Lawyer Lin Shanglun's article: AI becomes the universal excuse for layoffs! Tech companies' collective scapegoating

Large tech companies are conducting mass layoffs citing "AI replacing human labor," but from the perspective of frontline users, AI is currently far from capable of fully taking over professional work. The real reason is that over-expansion and misallocation of capital during the pandemic from 2020 to 2022 are coming due; AI's true role is as a "process accelerator," not a "full automation replacer." Using AI as an excuse for layoffs not only creates social panic but also distorts the correct application of AI.
(Background: "Silicon Valley Light Talk" Kenji announced quitting Phantom Wallet outright! Taking at least 5-10 years off, feeling nothing when salary hits the account)
(Additional context: Over 1,500 Meta employees jointly protested! Demanding to narrow the scope of "AI monitoring keyboard and mouse," allowing half an hour of daily pause)

Table of Contents

Toggle

• The flaw in the "distillation" myth: AI isn't yet powerful enough to cut so many people at once
• The real reason is the misallocation of capital over the past few years coming to light
• Everything is built on "At Will"
• The true role of AI: Accelerate, not replace
• Example: How are top US law firms using AI?
• The harm caused by using AI as a pretext for layoffs
• Give AI an honest narrative

In the past year or two, large tech companies have launched wave after wave of astonishing layoffs. Restructurings involving thousands or even tens of thousands of employees have become almost routine quarterly events. And each time, the official explanation increasingly revolves around one key word: AI.

"We are reshaping our organization with AI." "We will deploy human resources more intelligently." "We believe AI will bring a new productivity paradigm." These phrases are repeatedly copied and pasted, even evolving into more sensational versions—such as claims that the knowledge of senior employees is being "distilled" into models, allowing AI to take over their original work so these employees can be "gracefully let go."

It sounds very futuristic. But as a lawyer who has long used industry-leading AI tools, I believe: this story is highly questionable.

The flaw in the "distillation" myth: AI isn't yet powerful enough to cut so many people at once

I have been using cutting-edge AI tools for a long time, especially those designed to optimize legal workflows. From the frontline user’s perspective, current AI is nowhere near capable of "distilling a senior professional and having the model operate independently."

Why? Because AI's current focus is not on "replacing humans." The idea that you can compress a person into a model, which then automatically produces professional-level results—this simply doesn’t happen in practice. AI's current capabilities are fundamentally as a "powerful accelerator": it helps professionals quickly organize data, generate initial drafts, compare large volumes of files. But final judgments, strategies, and conclusions still require human decision-making.

If AI were truly capable of replacing thousands of engineers, designers, and operations staff at once, it should also be able to replace thousands of lawyers, accountants, and doctors—yet we all know that this has not happened at all.

Therefore, when companies tell you "because of AI, we no longer need these people," that narrative is actually very weak. It cannot withstand scrutiny from those who are actually using AI on the front lines.

The real reason is the misallocation of capital over the past few years coming to light

So why this wave of large-scale layoffs? If we rewind to 2020–2022, what was the scene? The pandemic caused a surge in digital demand, and every major tech company was frantically expanding. Doubling or tripling their workforce within two to three years was not unusual but the norm. At the same time, stories about the metaverse, Web3, VR/AR, autonomous vehicles, and other "next decade" visions kept coming, with capital investments reaching hundreds of millions or billions of dollars.

Many of these bets are now unrecoverable. The metaverse did not arrive as expected, VR device markets fell short of expectations, autonomous vehicle timelines were repeatedly delayed, and pandemic-driven benefits waned as life returned to normal. In other words, the current wave of layoffs in big tech is fundamentally a correction for over-expansion and wrong bets made in recent years—not proof of AI replacing humans.

Companies need a decent reason to justify cutting ten thousand jobs. Saying "we made a wrong call back then" is too hard to admit. Saying "we are embracing AI" sounds much better—aligning with capital markets’ tastes, stabilizing stock prices, and shaping a "future-facing" image. All in one move.

Everything is built on "At Will"

There is a structural background that Taiwanese readers especially need to understand. In the highly capitalist labor framework of the US, most employment contracts are "At Will." Employers can dismiss employees at any time, and employees can leave at any time, with minimal reasons required—this flexibility is built into the system.

In contrast, Taiwan? It’s practically impossible to sign such arbitrary employment contracts. Most employment relationships there are standardized, indefinite-term arrangements. These indefinite-term relationships are, frankly, quite frightening in Taiwan—dismissing an unfit employee is very difficult, costly, and can lead to retaliation or lawsuits after resignation, making the process very troublesome.

From this perspective, the wave of large-scale layoffs is essentially the extreme manifestation of the At Will system. For US companies, this flexibility is an advantage in global competition. But the question is—if you have At Will, why pretend that AI is the main reason? Why not just openly admit that it’s a business decision, organizational restructuring, or correction of past misjudgments?

The true role of AI: Accelerate, not replace

AI’s real strength has always been in accelerating and optimizing existing human workflows.

Even in the engineering sector, where AI impact is most anticipated, many large companies now require engineers to use AI for coding. But after AI writes code, what then? Engineers still need to debug, review architecture, and judge logic errors. AI makes them more efficient, not obsolete.

To help everyone better understand AI’s real role in frontline professional work, I can share a workflow I discussed with a US law firm that is actually in operation.

Example: How are top US law firms using AI?

Suppose there’s a patent infringement lawsuit involving a large number of patent documents, or a case where two companies accuse each other of patent infringement. These cases involve massive files: tangled patent descriptions, expert reports, evidence of infringement. In the past, lawyers had to read each document one by one.

Now, the approach is: let AI read everything first. What happens after AI has processed all the data? It marks every key point. For example, if you ask "When did the infringement occur?" AI will directly tell you: the specific actions of the accused infringer, the relevant communications, all presented at once.

But note: up to this point, AI is only "organizing" and "presenting." These initial drafts must be reviewed by lawyers. Every question, every follow-up, results in a new version of the organized data—restructured according to the lawyer’s needs. Then, the lawyer further writes legal claims (e.g., "The defendant infringed at this point") and finally drafts the complaint.

In other words, AI’s role in this process is as a powerful "pre-processor" and "data organizer," but the critical decisions—whether to claim infringement, how to construct legal arguments, which litigation strategy to adopt, the final conclusion of the complaint—are still firmly in the hands of the lawyer. Without legal judgment, AI-organized data remains just categorized information, unable to become a battle-ready legal document.

This is the true picture of cutting-edge AI workflows today. It’s not a sci-fi scenario where "distilling a lawyer and automatically producing a complaint"—it’s about "freeing lawyers from tedious data sorting, allowing them to focus on the most judgment-intensive parts."

The harm caused by using AI as a pretext for layoffs

Using AI as a pretext for layoffs directly causes misinformation and panic. When big tech companies claim "AI has replaced you," society is misled into believing AI is already powerful enough to fully take over human jobs. This fuels massive job anxiety, generational conflicts, and fears about the future. These emotions can even influence education choices, career planning, and societal attitudes toward new technology.

More problematic is that this narrative creates false expectations. Many small and medium-sized business owners see headlines about big companies "cutting jobs with AI" and think they can buy a few AI subscriptions, import some tools, and start laying off staff to save costs. But after implementation, they find AI doesn’t replace anyone; instead, it requires more skilled staff to operate, verify, and correct it. When expectations are dashed, AI becomes a scapegoat, and the implementation plan fails.

The deepest harm is the propagation of incorrect usage. When users initially treat AI as a "full automation replacer" rather than a "process accelerator," they interact with it wrongly—asking vague questions, expecting perfect answers, not checking, not following up, not correcting. The result is a pile of half-truths, leading to complete disillusionment with AI and missing out on its true value. A tool that could be a powerful assistant becomes misunderstood, misused, and undervalued due to false narratives.

Give AI an honest narrative

I am not opposed to layoffs by big tech companies. Under the At Will system and given the over-expansion in recent years, organizational restructuring is a reasonable business decision. The market will evaluate it fairly.

But please stop framing this as "AI replacing you," packaging a fundamentally "wrong past decision" as a future-oriented move. Also, stop inventing stories like "we distilled talent into models" that sound high-tech but are actually untestable. It’s unfair to laid-off employees—they are not lacking ability; their company simply bet on the wrong direction. It’s also unfair to AI—it’s being forced into a role it can’t fulfill, bearing a stigma it doesn’t deserve.

AI is a crucial tool of this era. Its true value lies in empowering professionals, compressing repetitive low-value tasks, and giving humans more time for judgment-intensive work. It should not be a scapegoat for big tech layoffs nor a shield for corporate misjudgments.

Because—AI isn’t that magical, but it also shouldn’t be so vilified.