1 . A Perspective article published in Nature this week says that AI systems pose a further risk: that researchers envision such tools as possessed of superhuman abilities when it comes to objectivity, productivity and understanding complex concepts. Scientists planning to use AI “must evaluate these risks now, while AI applications are still nascent (未成熟的), because they will be much more difficult to address if AI tools become deeply embedded in the research pipeline”, write co-authors Lisa Messeri and Molly Crockett.

In this article, Messeri and Crockett put together a picture of the ways in which scientists see AI systems as enhancing human capabilities. In one ‘vision’, which they call AI as Oracle, researchers see AI tools as able to tirelessly read and digest scientific papers, and so survey the scientific literature more exhaustively than people can. In both Oracle and another vision, called AI as Arbiter, systems are perceived as evaluating scientific findings more objectively than do people, because they are less likely to cherry-pick the literature to support a desired hypothesis or to show favouritism in peer review. In a third vision, AI as Quant, AI tools seem to surpass the limits of the human mind in analyzing vast and complex data. In the fourth, AI as Surrogate, AI tools simulate (模拟) data that are too difficult or complex to obtain.

Informed by anthropology and cognitive science, Messeri and Crockett predict risks that arise from these visions. One is the illusion of explanatory depth, in which people relying on another person — or, in this case, an algorithm — for knowledge have a tendency to mistake that knowledge for their own and think their understanding is deeper than it actually is.

Another risk is that research becomes skewed towards studying the kinds of thing that AI systems can test — the researchers call this the illusion of exploratory breadth. For example, in social science, the vision of AI as Surrogate could encourage experiments involving human behaviors that can be simulated by an AI — and discourage those on behaviors that cannot, such as anything that requires being embodied physically.

There’s also the illusion of objectivity, in which researchers see AI systems as representing all possible viewpoints or not having a viewpoint. In fact, these tools reflect only the viewpoints found in the data they have been trained on, and are known to adopt the biases (偏见) found in those data. “There’s a risk that we forget that there are certain questions we just can’t answer about human beings using AI tools,” says Crockett. The illusion of objectivity is particularly worrying given the benefits of including diverse viewpoints in research.

If you’re a scientist planning to use AI, you can reduce these dangers through a number of strategies. One is to map your proposed use to one of the visions, and consider which traps you are most likely to fall into. Another approach is to be deliberate about how you use AI. Deploying AI tools to save time on something your team already has expertise in is less risky than using them to provide expertise you just don’t have, says Crockett.

1. According to Paragraph 2, which might scientists agree with?

A．Human intelligence can’t be overshadowed in analyzing data.

B．Prejudice has been ruled out by AI tools in scientific evaluation.

C．Studies with desired hypothesis are usually favored in peer review.

D．AI can conduct a relatively thorough overview of the prior research.

2. What does the underlined word “skewed” in Paragraph 4 probably mean?

A．Declining.

B．Partial.

C．Dynamic.

D．Irrelevant.

3. What can we learn from this passage?

A．Relying on AI to fill gaps in the team’s knowledge is much preferable.

B．AI isn’t limited to the outlook and biases present within its training data.

C．Instant actions must be taken to rate AI’s risks before it is fully integrated into research.

D．Users counting heavily on AI tend to underestimate their insight of a concept.

4. Which would be the best title for the passage?

A．A Call for Considered Use of AI in Science

B．The Unseen Biases of AI in Social Science

C．AI’s Role in Enhancing Human Expertise

D．How AI Is Shaping Scientific Discovery

2 . All over the world, formal education supplies the economy with workers who will increase productivity to fuel the economic machine. But this machine now threatens our very survival. If the entire world reaches the levels of consumption seen in high-income countries today, we’ll need multiple planet Earths to supply the resources. The absurd (荒谬的) idea of infinite growth within a finite territory is at the heart of our economic system.

To keep this machine running, formal education generates ever more efficient “human capital”. Increasing productivity metrics (指标) rather than the individuality of students drives our civilization’s approach to schooling our young people. Whereas the Sustainable Development Goals call for turning education into a force for sustainability, the opposite is often true: The ways Western societies have come to think about education undermine our ability to deal with the environmental crisis. To get through this crisis, we need to cultivate our imagination, not undermine it.

Growing up, none of my schooling fostered my ability to imagine a world different from what I saw around me. Besides, I realize the suppression (抑制) of children’s imagination doesn’t take place only in underresourced communities, but in “elite” institutions that tout “critical thinking”. Schools want to see their graduates succeed, and success is too often about maintaining current structures — not about reimagining their foundations.

Essentially, our education systems shape children in the image of artificial intelligence. The perfect “worker”, AI, continually improves its own productivity but doesn’t challenge the larger structures within which it operates. It is one of the great paradoxes of our time that we invest so much into building supercomputers while marginalizing the imaginative potential of millions of human brains. What’s more, we even put our hope in solving the environmental crisis on AI. But AI, like our other technologies, can only treat the symptoms of the environmental crisis, not the causes

Throughout history, achievers of great change have relied on their imaginations to address fundamental flaws in society. In my country of birth, communists kept their dreams of democracy alive for decades by imagining different futures. In South Africa, Nelson Mandela’s followers had to be radical (激进的) in their imagination to create a vision of a fairer society. Imagining democracy when living under a totalitarian regime (极权主义政权) isn’t that different from imagining degrowth when living in a world of infinite growth.

The kind of intelligence that Nelson Mandela and such possessed was not artificial. The ability to reimagine the future and disrupt the current situation remains a distinctly human quality. Unlike AI, children are naturally imaginative and question the premises of society. In my research, I have observed that younger children are often the most radical in imagining different futures.

As long as our imagination is curtailed, ideas like degrowth sound utopian (乌托邦的) to many, Cultivating imagination means learning from history’s disrupters who made the   impossible possible. Instead of dismissing “childish” ideas about the world’s future, it means seeing inspiration in children’s imaginations. In an education system that celebrates imagination, arts and creativity are as important as math and science. Idealism coexists with pragmatism. The environmental crisis is not a crisis of technology or science, it is a crisis of imagination. If we let children be our guides, we might be able to imagine our way to survival.

1. The author hopes education can play its role in ________.

A．developing human resources	B．promoting sustainable development
C．increasing productivity	D．maximizing economic growth

2. From the passage, we can learn that _________.

A．imagination isn’t well developed in schools

B．AI helps exploit the potential of human brains

C．graduates’ success changes the social structures

D．AI can address the root cause of economic crisis

3. Nelson Mandela is mentioned as _________.

A．a success in building a fair society

B．a leader who had a great influence

C．a pioneer possessing the quality of AI

D．an inspiration to solve social problems

4. According to the passage, the author may agree that _________.

A．children’s imagination ensures human’s survival

B．the environmental crisis results from technology

C．imagination can help solve environmental crisis

D．the “childish” ideas will ruin the world’s future

3 . 2024 Harvard Winter Program in Shanghai, China

The city of Shanghai has brought together diverse peoples, cultures, and economies from various parts of China and around the world. Blending cultural studies with social sciences, 2024 Harvard Winter Program in Shanghai, China will help you understand the city.

Program dates: November 8, 2024 to December 21, 2024

Apply by: August 25, 2024

Program Structure

This six-week program includes two courses on Shanghai’s cultural history and economic development experiences. You start each week with two days of seminars (研讨会) designed to let you see Shanghai differently. On Wednesdays and Thursdays, the group moves to a field site for a hands-on research practicum (实践操作课) in companies, government offices, and communities. Over the weekend, you try your own hands at exploring the city and its economy, working on creative projects and/or research papers, with a flexible final project format that accommodates student priorities.

The program counts as two semester-long courses of degree credit (学分), 4 credits each. Both program courses can count towards East Asian Studies credit.

Accommodations

You will live in an on-campus hotel in one of China’s most distinguished academic institutions. Hotel services include wi-fi, daily housekeeping, safe for storage of valuables, cafeteria and gym. Apart from all the practical conveniences of a student district, the campus also offers library access and will be an ideal base for your creative and research projects.

Application

To apply, you must:

•Be at least 18 years old

•Have completed at least one year of college

•Be in good academic standing

Students enrolled at any university are welcome to apply.

The Winter 2024 application will be available in early July. Each program has unique requirements included in the online application. Beginning your application early is the best way to ensure that you have sufficient time to review and complete the application requirements by the deadline.

If you have questions about the application, please contact the Harvard Study Abroad Office by email at winterabroad@winter.harvard.edu.

1. If you take 2024 Harvard Winter Program, you _________.

A．will live in a hotel near Harvard campus	B．will spend twelve days in total on seminars
C．can get 4 credits towards East Asian Studies	D．must complete a research paper for a project

2. If you want to apply for this program, you must _________.

A．be a high school senior

B．have received hands-on training

C．have work experience in companies

D．finish no less than one year of college

3. What is the main purpose of the passage?

A．To stress the value of taking a winter program.

B．To attract more people to apply for a winter program.

C．To share a memorable experience at Harvard University.

D．To introduce requirements for a final project at Harvard University.

4 . How objective are you? How often are you over-reacting to situations, taking things personally or judging people unfairly? We all do this at some time or another.

    1     Under stress, most of us draw solely on our past experiences and biases, which cloud our ability to see things clearly and make sound judgments. Moreover, our desire to succeed is often so strong, that, under these intense circumstances, our limiting mental models tend to get in our way, creating even more stress and causing us to react in ways that we may regret.

Our ability to be objective depends on our willingness to question our mental models, the lens through which we perceive, interpret and respond to our world. If our mental models are incorrect, then our understanding of what is going on and our response to it, are often incorrect.     2    

The good news is that with practice, we can interrupt our automatic reactions, and choose a different response. Each time we do this, we are re-wiring our neural network by creating new pathways based on new models: new ways of perceiving and responding to our world.     3    

One of the most powerful mental model transformation catalysts (催化剂) is new knowledge or logic that challenge old ways of thinking.     4     They’ve usually been with us a while, so we tend to trust them. For most of us, we have never been taught about mental models and how to evaluate them to determine if they are helpful or harmful.

To transform unproductive mental models, we must change our mind! We have to decide, through our own logic and reason, whether our way of seeing the world is no longer valid for us.     5     It is in the wake of this new knowledge that transformation takes place. At that moment, a set of new connections is created in the brain. It’s these new connections that have the potential to enhance our mental resources to help us transform limiting mental models.

A．We can actually learn to think smarter!

B．An objective leader must judge and treat people fairly.

C．This requires that we be open to new knowledge and reasoning.

D．The key is to accept a problem as it occurs and not take it personally.

E．This is why we sometimes misjudge situations and take things personally.

F．As we have seen, mental models are deep — rooted beliefs we tend to hold onto.

G．The challenge is that when we are under pressure, we tend to be less objective.

5 . For years, scientists have been trying to figure out whether “brain workouts” such as puzzles (字谜) and online cognitive games could strengthen our minds and slow the process of aging.

Now, a study has found that regularly attempting a crossword may help slow decline in some people with mild cognitive impairment (障碍), an early stage of faltering memory that can sometimes progress to dementia (痴呆症).

While the study didn't investigate whether crosswords benefit younger adults who are not dealing with cognitive decline, it suggests that keeping your mind active as you age may benefit your brain. And the research offers hope to those diagnosed with mild cognitive impairment that they may be able to stave off further declines in the memory, language problems and decision-making that are the hallmark of the condition.

The American Academy of Neurology estimates that mild cognitive impairment affects about 8 percent of people ages 65 to 69; 10 percent of people ages 70 to 74; 15 percent of people ages 75 to 79; 25 percent of those ages 80 to 84; and about 37 percent of people 85 and older.

The research, which was funded by the National Institute on Aging, recruited 107 adults ages 55 to 95 with mild cognitive impairment. For 12 weeks, they were all asked to play one of two types of games, four times a week—spending either 30 minutes on Lumosity, a popular cognitive training platform, or 30 minutes attempting a digital crossword. After 12 weeks, the participants were reevaluated and given "booster" doses of game play six more times during the 78-week experiment.

By the end of the study, participants were given standard assessments used to measure cognitive decline, and friends and family reported on their day-to-day functioning. MRI scans also were used to measure brain volume changes.

Researchers found that in key measurements—cognitive decline scores, functional skills and brain volume changes—the regular crossword players fared better than the game players.

The finding surprised the scientists behind the study who had expected that challenging web-based brain games, which were specifically designed to boost cognitive function, would offer the most benefit.

“Our study shows that in people with mild cognitive impairment, crossword puzzles beat the computerized games on multiple metrics, ” said Murali Doraiswamy, a co-author of the study. “So, if you have mild cognitive impairment, which is different from normal aging, then the recommendation would be to keep your brain active with crossword puzzles.”

People with higher degrees of cognitive impairment appeared to benefit the most from doing the crossword, which was designed to be a moderately difficult puzzle comparable to the Thursday game.

The study has limitations. Some of the participants may have just been more familiar with crosswords and that's why they responded better to the puzzles than to Lumosity's computer games. More years of follow-up also are needed to determine whether interventions such as crossword puzzles can “truly prevent dementia,” Doraiswamy said.

1. What does the underlined phrase "stave off" in Paragraph 3 probably mean?

A．Promote.

B．Prevent.

C．Experience.

D．Cause.

2. Paragraph 4 is written to_________.

A．show a concerning condition	B．compare different groups
C．present real-life examples	D．prove an assumption

3. What can we learn from the passage?

A．One's degree of cognitive impairment depends on how old he is.

B．Solving crossword puzzles is a recommendation to elderly people.

C．The efficiency of puzzles in preventing dementia has been proved.

D．Crosswords' effects are linked to how serious cognitive impairment is.

4. What makes the findings of the study far from conclusive?

A．The lack of variables.	B．The weaknesses of assessments.
C．The insufficiency of time span.	D．The subjects' preference for games.

6 . What Came First, the Chicken or the Egg?

Chickens are one of the most economically important animals in the world today.     1    , Recent advancements in archaeological methodologies have unveiled that numerous bone discoveries, once believed to be evidence of early chickens, actually belong to wild birds.

Now, in a new publication, an international team of archaeologists, historians, and biomolecular scientists present the earliest clear evidence for the raising of chickens for egg production. Using eggshell fragments collected from 12 archaeological sites covering roughly 1500 years, the researchers show that chickens were widely raised in Central Asia from approximately 400 BCE to 1000 CE and were likely spread along the ancient Silk Road.     2    . It was this trait of productive egg laying, the researchers argue, that made the domestic chicken so attractive to ancient peoples

    3    . They collected tens of thousands of eggshell fragments from sites located along the main Central Asian corridor of the Silk Road. They then used a method of biomolecular analysis called ZooMS to identify the source of the eggs.   The identification of these shell fragments as chickens, and their abundance at each site, led the researchers to an important conclusion.     4     . In the wild, this kind of bird nests once per year and typically lays six eggs per clutch (窝). This is the earliest evidence for the loss of seasonal egg laying yet identified in the archaeological record. Meanwhile, it is an important clue for better understanding the mutualistic (共生的) relationships between humans and animals that resulted in domestication

Taken together, the new study suggests an answer to the age-old riddle of the chicken and the egg.     5    .

A．To reach these conclusions, the team did a lot of work.

B．The abundance of eggshells further suggests that the birds were laying out of season.

C．Yet, the story of how they originated and spread throughout the ancient world remains largely unclear.

D．Research shows chickens were bred throughout southern Central Asia, and were spread along the ancient Silk Road.

E．The birds must have been laying more frequently than their wild ancestor, the red jungle fowl

F．They argue that the loss of seasonal egg laying was the main drive for the spread of domestic chickens across Eurasia and northeast Africa

G．As the evidence in Central Asia suggests, the ability to lay a large number of eggs is what made the chicken the chicken we know today.

7 . 阅读下面短文，根据题目要求用英文回答问题。

What We Really Know About ASMR

If you spend time on YouTube or TikTok, you may have come across videos of someone whispering into a microphone, carefully slicing stacks of slime, or slowly ripping strips of paper. These videos are designed to trigger an autonomous sensory meridian (经络) response, or ASMR, a pleasant feeling that travels from the top of the head down the back of the neck to the upper spine.

Watching videos that trigger ASMR is thought to lead to positive responses throughout the body including enhanced mood, relaxation, and even pain relief. But can you really achieve these benefits just by watching a short video?A systematic review published seeks to answer that question. German researchers combined the data from 54 studies to better help us understand ASMR and its effects.

But first, where did this idea come from?There are references to feelings similar to ASMR in classic literature dating back to the early 1900s. The contemporary definitions of ASMR sprang from a discussion forum on a health-focused website in2007. A YouTuber named Jennifer Allen coined the term in 2010.

Based on the previous studies, the researchers concluded that examples of triggers include whispering, tapping, and natural sounds, and it is important to note that specific triggers of ASMR and the intensity of the response vary widely by individuals. They also explored its benefits. Clear evidence shows that ASMR can reduce stress and anxiety. Data also shows that it can improve short-term memory. In specific cases, participants watching ASMR videos had lower heart rates and blood pressure, physical signals of relaxation. In addition, they found that ASMR triggers brain activity patterns associated with relaxation and stress relief.

Despite the encouraging results, the authors noted that many of the ASMR studies had small sample sizes and a lack of diversity among participants; more research is needed to truly understand the impacts of ASMR.

Luckily, it’s true that watching sensory videos of people brushing their hair, whispering, and making soft repetitive noises can lead to relaxation and stress relief. The best thing about ASMR is that there are minimal side effects. So if you’re feeling anxious or stressed, it might be worth watching a few minutes of ASMR videos to see if they help!

1. What is an autonomous sensory meridian response (ASMR)?
_____________________________________
2. According to the studies, what benefits can ASMR bring us?
_____________________________________
3. Please decide which part is false in the following statement, then underline it and explain why.
>Although triggers of ASMR are different, they may cause reactions of similar intensity among different people.
_____________________________________
4. Besides watching ASMR videos, what else do you do when feeling anxious or stressed?
_____________________________________

8 .

The violence of shaking from an earthquake is usually focused in the direction the quake is traveling. But a boomerang quake may spread the intense shaking across a wider zone. The new study, published today in the journal Nature Geoscience, is a major step toward clearing up the complex physics behind boomerang quakes and understanding their potential damage.

The latest boomerang was recorded near the mid-ocean ridge in the Atlantic in 2016. By examining the position of the epicenter (震中) and the energy released by each shaking phase, Stephen Hicks, an earthquake seismologist at the Imperial College London and first author of the new study, identified what seemed to be two steps of the quake:The quake initially headed eastward, but then turned back west. “This was weird, ”he says. This second leg of the quake moved remarkably quickly, at so-called supershear speeds. The quake unzipped the surface at an estimated 11, 000 miles per hour—fast enough to dart from New York to London in 18. 5 minutes, so fast that it caused the geologic version of a sonic boom.

How often does this happen?A boomerang quake at supershear speeds, as Hicks and his team observed in the Atlantic, may be a fairly rare type. But wider evidence of boomerang quakes is mounting. These back-tracking events have been studied in computer models as well as simulated in lab experiments. “The theory says that it’s there, but it’s quite difficult to see that [in the real world], ”says Hicks. The 2016 earthquake in Kumamoto also seems to have ruptured (破裂) in a similar process. For that event, the initial shake triggered two other quakes, one of which raced backward to partially overlap the initial break.

Understanding when and why these boomerang events happen is vital to grappling with risks earthquakes present. “Studies like this help us understand how past earthquakes ruptured, how future earthquakes may rupture, and how that relates to the potential impact for faults (断层) near populated areas, ”says Hicks.

These boomerangs maybe obscured by conventional approaches used to analyze quakes, which are based on an assumption that a quake rushes in one direction. “Naturally we’re not looking for it, we don’t expect it to exist,” Hicks says. Yet for earthquakes, it seems, complexities might be the norm rather than the exception. As Hicks puts it: “The more and more we look at earthquakes in more detail, of course we see stranger things.”

1. A boomerang earthquake differs from other quakes because ________.

A．it moves at a super-fast speed

B．it triggers several small quakes

C．it releases immeasurable energy

D．it shakes in two opposite directions

2. From the passage, we can learn that _________.

A．the coming of the next boomerang can be predicted

B．boomerangs are too complex to be fully understood

C．boomerang events are commonplace in the real world

D．common methods are reliable in analyzing boomerangs

3. What does the underlined phrase “grappling with” in paragraph 4 probably mean?

A．putting off	B．coping with
C．bringing about	D．fighting against

9 . Student Volunteer Internships 2023-2024

A limited number of unpaid internships (实习) are available in Washington, D. C. within the U. S. Environmental Protection Agency's (EPA) Office of Water. You must be a full-time undergraduate or graduate student. A diversity of candidates within various fields of study are encouraged to apply, including degree candidates in the following subjects:

·Hydrology, aquatic biology, marine science, environmental engineering, geology, meteorology, soil science, etc

·Public health, toxicology, sanitation, environmental justice, etc

·Public administration, finance, management, communications, etc.

Candidates should have a demonstrated interest or career goals in a water-health or public-health related field, including sustainability and resilience. Demanded qualities include strong analytical, written and oral skills, the ability to work independently and on a team, and a strong desire to take initiative.

The volunteer intern will work for a minimum of 12 weeks on a full-time (40 hours per week) or part-time (20 hours per week) schedule. Participants will not be paid, but are qualified to apply for a travelling expense. Participants do not become EPA employees during their stay.

Applications will be accepted on a rolling basis;however, most projects will arise for internship opportunities during the   Summer   Semester. Please submit your application at least 2 months before each semester begins. We recognize that timeframes may vary due to class schedules and we will work with individual candidates to specify exact timing and duration as necessary. The general timeframes by semester are as follows:

Summer Semester (late May through late August)

Fall Semester (September through December)

Spring Semester (January through April)

To apply, email a resume, cover letter, and a short writing sample (maximum 5 pages) to: braschayko. kelley@ epa. gov.

1. According to the passage, applicants are required to

A．be full-time graduate students

B．speak foreign language fluently

C．have the ability to work with teams

D．major in a public-health related field

2. What will EPA provide for the volunteer interns?

A．Financial support for transport.

B．Personalized work schedules.

C．A letter of recommendation.

D．Employment opportunities

3. What should candidates do to apply for the internship?

A．Email your intern timeframes.

B．Provide the degree certificate.

C．Work in a team for 12 weeks

D．Offer a short writing sample.

10 . 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.

1. Bamman and his colleagues designed the task to_________.

A．compare the accuracy rate of ChatGPT and GPT-4

B．test the range of knowledge of ChatGPT and GPT-4

C．show how ChatGPT and GPT-4 memorize many books

D．check what ChatGPT and GPT-4’s training data consist of

2. What can we learn from this passage?

A．AIs were trained more on copyrighted works than those out of copyright.

B．Guadamuz thinks what AIs have done is a kind of copyright infringement.

C．AI companies need to uncover copyrighted materials used as training data.

D．The permission for the use of copyright works becomes a legal challenge.

3. The passage is mainly about_________

A．The training process of AIs.	B．The legal uncertainty of AIs.
C．The future regulation of AIs.	D．The training materials of AIs.