Easter(复活节) is still a great day for worship, randy in baskets and running around the yard finding eggs, but every year it gets quite a bit worse for bunnies.

    And no, not because the kids like to pull their ears. The culprit is climate change, and some researchers found that rising temperatures arc having harmful effects on at least five species of rabbit in the US.

    Take the Lower Keys March rabbit, for instance. An endangered species that lives in the Lower Florida Keys, this species of cottontail is a great swimmer — it lives on the islands! — but it is already severely affected by development and now by rising levels. According to the Center for Biological Diversity, an ocean level rise of only 0. 6 meters will send these guys jumping to higher ground and a 0.9-meter rise would wipe out their habitat (栖息地) completely.

    The snowshoe hare, on the other hand, has a color issue. Most of these rabbits change their fur color from white in the wintertime to brown in the summer, each designed to give them better cover from predators（捕食者）. As the number of days with snow decreases all across the country, however, more and more bunnies arc being left in white fur during brown dirt days of both fall and spring, making them an easier mark for predators. Researchers know that the color change is controlled by the number of hours of sunlight, but whether the rabbit will be able to adapt quick enough to survive is a big question. The National Wildlife Federation has reported that hunters have noticed their numbers are already markedly down.

    American pikas or rock rabbits, a relative of rabbits and hares, might be the firs' of these species to go extinct due to climate change. About 7-8 inches long, pikas live high in the cool, damp mountains west of the Rocky Mountains. As global temperatures rise, they would naturally migrate (迁徙) to higher ground — but they already occupy the mountaintops. They can't go any higher. The National Wildlife Federation reports that they might not be able to stand the new temperatures as their habitat beats up.

    The volcano rabbit has the same problem. These rabbits live on the slopes of volcanoes in Mexico, and recent studies have shown that the lower range of their habitat has already shifted upward about 700 meters, but there are not suitable plants for them to move higher, so they are stuck in the middle. Scientists are concerned about their populations.

    Native to the US, pygmy rabbits weigh less than 1 pound and live in the American West. They are believed to be the smallest rabbits in the world. Their habitats have been destroyed by development. Several populations, such as the Columbia Basin pygmy, almost went extinct and were saved by zoo breeding programs. Pygmy rabbits also rely on winter cover by digging tunnels through the snow to escape predators, but lesser snowfall is leaving them exposed.

All of this gives new meaning to dressing up in a giant bunny costume this Easter.

    Dreams are expressions of thoughts, feelings and events that pass through our minds while we are sleeping. People dream about one to two hours each night. We may have four to seven dreams in one night. Everybody dreams. But only some people can remember their dreams.

    The word “dream” comes from an old word in English that means “joy” and “music”. We dream in color. Our dreams often include all the senses—smells, sounds, sights, tastes and things we touch. Sometimes we dream the same dream over and over again. These repeated dreams are often unpleasant. They are bad dreams that sometimes frighten us.

    Artists, writers and scientists sometimes say they get ideas from dreams. For example, the singer Paul McCartney of the Beatles said he awakened one day with the music for the song Yesterday in his head. The writer Mary Shelley said she had a very strong dream about a scientist using a machine to make an animal come alive. When she woke up, she began to write her book about a scientist named Frankenstein who made a frightening monster(妖怪).

    In China, people believed that dreams were a way to visit with family members who had died. Some native American tribes(部落) and Mexican civilizations believe dreams are a different world we visit when we sleep.

    For many students who desire to move around but not far freely, one of the most common vehicles is the bicycle. For such a seemingly simple invention, its story is not that simple.

    Most historians trace its origin back to 1817, when a German nobleman named Karl von Drais invented a wooden, two­wheeled machine that riders moved forward with their feet. His invention became popular in both England and France, where it eventually became known as the velocipede. Unfortunately, it was eventually banned as a danger to pedestrians and was rarely seen after the early 1820s.

    Things were quiet for several decades until the bicycle development took off in the 1860s. An important milestone happened in Paris in 1863 when pedals were added to the front axle (轴). This occurred in Pierre Michaux's workshop, but it's unclear whether he or his employee, Pierre Lallement, should be given credit for the innovation Lallement moved to the United States, where he obtained a patent for "improvements in velocipedes" in 1866. These new machines proved to be popular, and the name "bicycle" had come into use by 1869. However, many people referred to them as "bone shakers", which described their clunky ride due to a heavy wooden frame and steel wheels.

    In the 1870s, "high wheelers" or "penny­farthings" became popular. However, with a large front wheel and a much smaller rear wheel, they could be dangerous, if riders had to stop suddenly, as they would "take a header" when their momentum (动量) carried them over the front wheel onto their heads. Eventually, English inventor John Kemp Starley designed a "safety bicycle" with two same small wheels, a chain drive, and a set of gears. With tires added in and brake systems bettered in the following decades, bicycle production had skyrocketed to over one million bicycles by 1899.

    Mass production of bicycles increased their popularity greatly, since they became affordable for the average person. Over the course of the 20th century, manufacturers continued to improve the features and design of bicycles as new technologies appeared.

    "What kind of rubbish are you?" This question might normally cause anger, but in Shanghai it has aroused complaints over the past week. On July 1st the city introduced strict rubbish-sorting regulations as a model for the country. Residents must divide their waste into four separate categories and throw it into specific public bins. They must do so at scheduled times, when monitors are present to ask the nature of one's rubbish, otherwise, they will face fines or worse.

    Shanghai authorities are responding to an obvious environmental problem. It generates 9 million tonnes of garbage a year, more than London's annual output and rises quickly. But like other cities in China, it lacks a recycling system. Individuals who fail to recycle could be fined up to 200 yuan (＄29). For repeat, the city can add black marks to their credit records, making it harder for them to obtain bank loans or even buy train tickets.

    Some object to this. Peng Feng of the Shanghai Academy of Social Sciences warns that the use of the credit system is overkill, raising a risk that officials will abuse their power. But others say a tough campaign is necessary. "Slowly people will get used to it," says Li Changjun of Fudan University.

    Many residents support recycling but are frustrated by the details. Rubbish must be divided according to whether it is food, recyclable, dry or harmful, the distinctions among which can be confusing, though there are apps to help work it out. Some have complained about the rules surrounding food waste. They must put it straight in the requisite public bin, forcing them to tear open plastic bags and toss (撕开) it by hand. Most upsetting are the short windows for throwing rubbish, typically a couple of hours, morning and evening. Along with the monitors at the bins, this means that people go at around the same time and can keep an eye on what is being thrown out; no one wants to look bad.

A new study examined scientists' peer reviews, or researchers' official statements on others' work, across multiple AI-related conferences. At one such conference, those peer reviews used the word "meticulous" — a buzzword often associated with generative A.I., like ChatGPT — almost 3,400 percent more than the previous year. Other major conferences showed similar patterns. In other words, many researchers were handing, at least, parts of their peer review over to A.I.

What's going on in science is a slice of a much bigger problem. Any viral post on social media now almost certainly includes A.I.-generated elements. There are synthetic videos for children on YouTube, like music videos about parrots where the birds have eyes within eyes, singing in an unnatural voice. The narratives make no sense, and characters appear and disappear randomly.

As a neuroscientist, this worries me. Isn't it possible that human culture contains within it cognitive micronutrients — things like reasonable sentences, narrations and character continuity — that developing brains need? Einstein supposedly said: "If you want your children to be intelligent, read them fairy tales. If you want them to be very intelligent, read them more fairy tales." But what happens when a child is consuming mostly A.I.-generated waste? We find ourselves in the middle of a vast developmental experiment.

A.I.'s cultural pollution is driven by a desire to fill the Internet's appetite for content as cheaply as possible, which in turn pollutes our culture. And despite public appeals to act against it, A.I. companies are dragging their feet because it goes against the industry's bottom line to have detectable products, which they fear might weaken the model's performance, although there is no current evidence.

To deal with this general refusal to act, we need a Clean Internet Act. Perhaps the simplest solution would be to force built-in watermarking to A.I. generated outputs, like patterns not easily removable. Just as the 20th century required action to protect the shared environment, the 21st century is going to require actions to protect a different but equally critical resource: our shared human culture.

 Imagine this. You need an image of a balloon for a work presentation and turn to an AI text- to- image generator, like Midjourney or DALL-E, to create a suitable image. You enter the prompt(提示词)" red balloon against a blue sky" but the generator returns an image of an egg instead.

What's going on? The generator you' re using may have been" poisoned". What does this mean?

Text- to- image generators work by being trained on large databasets that include millions or billions of images. Some of the generators have been trained by indiscriminately (任意地) scraping online images, many of which may be under copyright. This has led to many copyright infringement(侵害) cases where artists have accused big tech companies of stealing and profiting from their work.

 This is also where the idea of" poison" comes in. Researchers who want to empower individual artists have recently created a tool named" Nightshade" to fight back against unauthorised image scraping. The tool works by slightly changing an image's pixels(像素) in a way that confuses the computer vision system but leaves the image unchanged to a human's eyes. If an organization then scrapes one of these images to train a future AI model, its data pool becomes" poisoned". This can result in mistaken learning, which makes the generator return unintended results. As in our earlier example, a balloon might become an egg.

 The higher the number of" poisoned" images in the training data, the greater the impact. Because of how generative AI works, the damage from" poisoned" images also affects related prompt keywords. For example, if a" poisoned" image of a Picasso work is used in training data, prompt results for masterpieces from other artists can also be affected.

 Possibly, tools like Nightshade can be abused by some users to intentionally upload" poisoned" images in order to confuse AI generators. But the Nightshade's developer hopes the tool will make big tech companies more respectful of copyright. It does challenge a common belief among computer scientists that data found online can be used for any purpose they see fit.

 Human rights activists, for example, have been concerned for some time about the indiscriminate use of machine vision in wider society. This concern is particularly serious concerning facial recognition. There is a clear connect ion between facial recognition cases and data poisoning, as both relate to larger questions around technological governance. It may be better to see data poisoning as an innovative(创新的) solution to the denial of some fundamental human rights.