Post
JA EN

Is 'Loving Programming' No Longer Enough? — A Multifaceted Look at the AI Era Debate

Posted
By
13 min read
Is 'Loving Programming' No Longer Enough? — A Multifaceted Look at the AI Era Debate
  • Target audience: Software engineers, technologists interested in AI-era careers
  • Prerequisites: Basic experience with AI tools like GitHub Copilot, ChatGPT, Claude
  • Reading time: 20 minutes

Summary

“If you love programming, don’t come to the IT industry anymore” — this provocative title sparked heated discussions on social media. The argument: with AI’s emergence, “technical skills alone” no longer provide differentiation.

On the other hand, there’s a counterargument that “the ‘love’ of programming is multi-layered.” The physical sensation of coding, the intellectual work of design, the creative act of development — AI’s impact differs for each layer.

This article organizes the background of this debate and examines it from multiple perspectives. The meaning of “love,” the changing skill sets in the AI era, and both the warnings and hopes from cognitive science. Rather than rushing to conclusions, we aim to provide materials for readers to find their own answers.

The Origin of the Debate — Two Articles

Article 1: “If You Love Programming, Don’t Come to the IT Industry”

The note article’s argument is clear1:

“Before LLMs emerged, the IT industry was a ‘paradise’ for socially awkward people like us”

The IT industry used to be an environment where one could thrive on technical ability alone, even without strong communication skills. However, with AI’s arrival, it’s become difficult to differentiate through “coding ability” alone. In an environment where technical skills alone don’t maintain competitive advantage, interpersonal skills become more important — that’s the article’s thesis.

Article 2: “The ‘Love’ of Programming is Multi-layered”

The Hatena blog article provides a more nuanced analysis2:

Programming “love” has multiple layers:

  • Physical sensation: The pleasure of typing code, the rhythm of striking keys
  • Intellectual work: The joy of design and problem-solving
  • Creative act: The feeling of building something with one’s own hands

The author notes that because “design thinking was the core” for them, their satisfaction remained even when delegating code generation to AI.

The article also points out the importance of viewing the relationship with AI as “collaboration” rather than “inspection.” Maintaining agency by articulating intentions and iterating feedback, rather than passively checking outputs. The author treats AI as “a capable but judgment-lacking junior colleague.”

Importantly, the author qualifies this as being “fortunate by chance.” They acknowledge the valid concerns of engineers who feel their creative agency has been taken away, and offer no definitive answer about career sustainability.

Why Is This Debate Emerging Now?

Background 1: Rapid Adoption of AI Coding Tools

According to the Stack Overflow Developer Survey 2025, 65% of developers use AI tools weekly or more[^3]. In 2024, 35% of developers thought “AI tools struggle with complex tasks,” but by 2025, this dropped to 29%.

Background 2: Impact on Junior Engineer Employment

Stanford University research reports that employment of software developers aged 22-25 declined by approximately 20% between 2022 and 20253. This timing coincides with the rise of AI coding tools.

Background 3: Growing Concerns About “Skill Atrophy”

One engineer’s testimony is symbolic3:

“I had been heavily using AI tools in my daily work, but when I started a side project in an environment without access to those tools, I noticed I was struggling with things that used to come naturally. ‘What was once instinct felt manual and, at times, tedious.’”

Background 4: Re-recognition of Soft Skills Importance

The World Economic Forum’s Future of Jobs Report 2025 reports that communication, leadership, and adaptability have seen the largest increases in employer priority over the past two years4.

In a UK survey, 74% of tech companies responded that they “value soft skills equally with technical knowledge”4.

Multifaceted Analysis — Four Perspectives

Perspective 1: How Is AI Changing Developer Productivity?

The note article’s claim — “it’s become difficult to differentiate through technical skills alone” — has some validity. However, this doesn’t mean “technical skills are no longer needed.” Rather, we should see it as the way technical skills are demonstrated is changing.

Multiple surveys report productivity improvements from AI tools.

According to JetBrains Developer Ecosystem Survey 2025, about 90% of developers using AI tools save at least one hour per week, with 20% saving eight hours or more — a full workday5.

Google’s RCT (2024), an experiment with approximately 100 Google engineers, found that the AI-using group completed tasks 21% faster than the control group (96 minutes vs 114 minutes)6. Interestingly, senior developers showed slightly larger effects than juniors — suggesting that experienced developers can leverage AI more effectively.

Faros AI’s telemetry analysis of over 10,000 developers reported that high AI-adoption teams handle 9% more tasks and 47% more PRs7.

Of course, there are studies showing different results. METR’s (2025) research reported that experienced developers using AI on familiar repositories took 19% longer to complete tasks. However, this study had a limited scale of 16 developers and 246 tasks, and the results are from the special condition of “familiar codebases”8.

Implication: The impact on productivity varies significantly by context (familiar vs new codebase), usage patterns (how much to trust AI), and individual aptitude. Generalizing “AI makes you faster” or “AI makes you slower” is difficult, and finding the right approach for yourself is important.

Perspective 2: Redefining “Love” — Asking What You Actually Love

The Hatena blog article’s point that “programming ‘love’ is multi-layered” is crucial.

flowchart LR
    subgraph Layers["Multi-layered Structure of Programming 'Love'"]
        direction TB
        L1["Physical Sensation Layer<br>(Keyboard input, typing rhythm)"]
        L2["Intellectual Work Layer<br>(Design, problem-solving, debugging)"]
        L3["Creative Act Layer<br>(Feeling of building with one's own hands)"]
    end

    subgraph AIImpact["AI's Impact"]
        direction TB
        I1["Automation of code generation"]
        I2["Support for analysis and suggestions"]
        I3["Human judgment still necessary"]
    end

    L1 --> I1
    L2 --> I2
    L3 --> I3

Impact on the Physical Sensation Layer: The joy of typing, the pleasure of seeing code appear on screen — this is certainly diminished by AI. Copilot auto-completes, Cursor rewrites. Those whose “love” is rooted in this layer are likely to feel a sense of loss.

Impact on the Intellectual Work Layer: Design, architecture, problem decomposition — AI can “support” these but can’t easily “replace” them. Those for whom this layer is central can potentially use AI as a “thinking partner.”

Impact on the Creative Act Layer: The feeling of “I made this” — this is subjective and varies by person. Some feel their work is “still mine” even when collaborating with AI, while others feel “it’s no longer mine” the moment AI is involved.

The important question: Which layer is your “love of programming” rooted in?

Perspective 3: Warnings from Cognitive Science — The Risk of Skill Atrophy

Cognitive science research warns of the risks posed by AI dependency.

Gerlich’s 2025 study (n=666) found a significant negative correlation (r = -0.68) between frequent AI tool use and critical thinking ability9.

“Through cognitive offloading as a mediator, higher frequency of AI tool use tends to correlate with lower critical thinking ability. Younger participants showed higher AI dependency and lower critical thinking scores compared to older participants”

Another study published in MDPI explains the mechanism of “cognitive atrophy”10:

“Prolonged reliance leads to cognitive passivity. Users cease to validate, interpret, and reconstruct AI-generated information, accepting algorithmic outputs as epistemic authority. The brain no longer rehearses analytical processes and progressively loses the capacity to regenerate them without external support”

However, there is hope.

The same research group emphasizes the difference between using AI as a “scaffold” versus a “substitute”10:

  • Scaffold: Temporary, adaptive, empowering — aims to strengthen internal abilities and gradually reduce dependence on technology
  • Substitute: Permanent, dependent — technology assumes responsibility for adjustment, reducing intrinsic skills

Implication: The question is not “whether to use AI” but “how to use it.”

Perspective 4: Changing Skill Portfolios

So what is becoming required? The direction indicated by multiple surveys is consistent.

Changes in Technical Skills:

  • Code generation ability → AI output evaluation and verification ability
  • Mastery of specific languages → Cross-tool understanding
  • Individual implementation ability → System design with AI collaboration

Importance of Soft Skills: According to McKinsey research, demand for social and emotional skills is predicted to increase by 26% in the US and 22% in Europe between 2016 and 2030. Meanwhile, demand for basic data entry and processing skills is predicted to decrease by 19% in the US and 23% in Europe over the same period11.

Skill sets considered important in the AI era4:

  1. Technical skills: Data analysis, AI/ML understanding, systems thinking
  2. Human skills: Empathy, creativity, emotional intelligence
  3. Meta skills: Communication, problem discovery, dealing with ambiguity

A Framework for Constructive Thinking

It’s not productive to frame this debate as “programming lovers are finished” vs “everything’s fine, nothing changes.”

Framework 1: Decomposing Your “Love”

flowchart LR
    subgraph Question1["Question 1: What do you love?"]
        direction TB
        Q1A["The act of writing code itself"]
        Q1B["The problem-solving process"]
        Q1C["The achievement of making something work"]
        Q1D["Continuously learning technology"]
    end

    subgraph Question2["Question 2: How does AI change that?"]
        direction TB
        Q2A["Gets replaced"]
        Q2B["Gets supported/accelerated"]
        Q2C["Can be realized in new ways"]
        Q2D["Not affected"]
    end

    Q1A --> Q2A
    Q1B --> Q2B
    Q1C --> Q2C
    Q1D --> Q2D

AI’s impact differs depending on which layer your “love” is rooted in. Not everything is “finished” nor “fine.”

Framework 2: Using AI as a “Scaffold”

Based on cognitive science insights, using AI as a “scaffold” rather than a “substitute” may prevent skill atrophy while improving productivity.

Examples of using AI as a “scaffold”:

  • Form your own hypothesis before having AI generate an answer
  • After receiving AI output, verify “why this works”
  • Regularly challenge the same tasks without AI to confirm your abilities
  • Treat AI not as “something that produces answers” but as “a conversation partner”

Examples of AI becoming a “substitute”:

  • Ask AI before thinking
  • Adopt AI output as-is
  • Move forward without understanding “why it works”
  • Unable to do anything without AI

Framework 3: Reconsidering “T-shaped Skills”

Rather than relying on a single pillar of “loving programming,” be conscious of a T-shaped skill structure.

flowchart TB
    subgraph TShape["T-shaped Skills"]
        direction TB
        H["Horizontal bar: Broad foundation<br>(Communication, domain knowledge, business understanding)"]
        V["Vertical bar: Deep expertise<br>(Technical ability including AI evaluation skills)"]
    end

    H --> Value["AI-era value"]
    V --> Value

With a horizontal bar (broad foundation), you can adapt even if the shape of the vertical bar (deep expertise) changes. Conversely, relying only on the vertical bar makes you vulnerable when that field is replaced by AI.

Don’t Rush to Conclusions

The claim “if you love programming, don’t come to the IT industry anymore” has an element of truth. At the same time, the counterargument that “programming ‘love’ is multi-layered, and the impact varies by person” is also valid.

What we can say with certainty:

  • AI is automating code generation, and the value of “writing code” is changing
  • The importance of soft skills, especially communication and problem discovery, is increasing
  • Excessive AI dependency carries risks of cognitive decline
  • However, using AI as a “scaffold” may improve productivity while maintaining skills

What remains uncertain:

  • How fast this change will progress
  • How much “loving programming” as a motivation will support careers in the future
  • What the optimal adaptation strategy is for each individual

The most important question:

The “correct answer” to this debate differs for each person. What’s important is perhaps having your own answers to these questions:

  1. What does your “love of programming” mean?
  2. How is that “love” affected by AI?
  3. How will you compensate for the affected parts and leverage the unaffected parts?

See other articles related to this theme:

References

References are listed in order corresponding to citation numbers in the main text.

[^3]: [AI2025 Stack Overflow Developer Survey](https://survey.stackoverflow.co/2025/ai) - Stack Overflow (2025). [Reliability: High]

Additional References (Not Numbered in Main Text)

On Citation Accuracy: The research cited in this article has been verified through the following methods:

  • Confirmation via academic databases (Google Scholar, MDPI, etc.)
  • Information verification on official journal and survey organization websites
  • Cross-verification through multiple independent sources

Some blog articles are cited as primary sources with inherent limitations but are included to illustrate the context of the debate.

  1. プログラミングが好きな人は、もうIT業界に来るな - igz0 (2026). note. [Reliability: Medium] ↩︎

  2. プログラミングの「好き」は多層的 - syu-m-5151 (2026). Hatena Blog. [Reliability: Medium] ↩︎

  3. AI coding is now everywhere. But not everyone is convinced. - MIT Technology Review (2025). [Reliability: High] ↩︎ ↩︎2

  4. Developer Soft Skills: 85% of Career Success - IT Support Group (2026). Citing World Economic Forum Future of Jobs Report 2025. [Reliability: Medium-High] ↩︎ ↩︎2 ↩︎3

  5. The State of Developer Ecosystem 2025 - JetBrains (2025). [Reliability: High] ↩︎

  6. The reality of AI-Assisted software engineering productivity - Citing Google RCT (2024). [Reliability: Medium-High] ↩︎

  7. What METR’s Study Missed About AI Productivity in the Wild - Faros AI (2025). [Reliability: Medium-High] ↩︎

  8. Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity - METR (2025). n=16, 246 tasks. Results under special conditions of working on familiar repositories. [Reliability: High] ↩︎

  9. AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking - Gerlich, M. (2025). Societies, 15(1), Article 6. DOI: 10.3390/soc15010006. [Reliability: High] (Peer-reviewed) ↩︎

  10. Cognitive Atrophy Paradox of AI–Human Interaction: From Cognitive Growth and Atrophy to Balance - MDPI Information (2025). [Reliability: High] (Peer-reviewed) ↩︎ ↩︎2

  11. Skill Shift: Automation and the Future of the Workforce - McKinsey Global Institute (2018). [Reliability: High] ↩︎

AI・Technology
AI Career Programming Skills Identity Cognitive-Science
This post is licensed under CC BY 4.0 by the author.