Volume 2, Number 4 in Global TECHtonics: U.S./China Fault-line series

 

The weekend’s big development in the technology arena is Beijing’s eleventh-hour move to alter the timing and trajectory of the sale of TikTok’s U.S. operation.

We touched on the Trump Administration’s August moves against TikTok’s parent Bytedance in the U.S./China De-Coupling: 4 Levels of Risk post two weeks ago.  On August 6th, President Trump signed two executive orders which started a 45-day time-clock involving two Chinese companies with hugely popular social media apps – ByteDance (owner of TikTok) and Tencent (owner of WeChat).  According to those orders, U.S. citizens and businesses would be barred, once the 45-day period expired, from any transaction involving the company and/or its products.  On August 14th, the Trump Administration modified the order as far as it affected TikTok by putting a new order in place, giving TikTok 90-days within which to complete the divestiture of its U.S. operation to an approved U.S. corporate buyer.

The widely-presumed reason for this change being made so shortly after the announcement of the original order is that U.S. potential buyers interested in acquiring the U.S. operations of TikTok had pitched their interest to the White House.  It is not surprising that U.S. potential acquirers would be focused on TikTok and not WeChat.  The number of TikTok users in the U.S. is estimated at 80 million in comparison with 19 million for WeChat.  Its growth rate in global markets is far faster and, critically, its algorithms have nearly ubiquitous applicability whereas WeChat algorithms are more geared to Chinese user behavior and are so less replicable in other world markets.

Two groups of interested buyers have emerged publicly since the August 14th announcement:

  • Microsoft/Walmart: As Instagram and other social networks edge into offering shopping features, Microsoft and Walmart are looking to establish themselves at the strategic center of this opportunity with one bold acquisition  move.   Put simply, Walmart would provide the e-commerce component for TikTok while Microsoft would manage the crucial cloud-computing infrastructure.  The deal offers competitive advantages to both firms – Walmart would become better positioned to compete with Amazon and Microsoft would gain experience with an innovative and cutting-edge set of algorithms and data-sets.
  • Oracle: According to analysis by the New York Times business reporter Mike Isaac, “Oracle could use TikTok’s data about social interactions to benefit its cloud, data and advertising businesses.” Also, like Microsoft and Walmart, Oracle is interested in the opportunity the deal would afford “to offer customers a hyper-personalized experience in both content and commerce.”

Going into the weekend, the expectation was high that Bytedance’s preferred acquisition partner would become known and that negotiations would shift to a new phase of negotiation with only that chosen partner.

So, what was the development over the weekend which changed the trajectory and pace of this deal?  The Chinese government announced late in the day on Friday that any sale of Bytedance’s assets would be subject to a brand-new set of restrictions affecting artificial intelligence exports.  As reported in still-developing coverage in the Wall Street Journal, “the new Chinese restrictions highlight the extent to which TikTok, a breakout social-media hit—especially with younger U.S. users—has been thrust into a geopolitical contest between the U.S. and China over the future of global technology.”

I’ll limit my commentary on this development to three main points – a historical observation, a key point having to do with the present-day competition in advanced technologies between the U.S. and China, and my personal handicapping of where this deal is likely to go in the weeks ahead.

 

Historical Antecedent: The U.S.-Japan Trade War

While observers sometimes invoke the U.S.-Japan Trade War as a template for understanding our current tensions with China, the contrasts between the two are probably more instructive than the similarities.  A future post will return to the broad comparison.  For our purposes here, I will single out one important point of contrast.  The U.S.-Japan Trade war became incandescently hot as a political issue in the lead-up to the 1992 U.S. Presidential election.  But while that was happening, commercial developments on the ground were already in motion to begin lowering the heat.  The industry sector in which the grass-roots transformation took root and started having great effect was the automotive sector.  The seed for that bottom-up transformation was the fact that, post-war, Japan had developed intellectual property in their domestic market  that made them more competitive than the U.S. industry in a number of vital areas of automotive manufacturing (e.g., inventory management, quality control, customer-based innovation, etc). Led by Toyota, the Japanese and U.S. industries started reaching an accommodation even before politicians in the U.S. turned up the volume on their anti-Japan megaphones.  Japan would license out its intellectual property and bring its production closer to its customers in the U.S. by building factories and supplier networks in the U.S.  In return, American companies would gain access to know-how in areas where its competitiveness was lagging and also gained greater access to the restricted Japanese market.  At a political level, investments in new state-of-the-art production facilities in the non-unionized south brought jobs into key congressional districts.  Of equal importance, auto workers, their families and their communities started having the experience of working alongside Japanese managers on U.S. soil.  In the process, real-world people-to-people experiences built on collaboration replaced the one-dimensional caricatures being amplified by politicians and the media.

The Chinese have studied this experience whereby Japan lessened the political tension of the U.S.-Japan Trade War while, simultaneously expanding access to the lucrative U.S. market and affluent U.S. consumers.  For various reasons, they have not been as successful in applying the model.  We’ll examine the broader set of reasons in a future post but, for present purposes, one salient reason is that China, generally speaking, has not developed the portfolio of intellectual property focused in high-value industries (like, for Japan, automotive and consumer electronics) and highly sought after by U.S. companies.  Except, that is, until now as China emerges with competitiveness in advanced technology fields such as artificial intelligence, robotics, and autonomous vehicles.

 

Looking at Both Sides Now:

The U.S. innovation ecosystem represented by Silicon Valley is, and is likely to remain for the foreseeable future, peerless in many important respects – depth of talent and experience, access to capital, connectivity to leading universities, basic research capability and innovation mindedness.  In three respects, however, emerging tech competitors in China enjoy advantages which U.S. firms can’t match.  First, China has been for years the biggest and fastest growing market in the world and U.S. companies can’t afford to cede that base of users entirely to their Chinese competition to monopolize.  However, the ability of U.S. firms to access those consumers is highly constrained by a whole raft of protections – many non-WTO compliant and others not yet covered by WTO ground-rules — by which the Chinese government limits foreign access to its home market and by which it supports its home-grown champion companies.  Second, China may enjoy a tactical advantage through its laser-focus on market applications (as opposed to research and academically-based innovation). Third, AI firms in China definitely enjoy a leg-up in algorithm development because they have direct access to the world’s largest user-base for smart phones and are less constrained by privacy protections for those users.  These latter two advantages for Chinese tech firms are persuasively presented by the former President of Google China, Kaifu Lee (a Taiwanese national whose computer science PhD thesis at Carnegie Mellon gave birth to the world’s first speaker-independent, continuous speech recognition system) in his book AI Superpowers: China, Silicon Valley and the New World Order.  In Lee’s view, “the United States may have been a first mover in AI but that advantage will not last forever. The AI era will reward the quantity of solid AI engineers over the quality of elite researchers. Strength will come from an army of well-trained engineers and entrepreneurs, and China is training just such an army.”

So, stepping back, there is now for the first time since normalization of U.S.-China relations a strategically-important (emerging) industry where Chinese firms hold important competititve advantages over the U.S.  Unlike democratic Japan, this high-stakes competition is associated with a Communist regime with all that that entails for public attitudes in the U.S.  And there is little in the of way local ties-that-bind being built quickly and effectively on a people-to-people basis.  Nothing that can match the stabilizing experience with Japan investment into the U.S. in the 1990s. Together, these three factors go a long way to illuminating the huge pressures that have been building up under the U.S./China technology faultline on both sides of the U.S. political aisle.

 

Where’s The TikTok Deal Likely to Go?

Despite the fact that practically nothing is known yet about the details of the PRC government restrictions announced on Friday, two things can be safely said.   First, the fact that the PRC government is invoking national security as a basis for governing the commercial activities of its leading artificial intelligence firms is hardly surprising.  The competition between the U.S. and China is, for reasons just examined, acute.  The U.S. and other countries routinely monitor and manage international commercial activity for their technologically-advanced products and services, especially those that are ‘dual-use’ in both commercial and military applications.  The second point is that the timing of the announcement tends to be viewed in the U.S. as so transparently tied to the on-going negotiation involving TikTok that it will be viewed more as a political beanball, than a fair pitch.  This despite the fact, as pointed out by an astute comment (see below), that these new regulations had been proposed prior to Trump’s August 6th announcement and were in a public comment process.

The Chinese government action raises the prospect that key algorithms and other vital data – everything that makes TikTok tick — may be stripped out of the sale by its Chinese parent corporation as a new requirement of Chinese law.  That result would fundamentally change the value proposition for both the Microsoft/Walmart and Oracle bidding teams.  It’s like the difference in value between a top-of-the-line computer and that same computer with all its electronics removed.  At the very least, the PRC government action will force all parties to slow the pace of their negotiations and delay the deal being sealed until there’s greater clarity about what will ultimately be allowed.

With Friday’s move, it’s likely that the Chinese government will be satisfied with slowing the deal and changing the trajectory of its fall-out for global technology competition.  Scuppering the deal entirely would risk dramatically escalating the issue with Trump and his Administration.  That would go against China’s temporary strategy of muted response to the Trump Administration’s recent, pre-election flurry of jabs.  The idea in Zhongnanhai in the run-up to November 3rd is to give its wolf-warriors and nationalistic netizens enough to appease their appetites but not enough to risk fanning Washington-Beijing flames out of control.

So, with the clock ticking down to 64 days before the U.S. election and with 78 days before the Trump Executive Order 90-day deadline expires on November 12th, the endgame of this global chess match is now ruled by the time-clock.

TikTok, TikTok, TikTok …