Former astronaut Ellen Ochoa says that although she doesn’t have a definite idea of when it might happen, she ________ that humans will someday need to be able to live in other environments than those found on Earth. This conjecture informs her interest in future research missions to the moon. Which choice completes the text with the most logical and precise word or phrase?

Following the principles of community-based participatory research, tribal nations and research institutions are equal partners in health studies conducted on reservations. A collaboration between the Crow Tribe and Montana State University ___________ this model: tribal citizens worked alongside scientists to design the methodology and continue to assist in data collection. Which choice completes the text with the most logical and precise word or phrase?

In 2007, computer scientist Luis von Ahn was working on converting printed books into a digital format. He found that some words were distorted enough that digital scanners couldn’t recognize them, but most humans could easily read them. Based on that finding, von Ahn invented a simple security test to keep automated “bots” out of websites. The first version of the reCAPTCHA test asked users to type one known word and one of the many words scanners couldn’t recognize. Correct answers proved the users were humans and added data to the book-digitizing project. Which choice best states the main purpose of the text?

The following text is from Edith Wharton’s 1905 novel The House of Mirth. Lily Bart and a companion are walking through a park. Lily had no real intimacy with nature, but she had a passion for the appropriate and could be keenly sensitive to a scene which was the fitting background of her own sensations. The landscape outspread below her seemed an enlargement of her present mood, and she found something of herself in its calmness, its breadth, its long free reaches. On the nearer slopes the sugar-maples wavered like pyres of light; lower down was a massing of grey orchards, and here and there the lingering green of an oak-grove. Which choice best describes the function of the underlined sentence in the text as a whole?

The following text is from Maggie Pogue Johnson’s 1910 poem “Poet of Our Race.” In this poem, the speaker is addressing Paul Laurence Dunbar, a Black author. Thou, with stroke of mighty pen, Hast told of joy and mirth, And read the hearts and souls of men As cradled from their birth. The language of the flowers, Thou hast read them all, And e’en the little brook Responded to thy call. Which choice best states the main purpose of the text?

To You” is an 1856 poem by Walt Whitman. In the poem, Whitman suggests that readers, whom he addresses directly, have not fully understood themselves, writing, Which quotation from “To You” most effectively illustrates the claim?

Born in 1891 to a Quechua-speaking family in the Andes Mountains of Peru, Martín Chambi is today considered to be one of the most renowned figures of Latin American photography. In a paper for an art history class, a student claims that Chambi’s photographs have considerable ethnographic value—in his work, Chambi was able to capture diverse elements of Peruvian society, representing his subjects with both dignity and authenticity. Which finding, if true, would most directly support the student’s claim?

Herbivorous sauropod dinosaurs could grow more than 100 feet long and weigh up to 80 tons, and some researchers have attributed the evolution of sauropods to such massive sizes to increased plant production resulting from high levels of atmospheric carbon dioxide during the Mesozoic era. However, there is no evidence of significant spikes in carbon dioxide levels coinciding with relevant periods in sauropod evolution, such as when the first large sauropods appeared, when several sauropod lineages underwent further evolution toward gigantism, or when sauropods reached their maximum known sizes, suggesting that ___ Which choice most logically completes the text?

Public-awareness campaigns about the need to reduce single-use plastics can be successful, says researcher Kim Borg of Monash University in Australia, when these campaigns give consumers a choice: for example, Japan achieved a 40 percent reduction in plastic-bag use after cashiers were instructed to ask customers _____ whether wanted a bag. Which choice completes the text so that it conforms to the conventions of Standard English?

In ancient Greece, an Epicurean was a follower of Epicurus, a philosopher whose beliefs revolved around the pursuit of pleasure. Epicurus defined pleasure as “the absence of pain in the body and of trouble in the that all life’s virtues derived from this absence. Which choice completes the text so that it conforms to the conventions of Standard English?

British scientists James Watson and Francis Crick won the Nobel Prize in part for their 1953 paper announcing the double helix structure of DNA, but it is misleading to say that Watson and Crick discovered the double helix. findings were based on a famous X-ray image of DNA fibers, “Photo 51,” developed by X-ray crystallographer Rosalind Franklin and her graduate student Raymond Gosling. Which choice completes the text so that it conforms to the conventions of Standard English?

Scientists believe that, unlike most other species of barnacle, turtle barnacles (Chelonibia testudinari) can dissolve the cement-like secretions they use to attach to a sea turtle shell, enabling the barnacles to move short distances across the shell’s surface. Which choice completes the text so that it conforms to the conventions of Standard English?

The classic children’s board game Chutes and Ladders is a version of an ancient Nepalese game, Paramapada Sopanapata. In both games, players encounter “good” or “bad” spaces while traveling along a path; landing on one of the good spaces __________ a player to skip ahead and arrive closer to the end goal. Which choice completes the text so that it conforms to the conventions of Standard English?

In 1943, in the midst of World War II, mathematics professor Grace Hopper was recruited by the US military to help the war effort by solving complex equations. Hopper’s subsequent career would involve more than just as a pioneering computer programmer, Hopper would help usher in the digital age. Which choice completes the text so that it conforms to the conventions of Standard English?

In 1453, English King Henry VI became unfit to rule after falling gravely ill. As a result, Parliament appointed Richard, Third Duke of York, who had a strong claim to the English throne, to rule as Lord Protector. Upon recovering two years later, forcing an angered Richard from the royal court and precipitating a series of battles later known as the Wars of the Roses. Which choice completes the text so that it conforms to the conventions of Standard English?

What is 10% of 470?

4x + 6 = 18 Which equation has the same solution as the given equation?

The function g is defined by g(x) = x^2 + 9. For which value of x is g(x) = 25 ?

A printer produces posters at a constant rate of 42 posters per minute. At what rate, in posters per hour, does the printer produce the posters?

Right triangles LMN and PQR are similar, where L and M correspond to P and Q, respectively. Angle M has a measure of 53°. What is the measure of angle Q ?

y = -3x 4x + y = 15 The solution to the given system of equations is (x, y). What is the value of x ?

z^2 + 10z - 24 = 0 What is one of the solutions to the given equation?

Bacteria are growing in a liquid growth medium. There were 300,000 cells per milliliter during an initial observation. The number of cells per milliliter doubles every 3 hours. How many cells per milliliter will there be 15 hours after the initial observation?

Vivian bought party hats and cupcakes for $71. Each package of party hats cost $3, and each cupcake cost $1. If Vivian bought 10 packages of party hats, how many cupcakes did she buy?

Which expression is equivalent to 6x^8y^2 + 12x^2y^2?

A neighborhood consists of a 2-hectare park and a 35-hectare residential area. The total number of trees in the neighborhood is 3,934. The equation 2x + 35y = 3,934 represents this situation. Which of the following is the best interpretation of x in this context?

A circle in the xy-plane has a diameter with endpoints (2, 4) and (2, 14). An equation of this circle is (x − 2)^2 + (y − 9)^2 = r^2, where r is a positive constant. What is the value of r ?

A list of 10 data values is shown. 6, 8, 16, 4, 17, 26, 8, 5, 5, 5 What is the mean of these data?

g(x) = x^2 + 55 What is the minimum value of the given function?

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for.

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. 2. What is the writer doing in the second paragraph?

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. 3. Why does the writer mention the story of King Midas?

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. 4. What challenge does the writer refer to in the fourth paragraph?

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. 5. What does the writer suggest about the future of Al in the fifth paragraph?

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. 6. Which of the following best summarises the writer’s argument in the sixth paragraph?

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. Do the following statements agree with the claims of the writer in Living With Artificial Intelligence Reading Passage? YES – if the statement agrees with the claims of the writer NO – if the statement contradicts the claims of the writer NOT GIVEN – if itis impossible to say what the writer thinks about this 7. Machines with the ability to make moral decisions may prevent us from promoting the interests of our communities.

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. Do the following statements agree with the claims of the writer in Living With Artificial Intelligence Reading Passage? YES – if the statement agrees with the claims of the writer NO – if the statement contradicts the claims of the writer NOT GIVEN – if itis impossible to say what the writer thinks about this 8. Silicon police would need to exist in large numbers in order to be effective.

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. Do the following statements agree with the claims of the writer in Living With Artificial Intelligence Reading Passage? YES – if the statement agrees with the claims of the writer NO – if the statement contradicts the claims of the writer NOT GIVEN – if itis impossible to say what the writer thinks about this 9. Many people are comfortable with the prospect of their independence being restricted by machines.

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. Do the following statements agree with the claims of the writer in Living With Artificial Intelligence Reading Passage? YES – if the statement agrees with the claims of the writer NO – if the statement contradicts the claims of the writer NOT GIVEN – if itis impossible to say what the writer thinks about this 10. If we want to ensure that machines act in our best interests, we all need to work together.

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. Complete the summary using the list of phrases, A-F, below. Choose the correct option for the question number 11. Al currently has a limited role in the way 11 ………………………. are allocated in the health service. The positive aspect of Al having a bigger role is that it would be more efficient and lead to patient benefits. However, such a change would result, for example, in certain 12 ………………………. not having their current level of 13……………………….. It is therefore important that Al goals are appropriate so that discriminatory practices could be avoided.

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. Complete the summary using the list of phrases, A-F, below. Choose the correct option for the question number 12. Al currently has a limited role in the way 11 ………………………. are allocated in the health service. The positive aspect of Al having a bigger role is that it would be more efficient and lead to patient benefits. However, such a change would result, for example, in certain 12 ………………………. not having their current level of 13……………………….. It is therefore important that Al goals are appropriate so that discriminatory practices could be avoided.

Powerful artificial intelligence (Al) needs to be reliably aligned with human values, but does this mean Al will eventually have to police those values? This has been the decade of Al, with one astonishing feat after another. A chess playing Al that can defeat not only all human chess players, but also all previous human-programmed chess machines, after learning the game in just four hours? That’s yesterday’s news, what’s next? True, these prodigious accomplishments are all in so called narrow Al , where machines perform highly specialised tasks. But many experts believe this restriction is very temporary. By mid-century, we may have artificial general intelligence (AGI) – machines that can achieve human-level performance on the full range of tasks that we ourselves can tackle. If so, there’s little reason to think it will stop there. Machines will be free of many of the physical constraints on human intelligence. Our brains run at slow biochemical processing speeds on the power of a light bulb, and their size is restricted by the dimensions of the human birth canal. It is remarkable what they accomplish, given these handicaps. But they may be as far from the physical limits of thought as our eyes are from the incredibly powerful Webb Space Telescope. Once machines are better than us at designing even smarter machines, progress towards these limits could accelerate. What would this mean for us? Could we ensure a safe and worthwhile coexistence with such machines? On the plus side, Al is already useful and profitable for many things, and super Al might be expected to be super useful, and super profitable. But the more powerful Al becomes, the more important it will be to specify its goals with great care. Folklore is full of tales of people who ask for the wrong thing, with disastrous consequences – King Midas, for example, might have wished that everything he touched turned to gold, but didn’t really intend this to apply to his breakfast. So we need to create powerful Al machines that are ‘human-friendly’ – that have goals reliably aligned with our own values. One thing that makes this task difficult is that we are far from reliably human-friendly ourselves. We do many terrible things to each other and to many other creatures with whom we share the planet. If super intelligent machines don’t do a lot better than us, we’ll be in deep trouble. We’ll have powerful new intelligence amplifying the dark sides of our own fallible natures. For safety’s sake, then, we want the machines to be ethically as well as cognitively superhuman. We want them to aim for the moral high ground, not for the troughs in which many of us spend some of our time. Luckily they’ll be smart enough for the job. If there are routes to the moral high ground, they’ll be better than us at finding them, and steering us in the right direction. However, there are two big problems with this utopian vision. One is how we get the machines started on the journey, the other is what it would mean to reach this destination. The ‘getting started’ problem is that we need to tell the machines what they’re looking for with sufficient clarity that we can be confident they will find it – whatever ‘it’ actually turns out to be. This won’t be easy, given that we are tribal creatures and conflicted about the ideals ourselves. We often ignore the suffering of strangers, and even contribute to it, at least indirectly. How then, do we point machines in the direction of something better? As for the ‘destination’ problem, we might, by putting ourselves in the hands of these moral guides and gatekeepers, be sacrificing our own autonomy – an important part of what makes us human. Machines who are better than us at sticking to the moral high ground may be expected to discourage some of the lapses we presently take for granted. We might lose our freedom to discriminate in favour of our own communities, for example. Loss of freedom to behave badly isn’t always a bad thing, of course: denying ourselves the freedom to put children to work in factories, or to smoke in restaurants are signs of progress. But are we ready for ethical silicon police limiting our options? They might be so good at doing it that we won’t notice them; but few of us are likely to welcome such a future. These issues might seem far-fetched, but they are to some extent already here. Al already has some input into how resources are used in our National Health Service (NHS) here in the UK, for example. If it was given a greater role, it might do so much more efficiently than humans can manage, and act in the interests of taxpayers and those who use the health system. However, we’d be depriving some humans (e.g. senior doctors) of the control they presently enjoy. Since we’d want to ensure that people are treated equally and that policies are fair, the goals of Al would need to be specified correctly. We have a new powerful technology to deal with – itself, literally, a new way of thinking . For our own safety, we need to point these new thinkers in the right direction, and get them to act well for us. It is not yet clear whether this is possible, but if it is, it will require a cooperative spirit, and a willingness to set aside self-interest. Both general intelligence and moral reasoning are often thought to be uniquely human capacities. But safety seems to require that we think of them as a package: if we are to give general intelligence to machines, we’ll need to give them moral authority, too. And where exactly would that leave human beings? All the more reason to think about the destination now, and to be careful about what we wish for. Complete the summary using the list of phrases, A-F, below. Choose the correct option for the question number 13. Al currently has a limited role in the way 11 ………………………. are allocated in the health service. The positive aspect of Al having a bigger role is that it would be more efficient and lead to patient benefits. However, such a change would result, for example, in certain 12 ………………………. not having their current level of 13……………………….. It is therefore important that Al goals are appropriate so that discriminatory practices could be avoided.