영어 모의고사 (자동 편성)

Psychologists studying human judgment often describe the mind as ① operating through two distinct systems rather than a single unified process. System 1 works quickly and automatically, ② generating impressions and snap judgments without any noticeable sense of effort, much as a driver instinctively brakes when a ball suddenly rolls into the street. System 2, by contrast, is slow, deliberate, and demanding, engaged only when a problem ③ require careful calculation or conscious reasoning. Most daily decisions are handled almost entirely by System 1, which allows people ④ to function efficiently without exhausting their limited mental resources. This efficiency comes at a cost, however, because System 1 relies on simple shortcuts that can produce systematic errors, and System 2 is frequently too lazy or too occupied to double-check them. As a result, a judgment often feels most confident precisely at the moment when it has skipped the careful scrutiny that could have caught its own mistake, leaving the thinker unaware of how easily the impression was ⑤ formed.

Two years ago, I began volunteering at an animal shelter near my neighborhood on weekends. My tasks included feeding animals, cleaning cages, and helping visitors choose a pet to adopt. Over time, I noticed how much I ①enjoyed calming nervous or injured animals. One rainy afternoon, a frightened puppy refused to eat, and I sat quietly beside it for over an hour until it finally ②relaxed. Watching the puppy trust me changed something inside me. I began researching careers in veterinary medicine and animal behavior, subjects I had ③frequently considered before. My volunteer work, which started as a simple ④hobby, slowly turned into the ⑤clearest career direction I had ever found.

John Rawls proposed an unusual thought experiment to determine what principles of justice a society ought to adopt. Imagine a group of rational people gathered to design the basic rules of their future society, but placed behind what he called a veil of ignorance, a condition under which ① none of them knows their own race, gender, wealth, talents, or social position in advance. Since no one can predict whether they will be born into privilege or poverty, each person has a strong incentive to choose rules that ② protect the worst-off members of society, rather than rules that merely favor whichever group they happen to belong to. Rawls argued that rational people placed in this position would not gamble on inequality; instead, they would favor arrangements that guarantee basic liberties for everyone and permit unequal outcomes only when those outcomes actually ③ benefit the least advantaged. Justice, on this view, is what ④ impartial reasoning produces once self-interest is ⑤ maximized.

Opening most camera apps today reveals a row of filters that can smooth skin, brighten eyes, and reshape a jawline before a photo is even taken, all powered by AI models trained to recognize facial features in real time. What began as a playful novelty has become a ① default setting for many users, subtly adjusting nearly every selfie posted online. Psychologists have grown concerned about a specific consequence of this shift: people increasingly compare their unfiltered reflection in a mirror to a filtered version of themselves seen constantly on a screen, a comparison that earlier generations never had to ② face. Some social media platforms have responded by ③ labeling heavily edited photos or banning certain filters that alter body shape. Whether such small warnings can meaningfully ④ worsen the daily flood of filtered images, however, remains an ⑤ open question that researchers are only beginning to study.

Every day, email providers quietly block billions of ① unwanted messages before they ever reach a user's inbox, a task that would be impossible without an AI filtering system running constantly in the background. The filter ② examines each incoming message for suspicious patterns, including certain keywords, unusual sender addresses, and formatting tricks that scammers commonly rely on to slip past detection. What makes the system effective is that it ③ stops learning after launch, since whenever a user marks a message as spam or rescues a message mistakenly filtered out, that action becomes new training data, sharpening the filter's future decisions. This creates a constant back-and-forth, since spammers study which messages get blocked and then ④ adjust their wording to slip through again. As a result, spam filtering is less like a fixed wall and more like an ⑤ ongoing contest, with each side continuously adapting to outsmart the other.

A smart thermostat that learns a family's daily rhythm can lower the temperature automatically before everyone wakes up cold, while smart lights that follow a resident's schedule no longer need to be switched on and off by hand. Sensors collect data about temperature, motion, and light throughout the home, and an algorithm gradually learns which settings match each household's actual habits. Yet convenience comes with a trade-off that many households overlook: _______________.

Planning a multi-city trip once meant hours spent comparing flight schedules, reading hotel reviews, and mapping out which sights were close enough to visit in a single day. AI travel planners now compress much of that work into a short conversation, generating a full day-by-day itinerary within minutes. Many travelers appreciate having a solid starting point rather than a blank page. Still, seasoned travelers warn that _______________.

Instead of typing keywords into a search bar, a growing number of people now simply upload a photo and let an algorithm figure out what it contains. (A) The system works by breaking an image down into countless small features, such as edges, colors, and shapes, then matching that pattern against millions of stored examples. (B) This process, known as reverse image search, compares the visual patterns in a picture against a massive database, identifying everything from a landmark in the background to the exact model of a pair of sneakers. (C) It is not flawless, however; a blurry photo, an unusual camera angle, or an object partly hidden behind another can easily confuse the algorithm.

Musicians, meanwhile, have mixed feelings about this shift in how audiences discover new music. Open a music streaming app on any given morning, and a personalized playlist is usually already waiting, assembled overnight from songs the algorithm predicts will match a listener's current mood. ( ① ) This prediction relies on far more than genre alone; the system studies skip rates, replay counts, and even the time of day a song is typically played to build a remarkably detailed profile of individual taste. ( ② ) Listeners often praise the feature for introducing them to artists they never would have searched for on their own, effectively acting as a tireless music scout working around the clock. ( ③ ) A song that the algorithm decides not to promote may struggle to reach listeners no matter how well it was written. ( ④ ) This means that an artist's success can depend as much on pleasing an algorithm as on pleasing an actual audience. ( ⑤ )

Because the keyboard nudges people toward its own suggestions, some linguists worry that autocomplete may be quietly narrowing everyday vocabulary. Long before a user finishes typing a sentence, the keyboard on most smartphones has already begun guessing what word comes next. ( ① ) This feature relies on a language model trained to notice which words commonly follow one another, so it can offer a shortlist of likely candidates above the keyboard, saving users countless keystrokes over the course of a day. ( ② ) The system also adapts to each individual, gradually learning a person's favorite phrases, slang, and even the names of friends that no generic dictionary would recognize. ( ③ ) This constant personalization, however, creates a subtle side effect that few users notice. ( ④ ) This small habit, choosing a convenient suggestion rather than searching for the more precise or original word one might otherwise have chosen, can add up over years of daily use. ( ⑤ )