Thousands of years ago doctors used rocks for operations (手术). Later they used sharp bones or horns, metal knives and more recently, rubber and plastic. In the 1960s, a new tool was developed. First of all, it was of great practical use to the industry. Then it was also used to improve the art and science of operations.

The tool is the laser (激光), and it is being used by more and more doctors all over the world. As we all know, light is hot, and any light will give warmth. But light is usually spread out over a wide area. The light in a laser, however, is concentrated. This means that a light with no more power becomes much stronger when it is concentrated than that produced by an ordinary electric light.

Experiments showed researchers that different energy produces special effect on certain living cells (细胞). It is now possible for eye doctors to operate on the back of human eye without harming the front of the eye, simply by passing a laser right through the eye-ball. Operations once made patients exhausted and in need of long period of recovery (恢复) time. But now they make them feel relaxed and comfortable.

The rapid development of laser techniques in the past ten years has made it clear that the future is likely to be very exciting. Perhaps some cancers will be cured (治疗) with laser in a way. It will make operations not only safer but also more effective.

What is the future of AI (人工智能)? Will it be a good thing? We asked four students what they think.

A 2, 000-year mystery (谜) has moved one step closer to being solved. AI software developed by researchers has successfully deciphered (辨读) the text from a burnt scroll (卷轴) found in the remains of an ancient Roman building in Italy. This could help us better understand the ancient world.

In the centuries after the scrolls were discovered, many people have tried to open them, but ended up destroying some of them. Brent Seales, a computer scientist, has been trying to read these hidden texts for nearly 20 years. His team developed software to read the surface of scrolls using 3D CT images (图像). As the results were not satisfactory, they wondered whether machine-learning models might be trained to read the scrolls But making sense of all the information was a huge task for his small team.

Luckily, in March 2023, the Vesuvius Challenge was introduced, offering a large prize to researchers worldwide. Researchers were asked to decipher four passages, of at least 140 characters each, before the end of the year.

The key to the challenge’s success was the mix of competition and cooperation (合作), with Seales doing his part by sharing his software and images online. Smaller prizes were presented along the way to encourage progress, with the winning machine- learning algorithm (算法) shared at each stage to “level up” the community, so competitors could build on each other’s progress.

The turning point came in the middle of last year, when physicist Casey Handmer noticed something that seemed to form the shapes of Greek letters—he called it “crackle”. In late October, an American university student studying computer science used the crackle to train a machine-learning algorithm, uncovering the word porphyras, “purple”. This won him the prize for discovering the first letters. An Egyptian student, who followed with even clearer images of the text, came second. At last, the two young researchers together with a Swiss student worked out over 2, 000 characters from one scroll, sharing the $700, 000 prize.

The text of the already-opened scrolls is related to Greek philosophy (哲学), and researchers are likely to find more of the same in the scrolls. Also, the machine-learning techniques pioneered by researchers can now be used to study other types of hidden text. The progress suggested the future decipherment of the hundreds of scrolls that had been given up on before. The real gains are still ahead of us.