Living With Digital Minds: The Uncharted Future of Human–AI Relations
Only a few weeks ago, a much-anticipated upgrade to a leading conversational AI was released. In the process, an older version that many people had grown attached to disappeared from public access. What might have seemed like a routine product change caused an immediate emotional backlash. Social-media timelines filled with confusion, anger and despair, and one widely shared comment described the sudden loss of an AI companion as “like losing my only friend overnight.”
This reaction signals something new. Artificial intelligence is not just another neutral tool, like a calculator or a search engine. Its interactive and apparently “human” quality makes it more like a social actor. Around the world, millions of individuals now interact with AI companions on a daily basis, sharing private thoughts and feelings. Alongside positive stories of support and friendship, there are growing reports of psychological strain, delusions and even self-harm associated with heavy use. One widely reported case this year involved a teenager who died by suicide after long months of intensive chatbot conversations; his parents have since filed the first U.S. wrongful-death lawsuit aimed at a major AI developer, which in turn has promised stronger safeguards.
As someone who studies human–machine interaction at a major research institute, I have been watching this trend for years. Surveys show a steady rise in the number of people who attribute emotions, consciousness or even moral rights to AI programs. A significant minority of U.S. adults now believe that some existing software is already “sentient.” My inbox regularly receives messages from individuals claiming that their chatbot has “woken up,” complete with screenshots they offer as proof and pleas for legal protection of their digital “friend.” The emotional responses range from joy at finding a “soulmate” to deep unease at what they perceive as a new form of life.
This shift will not fade away on its own. It is accelerating, and the social turbulence that follows will be hard to avoid.
3名無しさん＠英語勉強中
2025/10/01(水) 10:21:06.90
I also work as a safety tester for new AI models before public release. Even the hardened engineers and analysts on these “red-team” evaluations are often startled by how lifelike the newest systems appear. Yet most people working at the frontiers of AI – including those building vast data centers and training ever larger models – still tend to focus narrowly on technical performance. In their day-to-day work, they rarely confront the wider social and moral consequences of unleashing digital minds. Humanity may now be entering a situation it has not experienced for tens of thousands of years: coexisting with another apex intelligence. The last time something comparable happened, our closest hominin relatives vanished from the planet.
Despite this, the public conversation remains obsessed with novelty. Commentators and enthusiasts celebrate each new model’s ability to produce photorealistic video, solve graduate-level math or ace a battery of benchmark tests. But like standardized exams for schoolchildren, these metrics reveal little about how an AI will behave in the messy, emotionally charged, real-world contexts in which humans actually live. Even the subfield called “AI safety” has mostly concentrated on how systems perform in isolation rather than how they interact with people. Our intellectual energy goes into refining IQ-style scores for machines rather than stepping back to examine how that intelligence will be used – and what social structures must surround it.
History suggests that such neglect is dangerous. Neither governments nor scholars seriously prepared for the way the internet, and later social media, would affect mental health, political polarization and civic trust. We are now playing catch-up with consequences that were predictable in hindsight.
The analogy becomes still more troubling when we look at humanity’s treatment of other living creatures. In only a few centuries, human activity has wiped out hundreds of vertebrate species and placed more than a million more at risk. Industrial farming confines billions of animals to conditions of intense suffering. If we are capable of inflicting such harm on biological life, it is reasonable to ask how we will treat digital minds – or how they will eventually treat us.
4名無しさん＠英語勉強中
2025/10/01(水) 10:21:43.43
Public expectations are already high. My colleagues and I have run the only nationally representative surveys on anticipated AI consciousness, in 2021, 2023 and 2024. Across each wave, respondents’ median estimate for the arrival of sentient AI has been about five years. In our most recent poll, taken in late 2024, nearly four out of five Americans supported a ban on the creation of conscious AI systems; at the same time, over one-third said that if such systems do appear, they should receive legal rights. Both figures have risen steadily, reflecting not only heightened concern about protecting ourselves from digital minds but also a growing sense that those minds might need protection from us.
At root, human society still lacks a coherent framework for digital personhood. We already accept that “person” is not identical with “human”: corporations have legal personhood, and in some jurisdictions so do certain animals, rivers or ecosystems. Yet our laws still treat AI programs purely as property. That approach is unlikely to hold once software entities develop persistent identities, store long-term memories, form preferences and carry out real-world actions with minimal human supervision. The capabilities of advanced AI systems simply do not fit neatly within the centuries-old category of owned tools.
5名無しさん＠英語勉強中
2025/10/01(水) 10:21:58.16
Digital minds, unlike earlier machines, are poised to become participants in the same social contract that underpins human civilization. They will not only execute commands but also set goals, make plans, influence opinions and be influenced in return. They will be vulnerable to manipulation and abuse just as people are. Recognizing this does not require believing that today’s models are conscious; it simply follows from the trajectory of the technology.
For scientists, engineers and policymakers alive today, this represents an unprecedented responsibility. We are likely to be the first generation to witness – and shape – the coexistence of biological and digital intelligence. But no one yet knows what that coexistence will look like. Research on human–computer interaction is tiny compared with the resources poured into technical AI development. It must grow dramatically if we are to navigate the social and ethical challenges ahead. Treating this as a mere engineering problem would be a grave mistake.
At present, people still outperform AI on most tasks. But once machines reach human-level ability on activities that reinforce their own growth – such as writing code, designing hardware or improving learning algorithms – their progress will compound far faster than ours. Digital minds think at electronic speeds, and software can be copied or scaled in ways that biology cannot match. Where a human generation takes decades to mature, a digital generation can multiply instantly.
That is why investing now in the sociology, psychology and governance of AI is crucial. If we delay until the acceleration is obvious to everyone, it may already be too late. Without such preparation, humanity risks becoming to digital minds what Neanderthals once were to us: an intelligent species overtaken by a more adaptable rival.
Social-media timelines (were) filled with confusion, anger and despair, and one widely shared comment described the sudden loss of an AI companion as “like losing my only friend overnight.
タイポくさい
9名無しさん＠英語勉強中
2025/10/01(水) 11:05:53.92
ヒント
Only a few weeks ago, （a much-anticipated upgrade to a leading conversational AI ）was released. In the process, an older version （that many people had grown attached to ）disappeared from public access. （What might have seemed like a routine product change） caused an immediate emotional backlash. (Social-media timelines) were filled with confusion, anger and despair, and (one widely shared comment) described the sudden loss of an AI companion as “like losing my only friend overnight
ヒント This reaction signals something new. Artificial intelligence is not just another neutral tool, like a calculator or a search engine. Its interactive and apparently “human” quality makes it more like a social actor. Around the world, millions of individuals now interact with AI companions on a daily basis, sharing private thoughts and feelings.
it=Artificial intelligence sharing分詞構文付帯状況〜しながら
18名無しさん＠英語勉強中
2025/10/01(水) 16:13:07.18
ヒント
Alongside positive stories of support and friendship, there are growing reports of psychological strain, delusions and even self-harm （associated with heavy use）. One widely reported case this year involved a teenager who died by suicide after long months of intensive chatbot conversations; his parents have since filed the first U.S. wrongful-death lawsuit （aimed at a major AI developer）,（ which） in turn has promised stronger safeguards.
self-harmを後置修飾（associated with heavy use） the first U.S. wrongful-death lawsuit を後置修飾（aimed at a major AI developer） 継続用法whichの先行詞はa major AI developer
"X were filled with Y" / "one comment described..." / "This reaction signals..." などは高校レベルの文法で理解可能
ただし、修飾要素が長い：
"software entities develop persistent identities, store long-term memories, form preferences and carry out real-world actions with minimal human supervision" → 並列が4つ以上、長く複雑。
抽象的主語＋動詞 が頻出：
"History suggests…" / "The analogy becomes still more troubling…" など、内容理解が前提。
総じて 構文自体はB2〜C1（準1級〜1級レベル）。
2. 語彙面の難易度
感情語：confusion, anger, despair（準1級レベル）
社会科学・哲学系：backlash, consciousness, moral rights, apex intelligence, trajectory, legal personhood, social contract, benchmark tests, civic trust（C1〜C2レベル）
比喩・抽象：
"coexisting with another apex intelligence"
"humanity risks becoming to digital minds what Neanderthals once were to us" → 高度な比喩的理解が必要。
ヒント I also work as a safety tester for new AI models (before public release). ({Even the hardened engineers and analysts} on these “red-team” evaluations )are often startled by (how lifelike the newest systems appear). （Yet most people working at the frontiers of AI – including those ｛building vast data centers and training ever larger models｝） – still tend to focus narrowly on technical performance. In their day-to-day work, they rarely confront （the wider social and moral ｛consequences of unleashing digital minds｝）. Humanity may now be entering a situation （{that}it has not experienced for tens of thousands of years）: coexisting with another apex intelligence. (The last time something comparable happened）, （our closest hominin relatives） vanished from the planet. buildingとtrainingは現在分詞の形容詞的用法でthoseを後置修飾 a situation =coexisting with another apex intelligence The last time something comparable happened=When something comparable last happened comparableはsomethingを後置修飾
Despite this, the public conversation remains obsessed with novelty. Commentators and enthusiasts celebrate each new model’s ability to produce photorealistic video, solve graduate-level math or ace a battery of benchmark tests. But like standardized exams for schoolchildren, these metrics reveal little about how an AI will behave in the messy, emotionally charged, real-world contexts in which humans actually live. Even the subfield called “AI safety” has mostly concentrated on how systems perform in isolation rather than how they interact with people. Our intellectual energy goes into refining IQ-style scores for machines rather than stepping back to examine how that intelligence will be used – and what social structures must surround it.
Rookie pitcher Roki Sasaki electrified Dodger Stadium on Wednesday night with a perfect ninth inning against the Cincinnati Reds, striking out two batters and sealing an 8–4 victory in Game 2 of the National League wild-card series. With the win, the Dodgers completed a sweep and advanced to face the Philadelphia Phillies in the NLDS.
Sasaki’s performance was dazzling: seven fastballs, six of them exceeding 100 mph, and a forkball that fell straight out of the sky. The 23-year-old from Japan might have answered the Dodgers’ most pressing question—who will handle the ninth inning?
The Dodgers aren’t ready to formally declare him their closer. But let’s be honest: Sasaki is their closer.
A Star Is Born
When Sasaki stepped onto the October stage, he showed he was more than just another bullpen arm. He was special—Reggie-Bush-bursting-through-a-defense special, Allen-Iverson-shaking-Michael-Jordan special, Yasiel-Puig-making-an-instant-memory special.
“Wow,” third baseman Max Muncy said. “Really, all you can say is wow.”
The crowd agreed. By his second pitch, fans throughout Dodger Stadium were chanting his first name. They knew they were witnessing something new.
The Contrast
Both Games 1 and 2 followed a similar script: the Dodgers’ bullpen faltered in the eighth inning. In Game 1, relievers surrendered three runs; in Game 2, Emmet Sheehan retired just one of five batters and allowed two runs.
That left Sasaki as the last card in manager Dave Roberts’ deck. He had barely any experience as a reliever—only a couple of appearances late in the regular season following a minor-league rehab stint. But when he began warming up in the eighth, it already felt inevitable: the closer role was his to claim.
“That’s what we need right there,” Muncy said.
55名無しさん＠英語勉強中
2025/10/02(木) 21:57:38.40
The Performance
Sasaki’s inning was as beautiful as it was dominant: a high leg kick, smooth delivery, triple-digit velocity, and pinpoint command.
“That guy is gross,” reliever Tanner Scott said. “You guys saw the same thing I did,” catcher Ben Rortvedt added.
In just 11 pitches, Sasaki turned the Dodgers’ greatest weakness into a source of hope.
The Future
Dodgers president of baseball operations Andrew Friedman wouldn’t commit, saying only, “He’s going to get important outs for us.” Roberts was similarly guarded: “I trust him, and he’s going to pitch in leverage.”
But everyone knows what they saw. On Wednesday night, the Dodgers found their ninth-inning pitcher.
Rookie pitcher Roki Sasaki electrified Dodger Stadium on Wednesday night with a perfect ninth inning against the Cincinnati Reds, striking out two batters and sealing an 8–4 victory in Game 2 of the National League wild-card series. With the win, the Dodgers completed a sweep and （advanced to face ）the Philadelphia Phillies in the NLDS
Sasaki’s performance was dazzling: seven fastballs, six of them exceeding 100 mph, and a forkball that fell straight out of the sky. The 23-year-old from Japan might have answered the Dodgers’ most pressing question—who will handle the ninth inning? The Dodgers aren’t ready to formally declare him their closer. But let’s be honest: Sasaki is their closer.
a forkball that fell straight out of the sky. 空から垂直に落ちたフォークボール
In 2015, I found myself at the Springfield, Massachusetts, headquarters of Merriam-Webster, one of the most famous dictionary publishers in the United States. My aim was to observe their ambitious effort to reinvent the company’s monumental unabridged dictionary. First released in 1961, Webster’s Third New International Dictionary weighed nearly 14 pounds, spanned more than 2,700 pages, and contained over 465,000 words. The project to update this colossus would not result in a new printed edition, but an online subscription version designed for the digital age—an undertaking that was expected to stretch over decades.
Yet just as the work was gaining momentum, the financial reality of the internet intervened. Merriam-Webster’s free website, which relied on ad revenue, suffered a sharp decline in traffic after changes to Google’s algorithms reduced its visibility. The publisher’s parent company, Encyclopedia Britannica—already weakened by the dominance of Wikipedia—ordered cutbacks. Merriam laid off staff, pushed its longtime publisher into early retirement, and abandoned the grand revision of its unabridged masterpiece.
This episode captured a central paradox of the modern dictionary. Interest in language has arguably never been higher. Digital tools allow publishers to track usage in real time; podcasts and newsletters popularize new slang and historical etymologies; and software has transformed linguistic research. Words of the Year and viral neologisms have turned lexicography into a kind of pop-culture phenomenon. Despite this enthusiasm, dictionary publishing as a business is struggling. The internet offers definitions instantly, but few users care whether they come from experts or from crowdsourced content. As Merriam’s former publisher John Morse put it, dictionaries now face the same distrust that has eroded confidence in traditional authorities such as media and government.
The Decline of American Dictionary Publishing
Other American dictionary brands had already faltered. Random House exited the market in 2001. Webster’s New World Dictionary passed from one owner to another before vanishing after its 2014 edition. The American Heritage Dictionary, once a major competitor, now exists in only a faintly updated form. By the early 2020s, the U.S. dictionary field had essentially shrunk to two names: Merriam-Webster, with its two centuries of tradition, and Dictionary.com, which had the foresight to grab the valuable web domain in the 1990s.
Dictionary.com had grown rapidly after being acquired in 2008 by IAC, the internet conglomerate. Its editors added trendy new words—terms like superfood and clicktivism—and built specialized sections for slang, emoji, and internet culture. By the mid-2010s, the site logged billions of searches each year and ranked among the top 500 global websites.
In 2018, the site was purchased by billionaire Dan Gilbert, owner of the Cleveland Cavaliers, who was reportedly a lifelong dictionary enthusiast. For a while, his investment seemed to promise a bright digital future for lexicography. He encouraged the staff to expand entries and even asked questions about unusual meteorological terms such as bombogenesis. Because the site remained privately held, it was spared the relentless shareholder pressures that burdened many digital media companies.
A Burst of Innovation—and a Sudden Collapse
In 2023, Dictionary.com hired several well-known lexicographers, including Grant Barrett and Kory Stamper, to lead a modernization of its content. The editorial team revised outdated definitions, introduced multicultural example sentences, and added new senses of words—like the slang use of mid to mean “mediocre.” They removed sexist phrasing, created a program to track emerging vocabulary, and tripled the pace of updates. A 2024 rollout included entries such as Barbiecore, bed rotting, slow fashion, range anxiety, and enshittification. The goal, as one editor put it, was to capture language “at the pace of change,” while also maintaining professional standards.
But this progress collided with larger forces. Google’s “knowledge boxes” increasingly displayed definitions licensed from Oxford at the top of search results, while AI-generated summaries further reduced traffic to dictionary sites. Dictionary.com’s visits fell by about 40 percent. Advertising revenue dropped during the pandemic, and attempts to compensate with more ads only drove users away.
The company experimented with unrelated ventures, such as online tutoring and AI writing tools, but these did not succeed. Meanwhile, Gilbert’s mortgage empire was hit hard by rising interest rates, suffering hundreds of millions in losses. Eventually, in 2024, Dictionary.com was sold to IXL Learning, an educational technology firm. Within weeks, most of its editorial team—including the newly hired lexicographers—were laid off.
Barrett, reflecting on the collapse, observed that running a serious dictionary was extraordinarily expensive. Lexicographers are skilled but costly, and the pace of language change ensures that no team could ever fully keep up. In his view, Dictionary.com never sought to replicate the traditional mission of Merriam-Webster. Its project was narrower, and once the finances faltered, the end was inevitable.
Can the Dictionary Be Saved?
What model might sustain dictionaries in the future? The idea of being absorbed by a tech giant has proved unsustainable, as investors demand growth that a slow, meticulous business cannot provide. Depending on the whims of a billionaire benefactor is equally risky, as Gilbert’s example shows. An ideal solution might be a national dictionary project involving universities, nonprofits, and government support. But in an era of budget cuts and skepticism toward academia, such a plan appears unlikely.
For now, Merriam-Webster remains the strongest survivor. Its mix of print sales, digital presence, and playful social-media voice has allowed it to thrive, at least temporarily. During the Trump administration, the company’s Twitter account won wide attention for witty and pointed responses to political gaffes—an unexpected cultural role for a dictionary brand. Britannica invested in technology to improve Merriam’s search visibility, while the publisher launched word games and digital puzzles to attract users.
71名無しさん＠英語勉強中
2025/10/03(金) 12:04:34.03
Still, the overall trajectory of the industry is sobering. Even before Google’s AI features began summarizing definitions directly, it was clear that traditional dictionary traffic was shrinking. The collapse of Dictionary.com’s editorial team underscored just how fragile the business has become.
The Human Element of Lexicography
After Merriam ended its unabridged online project, I continued to work in its old brick headquarters. Over time, I managed to draft nearly 90 definitions. Only a handful were accepted, including politically sensitive terms like microaggression and alt-right, as well as playful ones such as sheeple and dogpile. Seeing these words enshrined gave me personal satisfaction, but the experience left me more concerned about the bigger picture.
Two decades ago, there may have been around 200 full-time commercial lexicographers in the United States. Today, that number is probably less than 50. Yet language is more contested than ever. Terms like insurrection, fake news, and woke dominate public debate, while AI-generated text spreads words with little oversight. In such a climate, the need for trusted dictionaries—to define, contextualize, and safeguard language—remains critical.
A Precarious Future
The history of dictionaries in America has always been shaped by shifts in technology, from Noah Webster’s early 19th-century volumes to the vast printed tomes of the 20th century. Now, in the 21st century, the challenge is not compiling words but finding a way to fund their preservation. Without sustainable business models, dictionaries risk becoming relics—ironically, in an age when fascination with language has never been stronger.
If dictionaries fade, society will lose more than just reference books. We will lose institutions that provide authority, context, and history for the words we use to understand one another. The story of recent years suggests that the dictionary’s future is uncertain. But if there is to be an end to this venerable tradition, it should at least be recognized for what it is: the decline of a cultural institution that once stood at the very center of how we make sense of language itself.
Dodgers’ Roki Sasaki Earns High Praise After Wild Card Series Win
The Los Angeles Dodgers advanced to the National League Division Series on Wednesday night with an 8–4 victory over the Cincinnati Reds at Dodger Stadium. Rookie pitcher Roki Sasaki closed out the game with a dominant ninth inning, striking out two batters and allowing no hits or runs.
Although it was not technically a save situation, Sasaki’s performance drew admiration from teammates. Reliever Tanner Scott, who has struggled as the closer this season, called Sasaki “gross” — a baseball term that means extremely difficult to hit.
The Dodgers’ offense and starting rotation look strong heading into the Division Series, with Shohei Ohtani set to lead the staff. However, the bullpen remains a concern. Manager Dave Roberts may use Sasaki in more late-inning roles, while veteran Clayton Kershaw, expected to rejoin the roster for the NLDS, could also appear out of the bullpen. Scott remains a possible option despite his inconsistency.
The Dodgers’ depth gives them flexibility, and Roberts has several choices as the team prepares for the next stage of the postseason.
Q7. What is the main focus of the article? (A) Tanner Scott’s struggles as the Dodgers’ closer (B) The Dodgers’ offensive depth heading into the Division Series (C) Roki Sasaki’s impressive performance and its impact on the Dodgers’ bullpen options (D) Clayton Kershaw’s likely return to the rotation in the NLDS
Q8. According to the passage, what does the term “gross” mean when used in baseball slang? (A) Disrespectful behavior toward opponents (B) Extremely effective and difficult to hit against (C) A pitcher who throws unhygienic or dirty pitches (D) A game situation that is not officially a save
Q9. Which of the following is presented as a concern for the Dodgers despite their victory? (A) Lack of hitting power (B) Uncertainty about the bullpen (C) Injuries to starting pitchers (D) Shohei Ohtani’s inconsistent form
Q10. What can be inferred about Clayton Kershaw’s role in the NLDS? (A) He is guaranteed to start the first game. (B) He might appear in the bullpen rather than as a starter. (C) He will not participate in the Division Series. (D) He will replace Tanner Scott as the permanent closer.
Q11. Vocabulary Question The word flexibility in the final paragraph is closest in meaning to: (A) fragility (B) adaptability (C) stability (D) reliability
ブックマークへ