A lawyer’s hourly rate is $500. Yet this lawyer is now working at $400 per hour to teach an AI model the ‘correct legal answers.’

Local small business owners have yet to grasp the implications of this fact. However, it signals a fundamental shift in the structure of ‘money paid to experts.’

What is Happening—Experts Becoming ‘Teachers’ for AI

Major AI companies like OpenAI, Anthropic, and Google require massive amounts of ‘human expert feedback’ to improve their models’ accuracy. Only experienced doctors and lawyers can determine whether the AI’s answers to legal questions are correct or if there are errors in medical diagnostic logic.

This has given rise to a market for ‘AI gigs for experts.’ Platforms like Scale AI, Invisible Technologies, and Alignerr are offering tasks to doctors, lawyers, and engineers at hourly rates ranging from $100 to $400. According to reports from Bloomberg and Business Insider, some lawyers are earning between $5,000 and $10,000 a month from these gigs alone.

Why are they flocking to gigs despite having primary jobs? The reasons are simple:

• Flexibility in time. They can work at their convenience without being tied to clinics or law offices.
• Hourly rates comparable to their primary jobs. An hourly rate of $400 is nearly on par with the billable rates of mid-level lawyers. There are no travel or client interaction costs, allowing them to earn purely from their knowledge.
• Explosive demand. AI companies have raised hundreds of billions of dollars since 2023, with a portion flowing into ‘human feedback.’

In other words, expert knowledge is beginning to shift from ‘one-to-one service provision’ to ‘one-to-infinity AI training data.’ This is the crux of the structural change.

The Real Concern is ‘What Comes Next’—AI Starting to Replace Experts

Experts are training AI. So, what will a sufficiently trained AI do? It will start to replace expert jobs.

Signs of this are already emerging:

• Legal Field: Anthropic plans to release an experiment in 2025 where AI agents autonomously negotiate contracts between companies. This demo shows AI completing contract reviews that previously took lawyers dozens of hours in just a few minutes. Several U.S. law firms have reported a 60-80% reduction in the labor hours for legal research after implementing AI.
• Accounting and Taxation: U.S. startups like Pilot and Bench are handling the accounting of small businesses for a few hundred dollars a month, using AI and a small number of human staff. In Japan, companies like freee and Money Forward are rapidly increasing the accuracy of AI bookkeeping, shaking the cost structure of ‘outsourcing to tax accountants for 50,000 yen a month.’
• Medical Field: Google’s Med-PaLM 2 scored at a specialist level on the U.S. Medical Licensing Examination (USMLE). AI has already entered practical use in initial triage and image diagnostics.

Let’s summarize what is happening:

1. Experts train AI (the $400 gig).
2. The trained AI cheaply replaces expert jobs.
3. The ‘hourly rate as service providers’ for experts declines.
4. Outsourcing costs for small businesses structurally decrease.

This is not a temporary trend. Experts themselves are nurturing the AI that will take their jobs, marking an irreversible structural transformation.

How Much Will Outsourcing Costs for SMEs Decrease—Thinking in Concrete Numbers

So, what is the specific impact of this on local small businesses?

Let’s look at the typical costs that Japanese SMEs pay to experts:

This is not an optimistic forecast. The realities in web production and translation are already occurring. While complete automation in legal, tax, and labor fields will take longer due to regulatory and liability issues, transitioning to a model where ‘AI drafts → experts do the final checks’ will drastically reduce the working hours of experts. As working hours decrease, costs will naturally fall.

If a small business with an annual revenue of 100 million yen spends 2 to 4 million yen annually on tax advisors, labor management, spot legal services, and web-related tasks, achieving the above reduction rates would mean savings of 1 to 3 million yen per year. For a company with 10 employees, this amount is equivalent to one person’s salary.

It’s Not Just About ‘Becoming Cheaper’—The Real Change for SMEs

The impact of reduced costs is not merely about savings. A qualitative change occurs where ‘things that were previously out of reach become attainable.’

Here are some examples:

• Legal checks on contracts. Contracts that were previously handled in a makeshift manner because ‘it costs 50,000 yen’ can now be checked instantly by AI. The quality of trouble prevention dramatically increases.
• International Transactions. If translation costs become almost zero, local manufacturers can prepare materials for overseas exhibitions in ten languages. A town factory with an annual revenue of 500 million yen can suddenly become a global player.
• Specialized Analysis. Hiring a data analysis expert for 50,000 yen a month was impossible. However, with AI tools, trend analysis of sales data and demand forecasting can be done for just a few thousand yen a month.

In other words, the reduction in costs eliminates the ‘reasons not to do it.’ This is the true game changer for small businesses. Large companies have been able to hire experts for a long time. Small businesses have given up because ‘it’s too expensive.’ AI flattens that disparity.

Small businesses stand to gain significantly from this change.

So, What Should Be Done?

Simply waiting for ‘AI to evolve and become cheaper’ will not change anything. What can be done starting today is clear.

1. First, try replacing one outsourced task with AI.

Whether it’s checking contracts, summarizing meeting minutes, bookkeeping, or drafting recruitment materials, it doesn’t matter. Throw it at ChatGPT or Claude and see for yourself, ‘How does this compare to the quality of what I used to outsource?’ There are bound to be surprisingly usable cases.

2. Change contracts with experts from ‘full outsourcing’ to ‘AI checks + expert reviews.’

I’m not saying to cut ties with tax advisors or labor management consultants immediately. Just changing the model to ‘AI does 80%, and the remaining 20% is reviewed by an expert’ could potentially halve costs. Experts are also beginning to adapt to this trend. Those who cannot adapt will soon be eliminated.

3. Use the savings for ‘offensive’ purposes.

Cost reduction is not the goal. The real challenge is what to do with the 1 million yen saved. Experimenting with new businesses, hiring, and capital investment. Small businesses that can convert defensive costs into offensive investments will grow in the next five years.

This is Not a ‘Future That Will Come Eventually’

The fact that doctors and lawyers are flocking to AI gigs means that the value of specialized knowledge is shifting from ‘scarcity’ to ‘reproducibility.’ When something that was valuable because it was scarce becomes replicable by AI, prices will collapse.

This structural change presents the biggest opportunity for local small businesses in the past 30 years. An era is approaching where legal, accounting, and marketing services of the same quality as those provided by major companies in Tokyo can be obtained at one-tenth the cost.

The question is whether to recognize this change and act or to continue paying the same costs without realizing it. The difference will become an irreparable gap in competitiveness in three years.

I hope you will try just one thing today.