1 . Vast numbers of copyrighted books appear to have been memorized by ChatGPT and its successor GPT-4, posing questions about the legality of how these large language models (LLMs) are created.

Both artificial intelligences were developed by private firm OpenAI and trained on huge amounts of data, but which texts make up this training data is unknown. To find out more, David Bamman at the University of California, Berkeley, and his colleagues looked at whether the AIs were able to fill in missing details from a selection of almost 600 fiction books, drawn from sources such as nominees (被提名者) for the Pulitzer prize, and The New York Times’s bestsellers lists over the same time period.

The team picked 100 passages from each book that contained a single, named character. The researchers then blanked out the name and asked the AI to fill it in. This task was designed to expose if the AIs could return the exact right answer. “It really requires knowledge of the underlying material in order to be able to get the name right,” says Bamman.

Both AIs completed the task with high accuracy — as much as 98 percent for passages from Lewis Carroll’s 1865 book Alice’s Adventures in Wonderland — which is out of copyright — and 76 percent for J.K. Rowling’s Harry Potter and the Philosopher’s Stone, which is not. The researchers say this suggests the AIs were trained on significant proportions of both books.

These AIs don’t produce an exact duplicate of a text in the same way as a photocopier, which is a clearer example of copyright infringement. “ChatGPT can recite parts of a book because it has seen it thousands of times,” says Andres Guadamuz at the University of Sussex, UK. “The model consists of statistical frequency of words. It’s not reproduction in the copyright sense.”

“The use of copyright works without permission in training data sets for large language or image models has already emerged as one of the most pressing legal challenges to this novel industry,” says Lilian Edwards at Newcastle University, UK.

Bamman says that, ultimately, the legal system in each country will have to determine whether LLMs are infringing (侵犯) copyrights. “I think that’s an open question that a lot of court cases are going to decide for us in the coming months,” he says.

Regulation is also likely to play a key role: the European Union’s Artificial Intelligence Act, which has been two years in the making, will include a requirement that companies making generative AI tools need to disclose any copyrighted material used to train their models. That was a late change, added to the draft law in April, according to Reuter.

10 . Try Hard, but Not That Hard.

So many of us were raised with the belief of hard work and max effort, taught that what we put in was what we got out. Now, some coaches and corporate leaders have a new message.

Trying to run at top speed will actually lead to slower running times. Lifting heavy weights until you absolutely can’t any more won’t spark more muscle gain than stopping a little sooner. The trick — be it in exercise, or anything — is to try for 85%. Aiming for perfection often makes us feel awful, burns us out and backfires. Instead, count the fact that you hit eight out of 10 of your targets this quarter as a win.

“I already messed it up,” Sherri Phillips would regret after missing one of her daily personal goals. Last year, the COO of a Manhattan photography business began tracking metrics like her sleep quality and exercise time. It was only after she switched to aiming for 85% success over a week that she stuck with her efforts, instead of giving up when she missed a mark. 

Dialing in on the sweet spot of 85% can help us grow. In a 2019 paper, researchers used machine learning to try to find the ideal difficulty level to learn new things. The neural network they created, meant to mimic the human brain, learned best when it was faced with queries (疑问) set to 85% difficulty, meaning it got questions right 85% of the time. If a task is too hard, humans get demotivated, says Bob Wilson, an author of the study. “If you never make any errors, you’re 100% accurate, well, you can’t learn from the mistakes.”

Ron Shaich, a founder and former CEO of restaurant chain Panera, is skeptical of people who hit 100% on sales targets. He wonders if the goals are too low. They should be ambitious enough that you won’t always get there, he says. Now an investor, board member and author of a coming business book that stresses 80% equals success, Shaich is convinced most companies don’t even hit that number.

Years ago, as a consultant, Grace Ueng learned the “80-20 rule.” The idea was to stop once you were 80% complete on a project, she says. Ueng recently took up piano. She practiced for long hours and still grimaced when she performed for her music group. Then she started tackling small chunks of a piece instead of running through the whole thing again and again. Before a recent performance, she read a book and went to church instead of putting in extra hours at the piano. When it was time to perform, she played well—and actually enjoyed it. “You have to have the wisdom,” she says, “to know when to stop.”