正在学习

Impact on Relationships

Impact on Relationships

Late into 2024, one of us was preparing for an interesting evening where a group of us were planning to recite our favourite speeches or sonnets from Shakespeare. Ganesh, an avid fan of the play Julius Caesar, immediately chose Mark Antony’s declamation over Caesar’s murdered body, ‘O Pardon me thou bleeding piece of earth’, and asked the handy Meta AI on his phone to remind him of the lines. The AI provided some unsolicited information that this speech served to incite the crowd against Brutus and other conspirators. When Ganesh pointed out that it was a soliloquy and not the ‘Friends, Romans, Countrymen’ speech, the AI quickly retracted and, almost in jest, acknowledged its gratitude for being able to learn from Ganesh’s superior knowledge of the subject. Suddenly, it became almost human in its ability to retract, correct an earlier opinion and collaborate!

When this kind of emotional intelligence is developed further by AI agents, they may prove to be a more sought-after confidantes and companions to many humans than our human friends. Yuval Harari has rightly pointed out that while even our best friends might respond to a conversation in different ways depending on their state of mind on a particular day, an AI companion can be relied upon to respond based on its evolving understanding of the human, without being biased by any setbacks of its own. Does that mean that our future relationships with our AIs will be more meaningful and rewarding than those we enjoy with our own friends and family? Let us delve deeper into possibilities.

Scenario 1: A teacher admonishes her class for doing a shoddy job with their summer assignments and tells them they have a last chance to make good the following Monday. A smart student decides to use AI to get the assignment right and gets her personal AI to undertake a voyage of discovery into the teacher’s social media persona over the summer vacation and turns it into a perfect paper, contextualised to the new preferences the teacher has developed over the vacation. This success motivates the student and indeed most of the class to bring moods and experiences into assignments, and the entire process of education and learning undergoes a dramatic change.

Scenario 2: A young lady is walking down the street when a man approaches her with a half-smile of recognition. Her new version of Google Glass does a quick scan and sends three messages to her brain through an embedded neural network—who the person is, where they have met before and would it be worth her while to acknowledge his existence and try to build a relationship.

Scenario 3: A cricketer has been out of form and wants to rediscover his batting prowess to keep his place in the team. When he walks out to play on the third morning of a test match, the coach gives him some parting ideas on how to deal with the key fast bowler from the opposing team. The advice is to play with care for the first few overs while the bowler is fresh and then start scoring. On his way to the pitch, he switches on his AI and gets the suggestion that he should score aggressively off the first few balls before the bowler gets into his rhythm.

In all three cases, AI has access to something that is unique to the context of fast-changing situations in everyday life. AI could explore recent happenings through instantaneous access to social media, which is denied to teachers, coaches and individuals relying on memory and previous experience. As people get more experienced and comfortable with agentic AI and interacting with their AI as a trusted Jeeves in all situations, the need to ‘call a friend’ will go down further and further. Does this endanger the long relationships we have had with parents and elders in the family, teachers with predictable idiosyncrasies, partners, mentors and coaches? Will our friendships degenerate into transactional and opportunistic dalliances as AI slowly penetrates not just our mind but also our soul?

We will live and learn.

There are undoubtedly many positive outcomes that we will see in the foreseeable future enabled and often led by AI. Companionship to senior citizens at a time when more and more families are living cities or even countries apart is one of the most critical services that AI-powered apps and chatbots can provide on the mobile phones or computers of citizens. As capabilities mature, first-level diagnosis of ailments and treatment recommendations can come from AI without making a cumbersome visit to the doctor necessary, and even mental health patients can receive preliminary solutions on interacting with AI. In 2025, a woman used ChatGPT as a therapist initially, and eventually to write a suicide note.16 There is a danger of adverse outcomes on blindly following AI advice.

On the negative horizon, the early days of AI have already seen enough hallucinations that have resulted in erroneous and sometimes extreme generalisations, and as more and more usage of AI occurs, the risk of AI perpetuating existing biases is very real. Ongoing and regular use of AI copilots and agents could even eventually result in the diminishing of human intellect as the need to explore fades and results in unused human capabilities for search and reflection. Privacy concerns are already proliferating everywhere because of rampant use of social media, and the use of AI in fake audio and video messages is harassing every user with a presence on the internet. With more and more information and contextualised knowledge becoming available to AI, privacy is expected to be the predominant concern for all intelligent humans in future. The intrusion of AI into people’s private lives is, however, inevitable. SurveyMonkey mentions that 75 per cent of young people are using AI-powered dating apps, and Pew Research mentions that 40 per cent of relationships commence online.17 As a natural corollary, relationships and communications are mainly online, and with the ease of use of AI-mediated communications in both personal and business matters, voice, video and personal meetings are already not the preferred mode and will become increasingly rare in the years to come.

With AI already playing a significant role in psychiatry, therapy and counselling, the emotional intelligence of AI is increasing at a rapid pace, and it is likely that human relationships may actually improve rather than deteriorate because of the wise counsel provided by AI communication assistants and agents. As time progresses, the definitions of intimacy and relationships themselves may change, and we all have to be prepared for new ways of living, loving and working together.

练习题

In the opening anecdote, what human-like quality did the AI most clearly display after Ganesh corrected it?

A. It insisted on its first answer and refused correction.
B. It retracted its mistake, corrected itself, and collaborated.
C. It stopped responding because it lacked context.
D. It replaced Shakespeare with a modern summary.

According to the passage, why might some people prefer an AI companion over a human friend in conversation?

A. Because AI always agrees with every opinion.
B. Because AI has more emotions than humans.
C. Because AI can respond based on an evolving understanding of the person without its own mood-based setbacks.
D. Because AI never needs any data to respond.

Which of the following scenarios in the passage show AI using fast-changing contextual information to influence human decisions or actions? Select all that apply.

A. A student uses AI to tailor an assignment to a teacher's recent social media persona.
B. A wearable device identifies a man and suggests whether acknowledging him is worthwhile.
C. A cricketer receives AI advice that differs from the coach's strategy.
D. Ganesh recites a Shakespeare speech from memory alone.
E. A teacher grades papers without using technology.

The passage suggests that AI companions may become more appealing partly because they are less affected by their own day-to-day emotional state than human friends are.

In the cricketer scenario, the AI fully agreed with the coach's advice to begin cautiously against the fast bowler.

The author argues that growing trust in agentic AI could reduce the need to 'call a friend' for advice.

In the education example, the student's AI improves the assignment by making it "___ to the new preferences the teacher has developed over the vacation."

How could AI's access to recent social media information change human relationships and decision-making, according to the passage?

Explain one way this section connects with the prior idea that data privacy and cybersecurity are major AI risks for entrepreneurs.

A start-up founder wants to build an AI companion app for senior citizens that gives emotional support and basic guidance. Which concern from the text most directly suggests the product must include strong safeguards before encouraging users to rely on it deeply?

A.
B.
C.
D.

An entrepreneur is designing wearable AI glasses that identify people, recall prior interactions, and suggest whether building a relationship is worthwhile. Which considerations are supported by combining the current and prior knowledge points?

A.
B.
C.
D.
E.

\text{Because AI can access fast-changing contextual information and offer actionable insights from clean data, it may become a trusted adviser; however, } ___ \text{ remains a major concern when such systems use personal or social media information.}

登录后解锁笔记、知识点解析、AI 问答

立即登录