In an Era Where AI Provides "Answers," Do Young People Need "Thinking Skills"?

With the advent of generative AI, we can now obtain "answers" more easily than ever before.

There's no need to search for unfamiliar words one by one. Long texts can be summarized in seconds, foreign languages can be instantly translated, and AI can be tasked with structuring reports, solving math problems, and writing programs.

For working professionals, this is an incredibly reliable productivity tool.

However, the conversation shifts slightly when it comes to schools and universities.

This is because the purpose of education is not merely to "complete the correct answer."

Reading texts, pausing at points of confusion, researching materials, comparing others' opinions, and drawing one's own conclusions—this time-consuming process itself is learning.

Yet, generative AI can leap over these tedious parts in an instant.

It's precisely because it's convenient that it becomes challenging to handle in education.

"The assignment is complete. But what did the student learn?"

Let's say a student submits a 3,000-word report.

The writing is natural, and the structure is solid. There are citations, and at first glance, it appears excellent.

But what if most of it was created by generative AI?

We are in an era where students can produce high-quality submissions without deeply understanding the theme.

The issue here is not just simple "cheating."

It's the possibility of entrusting AI with the very skills that were supposed to be developed through the assignment.

Reading texts independently.

Identifying key points.

Comparing multiple sources.

Considering opposing views.

Constructing sentences.

And explaining it to others in one's own words.

All of these are tedious and time-consuming.

However, it is within this tedium that thinking skills develop.

Generative AI is strong in quickly creating finished products. On the other hand, it also poses the risk of shortening the "thinking time" necessary for human learning.

In the U.S., High School Reading Scores Have Been Declining Long-Term

Concerns about young people's learning abilities did not suddenly arise with the advent of generative AI.

According to the National Assessment of Educational Progress (NAEP) in the U.S., the average reading score for 12th graders in 2024 was 3 points lower than in 2019 and 10 points lower than in 1992.

In 2024, 35% of 12th graders were "Proficient" or above according to NAEP.

However, this needs to be viewed with caution.

We cannot definitively say that this decline is due to smartphones or AI.

Various factors such as the impact of the COVID-19 pandemic, home environment, income disparity, reading habits, and the education system are related to academic performance.

Nevertheless, alongside the transition to a digital society, there is indeed a growing awareness in educational settings that "the time spent reading long texts with concentration is decreasing."

Has Social Media Turned Humans from "Readers" to "Scanners"?

Think about the information you usually see on your smartphone.

Short social media posts.

Videos lasting only a few seconds.

News headlines.

Notifications.

Chats.

Before you can thoroughly consider one piece of information, the next one comes along.

Switching to different information every few seconds has become more common than spending 30 minutes or an hour on a single text.

Now, generative AI has joined the mix.

"Summarize this article in three lines."

"Tell me only the important parts of this book."

"Explain only the points likely to be on the test."

Indeed, it's convenient.

However, summaries always have "parts that were discarded."

What premises did the author start the discussion from?

What evidence was accumulated?

How were opposing views handled?

What doubts or ambiguities were there in reaching the conclusion?

Reading is not just about obtaining information.

It is the process of following the writer's thoughts and thinking for oneself.

If receiving only conclusions from AI becomes the norm, opportunities for this kind of training will undoubtedly decrease.

Are Digital Devices in Class Really Aiding Learning?

There is also concerning data in the OECD's PISA 2022.

About 30% of students in OECD countries reported being "distracted by digital devices in almost or all math classes."

Students who reported being frequently distracted scored an average of 15 points lower in math than those who were rarely distracted, even after adjusting for socio-economic background.

Of course, this does not prove causation.

Does using smartphones lower grades?

Are students with low interest in classes more easily distracted by digital devices?

It's not a simple matter to separate these factors.

However, one thing can be said:

"Simply increasing tablets or PCs in classrooms does not automatically improve the quality of education."

Technology itself does not improve education.

The value is determined by how, when, and for what purpose it is used.

On Social Media, Concerns That "Students Can No Longer Think Without AI"

Concerns about generative AI and education are being discussed quite specifically on social media.

 

According to the post, in group discussions, students often "repeat what AI said," during writing revisions, they "regenerate until it looks good," and in brainstorming sessions, they initially have AI generate a large number of ideas.

The post also points out that "some students don't start talking in groups without AI assistance."

This post received over 200 upvotes, indicating that many people share the confusion about university education in the AI era.

In another teacher-focused community, there are voices saying, "If students delegate the learning activity itself to AI, it's natural they won't acquire the concepts they were supposed to learn through that activity."

Some teachers have even reverted to paper-based assignments because even excellent students have started relying on AI in all situations and struggle to think of the main topic of a text on their own.

Of course, these are personal anecdotes from social media and not scientific research results.

Nonetheless, they provide insight into the changes that teachers and students are experiencing on the ground.

On the Other Hand, Some Say "It's Not Just AI to Blame"

There are also many opposing views on social media.

Rather than AI destroying education,

"The problem also lies with the education side that continues to assign tasks that can be easily handled by AI."

This perspective cannot be ignored.

Assigning students to write numerous reports without a clear purpose.

Having them memorize facts that can be quickly found through a search.

Continuing to assign the same homework themes for years.

If the task itself is aimed at "submission," it's natural for students to think about streamlining it with AI.

Even with the spread of calculators, math education did not disappear.

Even with the spread of word processors, writing education did not vanish.

Even with the advent of search engines, knowledge itself did not become unnecessary.

If we think of generative AI in the same way, the solution is not to simply retain old assignments and say "AI is prohibited."

There is a need to redesign the assignments themselves.

"Don't Use AI" vs. "Spot AI's Mistakes"

In fact, there are teachers on social media who deliberately incorporate AI into their classes.

They have AI write answers and then have students critique the content.

They have students verify the sources of information presented by AI.

They provide deliberately incomplete answers and ask students to find what's wrong.

They have AI write counterarguments and then have students further counter those.

With such uses, AI can become not a "machine to replace thinking" but a "partner for thinking."

Here lies a major turning point in education in the AI era.

Until now, schools have focused on

"Can you produce the correct answer?"

However, in a society where AI can produce plausible answers in large quantities,

"Can you judge whether that answer is truly correct?"

becomes more important.

"Liberal Arts" Gaining Attention Again in the AI Era

In this context, liberal arts are being re-evaluated.

In Japan, it is often associated with "general education" or "humanities," but it is originally much broader.

It involves thinking about a single issue from multiple perspectives while traversing literature, philosophy, history, politics, economics, social sciences, natural sciences, and mathematics.

The goal is not merely to increase the amount of knowledge.

"Is this information reliable?"

"Is this evidence or just an opinion?"

"What happens when viewed from a different perspective?"

"Is there an ethical issue?"

"Has there been a similar event in the past?"

"Can I explain my judgment to others?"

It is education that cultivates such skills.

As generative AI becomes capable of producing vast amounts of information, text, and images, these abilities become even more important.

On Social Media, "The Value of Liberal Arts Increases"

On LinkedIn, discussions linking the AI era with liberal arts are prominent.

Educators argue that critical thinking, ethical reasoning, communication, creativity, empathy, and systems thinking, which will become important in the AI era, are precisely the abilities that liberal arts have nurtured.

Another post finds the value of liberal arts in "pausing and taking time to think about a single issue."

This perspective is quite symbolic.

Current technology is generally moving in the direction of making everything faster.

Faster searches.

Faster writing.

Faster analysis.

Faster programming.

That's why, conversely,

"The ability to keep thinking without rushing to conclusions"

becomes rare.

As the cost of producing answers with AI decreases, the value of asking good questions, reading backgrounds, and spotting mistakes increases.

The Binary Choice of "Humanities or Sciences" Is Becoming Outdated

However, it's not about saying "because it's the AI era, you should only study the humanities."

On social media, there is also a discussion that simply retaining traditional liberal arts education is insufficient.

A minimum understanding of technology is necessary to think about AI ethically and socially.

What is data?

Why does AI make mistakes?

How do biases arise?

What data do AI services collect?

How much computing resources and power are used?

What issues are there with copyright and personal information?

Discussing ethics without understanding technology is not enough.

In other words, what is needed is not

"STEM or Liberal Arts"##HTML