humanity.
That sounds like science fiction. But the risk is real, experts warn. “We may fail to train AI systems to do what we want,” said Ajeya Cotra, AI research analyst at Open Philanthropy. “We can inadvertently train them to pursue goals that are at odds with humans.”
To take a hypothetical example, from Kelsey Piper at Vox: A program is asked to estimate a number. It turns out that the best way to do this is to use more of the world’s computing power. Then the program realized that humans had used that computing power. So it destroys all humans so its numbers can be estimated without hindrance.
If that sounds unreasonable, consider that existing bots already operate in a way that their creators did not intend. ChatGPT users have come up with alternatives to get it to say racist and sexist things, despite OpenAI’s efforts to suppress such responses.
As AI researchers admit, the problem is that no one fully understands how this technology works, making it difficult to control all possible behaviors and risks. However, it is already available for public use.
Life lived: In the 1940s and 1950s, singer Joyce Bryant released discs so sexy that some radio stations didn’t play them. She then left the show business for a while to do missionary work. She passed away at the age of 95.
SPORTS NEWS FROM Ethics
One big increase: Aaron Judge is sticking with the Yankees on a nine-year, $360 million deal after turning down a $213.5 million offer last season. Here’s what we know about negotiations.
Padres: After missing both Judge and the new Phillies shortcut Trea Turner, San Diego and the Red Sox longtime shortcut Xander Bogaerts agreed to an 11-year, $280 million contract last night.
Candidates: France and England, Argentina and Brazil, Messi and Ronaldo – quarterfinals everything but an easy road to the trophy.
Pan-Arab Solidarity: Millions of Moroccans celebrate their country’s victory over Spain, Arabs and Africans all over the world participated.
Loyal fans: A football-loving swath of India – otherwise a cricket-mad nation – has sided between Brazil and Argentina and put on giant displays of devotion.
Best Art of the Year
Museums have managed to bring visitors back with blockbuster shows this year, and the best exhibits left our reviewer ‘suffocated’. Whitney’s biennial reflecting on three years of national history is heart-wrenching, while “Matisse: The Red Workshop,” at the Museum of Modern Art, is one of the “greatest exhibitions of all time”, with only 10 items.
But some of our critics’ favorite shows have managed to attract new audiences. “Art and Racial Issues,” an exhibition by American-Pop artist Robert Colescott at the New Museum, is a harrowing revision of history. And “What is the use of Buddhist art?” at Columbia University’s Wallach Art Gallery brought religious objects “to life in a way museums rarely do.”