Conclusion First: “High Labor Cost Jobs” Can No Longer Be Protected

Amazon has deployed over 1 million robots in its warehouses. They work 24/7 without complaints or injuries. The cost of this system, which supports same-day and next-day delivery, is decreasing year by year.

Meanwhile, AI interviews have entered the hiring process, and AI surveillance has begun quantifying employee performance.

When we line up these three pieces of news, a structure becomes clear: the price of jobs that were once high because they were performed by humans is being rewritten on a weekly basis.

This is not just a story for large corporations. Local small and medium-sized enterprises (SMEs) must confront this structural change now.

—

The True Impact of Amazon’s 1 Million Robots

Reports indicated that Amazon had 750,000 robots by 2024, but by 2025, that number surpassed 1 million. The number of employees stands at about 1.5 million. This means we are now at a ratio of “1.5 humans per robot.”

What’s noteworthy is that despite the introduction of robots, Amazon has not significantly reduced its workforce. Robots handle heavy lifting and simple picking tasks, while humans are assigned to “exception handling that requires judgment” and “complex packaging that robots cannot manage.”

However, there is a catch here.

The remaining human jobs are required to align with the productivity of the robots.

On a line where robots pick 300 items per hour, if the human-operated process only manages 50 items per hour, the entire operation gets jammed. As a result, human work pace is set according to robot standards. Labor advocacy groups point out that “as automation progresses, the pressure on the remaining humans increases”—this is the structure at play.

Reports indicate that the injury rate among Amazon warehouse workers is higher than the industry average. While it is true that robots have “reduced dangerous jobs,” the increased density and speed of the remaining work have created a different kind of burden.

The implication for SMEs is this: **”It’s only half true that introducing robots and AI will make things easier. If you do not redesign human jobs to align with the automated parts, the workplace will break down.”

—

AI Interviews Expose the “Collapse of Skill Value”

In the UK, AI interviews are rapidly penetrating the hiring processes of major companies. AI analyzes candidates’ responses, expressions, and tone of voice, scoring them. HR personnel no longer need to interview 100 applicants one by one.

In terms of cost, the labor cost for traditional first interviews was between 5,000 and 10,000 yen per person (interviewer’s hourly wage plus adjustment costs). For 100 people, that amounts to 500,000 to 1 million yen. AI interview tools cost tens of thousands to over a hundred thousand yen per month and can process any number of candidates. Interview costs have dropped to less than one-tenth.

However, an interesting phenomenon is occurring here.

Candidates have also begun using AI. They generate model answers with ChatGPT and provide “AI-friendly responses” during AI interviews. One UK recruiter mentioned that “there is a suspicion that over 30% of candidates are using AI-generated responses.”

In other words, a structure of AI interviews vs. AI responses has emerged. This leads to a rapid decline in the value of skills that can be measured in interviews. Memorized knowledge, templated responses, and standardized problem-solving abilities—all of these become nearly worthless in the hiring market once “AI can answer them instead.”

Conversely, what increases in value? The human trust that makes someone say, “I want to work with this person,” proven records of handling unexpected situations on the ground, and concrete results that can be quantified. Only “the individual’s own experience and judgment,” which AI cannot generate, can maintain value in the hiring market.

This is good news for SMEs in hiring. As large companies advance efficiency with AI interviews, the strength of SMEs in hiring—”you can understand by meeting and talking”—increases relatively. The CEO can meet directly, show the workplace, and think together about “what can be done here.” This hiring process cannot be replaced by AI.

—

How AI Surveillance Lowers the Value of “Ordinary Jobs”

In the UK, the supermarket chain Co-op has implemented a system where employees’ job performance is constantly monitored by AI. The speed of checkout processing, inventory efficiency, and length of breaks—all are quantified.

What this means is clear. The moment “ordinary jobs” are quantified, it proves that “anyone can do it the same way.”

The speed of checkout can be quantified. Person A serves 80 customers per hour, while Person B serves 65. When this difference is visualized, companies use the person who can process 80 as the standard and demand training or reassignment for those below that threshold. Further down the line, the decision may be made to “automate 80 customers with self-checkouts and AI monitoring.”

Jobs that can be quantified will eventually be targeted for automation, sooner or later. This is not just a story for large corporations. Even local SMEs can automate quantifiable tasks like attendance management, inventory counting, daily report input, and invoice processing with SaaS solutions costing a few thousand to tens of thousands of yen per month.

Let’s look at some specific examples:

• Invoice Processing: Traditionally, accounting staff spent 20 hours a month on this task, which can now be reduced to 2 hours with AI-OCR and accounting software integration. This translates to a cost reduction from 40,000 yen to 4,000 yen.
• Inventory Management: What took 15 hours a month for manual counting and Excel input can now be done in 1 hour with barcode and cloud management.
• Daily Reports and Documentation: 30 minutes per person for 20 days, totaling 10 hours a month, can be reduced to 2 hours with voice input and AI summarization.

The costs of these “quantifiable jobs” have already dropped to less than one-tenth. The question is whether you are aware of this fact. If you are not, you will continue to pay labor costs for “jobs that could be done by AI for a few thousand yen” for many hours each month.

—

SMEs’ Criteria for Judgment: What Jobs Cannot Exist Without Humans?

When we overlay these three pieces of news, the criteria for judgment becomes clear.

“Can this job be quantified?”

If yes, it will be replaced by AI or robots sooner or later. Costs will continue to decline. Continuing to invest labor costs in that area is akin to burning money every month.

If no, it is a candidate for “jobs that cannot exist without humans.” However, there is one more filter.

“Does this job require a relationship of trust to exist?”

This is where local SMEs can win against large corporations.

They know the faces of their clients’ CEOs. They understand the quirks of the craftsmen on-site. They remember the family structure of their customers. This accumulation of “relationships” cannot be replicated by AI. Moreover, as large corporations push for AI and efficiency, the scarcity value of this “human connection” increases.

Specifically, here are three things SMEs should do right now:

1. Identify Quantifiable Tasks and Calculate Their Monthly Costs

Accounting, inventory management, daily reports, schedule adjustments, and quotation creation—how many hours and how much are you paying for these? Many SMEs do not even grasp these numbers. Start here.

2. Replace Tasks That Can Be Automated for Less Than 50,000 Yen Per Month

There is no need to implement a large system all at once. Start with SaaS solutions costing a few thousand yen per month. For invoice processing, use invoice OCR; for daily reports, use voice input plus AI summarization; for schedule adjustments, use calendar AI. “Try one first, starting next week” is the right approach.

3. Reinvest Saved Time into Building Trust

Use the 20 hours saved from automation for visiting existing customers, talking with new clients, and having one-on-one meetings with employees. This will become the true competitive strength of SMEs. As large companies streamline customer interactions with AI chatbots, the value of “the CEO calling directly” or “the representative visiting the site” increases.

—

The Value of “Human Jobs” Will Not Return to Previous Levels

Amazon’s 1 million robots. The proliferation of AI interviews. The normalization of AI surveillance.

The simultaneous progression of these three factors means that the price of jobs that were high because they were performed by humans will structurally continue to decline. Week by week, month by month, with each new tool, the number of jobs that “no longer need to be done by humans” will increase.

There is no point in resisting this trend.

What we should be asking is, “What jobs in our company will increase in value when done by humans?”

The answer is usually found on the ground. It lies in conversations with customers. It is in the voices of employees saying, “We need that person.”

Start by automating just one task next week. Then, use the time saved to meet with customers.

That is where the path for SMEs to survive lies.