Introduction

PwC’s latest research reveals that AI is not taking jobs but is instead tripling employee productivity, with a 2312% increase in database engineering positions. From recruitment to performance, AI is reshaping corporate operations.

Recently, Anthropic launched a blog called “Claude Explain,” primarily generated by their AI model, Claude.

The blog covers hardcore topics like “Simplifying Complex Codebases with Claude,” showcasing AI’s writing capabilities. However, this is not AI running wild. A spokesperson stated that all content is vetted by Anthropic’s experts and editorial team, who add insights, real-world examples, and background knowledge. The process requires human expertise and multiple iterations.

From a technical standpoint, “Claude Explain” demonstrates a collaborative approach: Claude creates educational content while the team reviews, refines, and optimizes it.

However, visiting the blog’s homepage might be confusing, as it appears to be entirely managed by Claude. Anthropic clarified that the blog aims to demonstrate the “human + AI” collaboration model, starting with educational content.

This is like giving work a “boost” with AI, providing users more value. We are not replacing humans but enabling experts to accomplish more with AI. Future topics will include creative writing, data analysis, and business strategies.

AI writing is not a new phenomenon. Meta’s Mark Zuckerberg aims to develop an end-to-end AI advertising tool, while OpenAI has created a model specifically for writing creative copy, claiming “AI can handle 95% of marketers’ tasks.”

Publishers are also experimenting with AI for news writing:

• Gannett uses AI for sports summaries and headlines.
• Bloomberg adds AI-generated summaries to articles.
• Business Insider laid off 21% of its staff while encouraging reporters to use AI for writing.
• The New York Times encourages staff to use AI for editing suggestions, headlines, and interview questions.
• The Washington Post is developing an AI-driven “story editor” called Ember.

However, many AI writing attempts have faced challenges, often producing inaccurate or misleading content. Business Insider recommended non-existent books and had to apologize, Bloomberg corrected numerous AI-generated errors, and G/O Media faced backlash for publishing flawed AI-written articles.

Anthropic, despite using AI for blogging, is still hiring for marketing, content, and editorial positions.

The Short-Lived AI Blog

Unexpectedly, the blog was shut down after just one month. Anthropic had enlisted experts and editors to refine Claude’s drafts, ensuring reliable content without AI inaccuracies. However, the blog’s performance was underwhelming.

Initially, the blog had its own page on Anthropic’s website and posted several articles, receiving links from 24 other sites within a month—decent performance. Yet, one weekend, Anthropic redirected the blog’s URL to the homepage, removing all articles.

The feedback on social media was not favorable. Some viewed the blog as a marketing gimmick, while others criticized Anthropic for not clarifying which content was written by Claude and which was edited by humans. The question remained: how much of the writing was truly AI-generated?

Moreover, AI writing is a minefield. Even top-tier AI like Claude can produce nonsensical content, leading to potential PR crises. Anthropic likely feared that Claude might generate something outrageous. The blog may have been a trial run, aiming to combine user-desired practical content with marketing goals. However, with lukewarm feedback and risks looming, they chose to pull the plug.

This decision seems wise—it’s better to cut losses than to risk further issues.

AI Enhancing Work Efficiency, Not Replacing Jobs

PwC’s latest research indicates that generative AI has not cost us jobs but has significantly upgraded the workplace. Particularly in coding and software development, AI has become a company’s “secret weapon” for profitability.

PwC analyzed nearly 1 billion job postings globally and found that since 2022, industries utilizing AI have seen employee productivity triple. This indicates that companies’ investments in AI are paying off.

Overall, AI is not leading to mass layoffs but is reshaping work models, increasing efficiency, wages, and skill requirements. This contradicts Anthropic’s CEO’s previous claim that AI could eliminate half of entry-level jobs in the next 1-5 years, raising unemployment to 10-20%.

The newly released Claude Code can write and debug code, nearly matching programmers’ capabilities.

Natural language programming is becoming popular and is likely to grow. How significant is AI’s impact on employment? Companies are hiring aggressively to feed AI, resulting in a 2312% increase in database engineering positions and a 382% rise in demand for statisticians.

Experis North America’s Kye Mitchell stated:

AI is not about taking jobs; it’s about fundamentally changing how we work. Technology will only truly replace a job when 80% of its tasks can be fully automated. Currently, most jobs are far from that point. AI enhances our skills, speeds up work, and allows us to focus on more valuable tasks.

PwC found that AI is impacting not just high-tech industries but also traditional sectors like mining and agriculture. Wages in AI-affected industries are rising twice as fast as in others, with workers possessing AI skills earning 56% more than their peers, a 31% increase from last year.

The speed of skill requirement changes is accelerating, with AI-affected positions experiencing a 66% faster evolution in skill demands. Despite an overall downturn in the job market, AI-related job postings continue to rise. The information and communication sector is aggressively seeking AI talent, while construction and healthcare are lagging.

In the U.S., positions closely related to AI are seeing increasing skill demands, indicating that AI is redefining roles rather than causing job losses.

In recent years, companies have become less focused on degrees when hiring. As of April this year, about half of tech job postings no longer required a bachelor’s degree.

The requirement for a bachelor’s degree in AI-related positions has dropped from 63% in 2019 to 53% in 2024. While degree requirements have eased, companies are pivoting towards AI. Due to uncertainty, the U.S. tech industry reduced 214,000 jobs in April.

To retain jobs or achieve promotions, it’s essential to quickly learn AI skills.

No AI, No Job!

Luis von Ahn, CEO of Duolingo, warned 900 employees that AI is now a top priority for the company. In an April memo, he stated:

• Tasks that can be handled by AI will no longer be outsourced.
• AI skills will be prioritized in hiring.
• AI usage will be included in performance evaluations.
• New hires will not be recruited unless AI truly cannot handle the task.

Some criticized AI for lacking human context in translation, arguing that language learning requires a human touch. Users threatened to uninstall the app, and others accused the company of prioritizing AI over employees.

In response to the backlash, Luis von Ahn quickly clarified:

Don’t get me wrong; I’m not trying to replace employees with AI. Our hiring speed remains the same as before. AI can accelerate work speed and improve quality. The sooner you learn to use AI, the better your future will be.

Companies like Meta, Shopify, and Box are adopting “AI-first” strategies for risk assessment, hiring, and performance evaluation. Meta uses AI for product privacy risk assessments, automating simple decisions to free teams for complex issues.

Shopify’s CEO stated that everyone must learn how to use AI for tasks like prototyping and writing performance reports, and before hiring new staff, they must prove that AI cannot do the job.

“AI will disrupt Shopify and our lives; we must go all in!” This statement went viral on social media.

Now, Google allows engineers to use AI tools during interviews to solve problems.

However, some companies have realized that AI is not a panacea. Swedish financial company Klarna cut 38% of its workforce, replacing over 800 full-time positions with AI assistants, only to find that AI struggles with tasks requiring human empathy and complex judgment.

AI is not infallible, especially in situations requiring nuanced communication; humans are still needed. The realization that “technology must be strong, but human touch must be stable” has emerged.

The CEO later admitted that they had overreached, and while continuing to invest in AI, they also rehired many gig workers.

Bloomberg Beta investor Roy Bahat stated that new companies must integrate AI into their core logic to keep up. The key is not to use AI to replace people but to teach people to collaborate with AI, delegating repetitive tasks to machines while humans focus on more creative endeavors.

Is AI Diminishing a Generation of Students?

However, there is a significant difference between “knowing how to use AI” and “only knowing how to use AI.” Recent media interviews with many teachers revealed that AI has become a cheating tool for students.

Companies are aggressively promoting AI tools to students, undermining teachers’ efforts to limit AI use in classrooms. When students log into their school’s Microsoft system, Copilot prompts them to summarize reading materials or draft essays.

A teacher in Los Angeles lamented, “About 40% of the assignments I grade are written by AI.”

An online English teacher expressed frustration, stating that some students treat AI as their brain!

During a class discussion, one student said, “Let me think,” then muted themselves, likely searching for answers on their phone. When they finally spoke, their response was hesitant, clearly having looked up the answer.

Research shows that increased use of ChatGPT correlates with poorer grades, diminished memory, and a decline in critical thinking skills. “My students no longer think for themselves; they lack interests and hobbies.”

A 12th-grade English teacher expressed anger, stating, “They have no original ideas and merely parrot what they hear on TikTok.”

Teachers are understandably upset, facing an AI-driven trend pushed by large companies that they cannot control. Ultimately, as children become accustomed to letting AI do the work, they risk losing not only problem-solving skills but also the essence of independent thought.