Tech Needs to Take a Stronger Stand Against Senseless Mercantilism

Better late than never: It’s an ancient proverb, a cliché that has become grizzled and toothless. And yet, the hoary saw can still give us a gummy bite.

I thought of “better late than never,” among other things, while reading an article earlier this week in the Wall Street Journal. In the article, Nvidia Chief Executive Jensen Huang decried the ineffectiveness of U.S. export controls restricting the sale of advanced chips to China. Huang pronounced the export controls a failure, saying they served only to give China an impetus to accelerate development of its own artificial-intelligence silicon.

Well, duh.

How could the export controls have produced any other result?

Of course, Huang is right, but he should have proactively, publicly, and vigorously opposed export controls at their point of ideation rather than waiting until they made Chinese chips great again. (Well, great for the first time, but that formulation wouldn’t have given me license to twist the verbal knife.)

U.S.export controls left China with no option but to blaze its own AI trail, developing its own chips and pursuing a made-in-China AI stack from top to bottom. The obvious question: what did the U.S. government expect China to do?

Doing Nothing: Not an Option

China, seeing that the supply of critical technologies was subject to arbitrary restrictions in the form of export controls, was forced to expedite development of homegrown substitutes. What other realistic alternatives were at hand? To do nothing? To submit and surrender? To abandon its exploration and pursuit of AI and other advanced technologies?

Those were never options, and it was ludicrous for successive U.S. administrations to assume that China would meekly roll over and succumb to the deprivations of export controls.

Instead, China’s government and the country’s domestic tech industry responded logically and determinedly to adverse circumstances. If your supply of a critical resource is about to be cut off, you must find a new source, preferably one that cannot be turned off. Huang should have known what was coming, and the Biden and Trump Administrations should have known what was coming. It was as predictable as night following day.

In practice, Chinese authorities took two primary courses of action when faced with U.S. export controls on GPUs and AI accelerators. First, of direct concern to Nvidia, Chinese officials and companies poured resources into GPU alternatives, into chips that could serviceably replace those previously purchased from Nvidia. This move was necessarily a long-term initiative, and it is just now beginning to produce meaningful fruit at Huawei and other Chinese companies.

In the interim, however, China and its technology ecosystem also sought to develop AI models more economically and efficiently, with programming and process advances that obviated the need for top-of-the-line AI chips from Nvidia. This was a smart move for two reasons: First, to reduce dependence on Nvidia’s high-end GPUs, which were increasingly foreclosed to Chinese companies and organizations; second, given that development of China’s homegrown AI chips would take considerable time to close ground on Nvidia’s market-leading technology, Chinese had to identify other advances that would render AI silicon less decisive to the realization of AI solutions.

When you think about it, the second objective is more important than the first, because of the very real possibility that Nvidia might maintain technical leadership over Chinese rivals well into the foreseeable future. Rather than playing catch up on the established game board, predicated on the brawn and brute force of AI accelerators, China sought to move proceedings to a different game board entirely. The new game board was oriented around enhancing the resource efficiencies of AI models rather than around an AI-accelerators arms race. Hence, the development of DeepSeek, with similar progeny reportedly on tap.

Making Chinese Silicon Great

We must remind ourselves, however, that these parallel initiatives were fast-tracked entirely because China had no recourse but to go its own way. The catalyst was U.S. export controls. In devising and implementing stringent export controls, the U.S. government sought to starve China of access to the world’s most advanced AI chips. China could allow itself to fall hopelessly behind — not an option, for what should be obvious reasons — or it could develop alternatives to and substitutes for a key technological resource that was becoming increasingly inaccessible.

From the Wall Street Journal article:

Nvidia has over the past four years lost market share to Chinese competitors because of the restrictions, Huang said.
“The local companies are very talented and very determined, and the export controls give them the spirit, energy and the government support to accelerate their development,” Huang said Wednesday in Taipei, where he is attending an industry conference.

I’m about to say something that will be controversial by today’s bizarre standards, but would have been unremarkable not that many years ago. During the golden age of U.S. technology hegemony, boisterous flag waving was kept to a minimum. The focus was on technology, not imperial hegemony.

Everybody knew that Silicon Valley was the heart and occasionally alienated soul of the technology industry, but individual companies, rather than the U.S. government, were at the forefront of the action. Whether we think of Intel, Microsoft, HP, Cisco, and other first-generation infrastructure stalwarts or of Web mainstays such as Google, Amazon, and Facebook, we think of them first as successful (ie wealth-generating) corporate brands, not as government proxies. (Maybe some furtive government service was performed behind the scenes, but , if so, it was done with circumspection and discretion.) Company brands and logos were the colors and banners that flew most prominently from the corporate redoubts that rose over the valley.

Yes, these were companies headquartered in the U.S., but they weren’t stridently jingoistic. The successful companies of the earlier tech era wanted their technologies and products to be bought and sold everywhere on this planet, not just in the United States. In taking a multinational or transnational orientation, these companies became commercial successfully and, like Hollywood in its halcyon period, served as exemplary ambassadors for the United States.

As such, they had the good sense not to wave the flag too strenuously. If you were a country in Europe or Asia, you bought a Cisco router or an HPE server not because it was an American product, but because it solved a business problem or addressed a technical need. You, as a customer, might even grow to trust that vendor as a consultative partner rather than merely as a transactional supplier.

The Cost of Alienating Customers

Somewhere along the way — we can identify the shift as beginning in the middle of the 2010s — things changed. At that point, VCs and technology companies in Silicon Valley were enlisted in what devolved into mercantilist trade wars. This might have made some tech CEOs and big-league VCs feel a rush of patriotic fervor, but the result was a cooling of passions among customers resident outside the United States. The changed dynamic introduced an uneasy, oppositional element to the supplier-customer relationship.

Part of the reason why Chinese technology products don’t succeed as much as they might outside China is because some customers, often in Western countries, wonder whether Chinese companies, such as Huawei, adhere strictly to commercial imperatives. Do these companies put customers first, or are they beholden to an overriding geopolitical mandate. If so, there’s a risk of an adversarial relationship, which a customer never wants to have with a vendor.

This is no longer strictly a Chinese problem.

When a technology vendor becomes enmeshed in aggressive nationalism, complicit with a government wielding tariffs like truncheons, customers outside the United States can’t be blamed for wondering whether something fundamental has changed. A potentially adversarial distrust begins to permeate the relationship between supplier and customer. The foreign customer begins to wonder about things — politics, geopolitics, precedents of creeping export controls, rampaging tariffs — that should ideally never enter into supplier-customer relationships. Exceptional brands might be exemplary ambassadors of their countries of origin, but they are typically not aggressive proponents of jingoism.

What I’ve just written doesn’t apply exclusively to US-headquartered companies’ dealings with China and Chinese customers. Given recent events, dealings with too many countries are at risk of becoming infected by a similar toxicity.

Can Europeans trust that the U.S. government won’t choke off supplies of key technologies through export controls, or that U.S. tariffs won’t make European products and services prohibitively costly in the U.S. market? (I wrote this post earlier in the week, but today’s news regarding Trump’s threat of 50% tariffs on the EU proves my point. Thanks for that, Trump.)

In the Wall Street Journal article, Huang says that he’s thankful that the Trump Administration walked away from what is descried as the AI diffusion rule. To wit:

At the same time, he (Huang) praised Trump for reversing one export-control plan that Nvidia had forcefully criticized. In the final days of the Biden administration, officials released a draft of an AI diffusion rule that would have put country caps on the sale of AI chips around the globe, not just to China. The rule was aimed at making it harder for China to access U.S. technology through third countries.
The Trump administration rescinded the rule this month before it took effect.
“The goal of the AI diffusion rule as specified in the past was to limit AI diffusion. President Trump realizes that it’s exactly the wrong goals,” Huang said.
“America is not the only provider of AI technology,” he said. “If the United States wants to stay in the lead and the U.S. would like the rest of the world to build on American technology, then we would have to maximize AI diffusion, maximize the speed.”

Getting Back on Course

I’d take this argument further. The best way to persuade the rest of the world to build on American technology is not use technology as a weapon in a mercantilist us-against-them trade war.

Instead, just as in earlier generations of the technology industry, the formula for success involves companies focusing relentlessly on advancing and innovating their technologies rather than in getting caught up in invidious geopolitical agendas. Companies, rather than countries, have the benefit of positioning themselves as self-interested commercial entities focused on revenue growth and profits rather than on mercantilism and geopolitical domination. (Shooting wars entail different obligations, of course, but that’s not what we’re discussing here.)

For individual companies, the relationship that matters involves their customers. In allowing themselves to be used as geopolitical pawns, technology companies alienate customers not only in China but all around the world. Consequently, foreign companies and organizations begin to think seriously about developing homegrown products and services, which don’t come with the baggage and uncertainty that beset suppliers under the aegis of a capricious government.

No customer wants to think that their technology supplier is arbitrarily beholden to and working directly for any national government, especially during a period when said government has taken an antagonist, bullying stance toward other countries, including longtime allies. As a customer, can you depend on those suppliers to protect and prioritize your interests? If U.S.-based companies want to maintain and grow their market shares beyond U.S. borders, they should strenuously defend their right to do business unencumbered by unreasonable restrictions placed on them by an increasingly unhinged U.S. government.

Companies should refrain from engaging in strident nationalism and stick to serving their customers with innovative products and technologies that address articulated (and sometimes unarticulated) challenges and problems. Perhaps these twisted times have skewed our historical perspective, but that’s how U.S.-headquartered companies managed to thrive in the past. It remains the best way forward.