The world inches closer and closer to a US-China grand bargain of sorts, after the two countries concluded yet another round of negotiation, this time in Madrid. When the previous round of talks in Stockholm wrapped up, I shared my view that the probability of a grand bargain stood at 55%, up from ~10% in December 2024 and ~30% in June. Now, post-Madrid, my odds are upped to 70%. Leaks of a Trump state visit to China are swirling around.
This higher re-rating is all because of TikTok, which was a major discussion item heading into the Madrid talk. The few pieces of details that have emerged about TikTok from both sides are that a deal for TikTok’s US operation to survive and continue is more or less done. Trump signed a 4th extension yesterday to December 16, so there is enough time to finalize and close the deal.
While the TikTok deal has been receiving most of the headlines, the real grand bargain-like impact is the emergence of a TikTok template. It is the contours of a roadmap and path, however murky and fickle, for other Chinese technologies to flow to the US without too much political headache. For people who hate Chinese technologies in all forms to their core simply because they are from China – the ideologues – this is bad news. For people who could care less and just want technologies they need that are quality and affordable – the pragmatists – this development is a breath of fresh air.
The TikTok deal is a vehicle to get to a TikTok template. Here is how I see this template working, when the deal is done.
Core Technology Licensing
The most important element of this template is a new path for Chinese technology to be deployed through licensing. Prior to this template, the only two paths are getting banned or forced sale to a US competitor. This licensing path was both hinted at by the deputy director of the Cyberspace Administration of China after the Madrid talk concluded, and corroborated by the Wall Street Journal’s reporting of the impending deal.
For TikTok, the core technology is its content recommendation engine, deployed in all of ByteDance’s family of apps. A version of these algorithms are what will presumably be licensed as a way to transfer and run that technology in the US. But that engine is not really the point, even though the social media industry has been wow’ed by its addictiveness, and Congress has deemed it digital poison and a vector of foreign influence on the American public, while the President believes it is the reason why he’s so popular with the American Youts.
This licensing path opens the door for more critical, strategic, and advanced technologies to flow from China to the US, e.g. LiDAR, batteries, and rare earths. In other words, technologies that actually matter to national competitiveness. These examples are some of the components that are already at the heart of building next generation electric vehicles and robotics. These are also components that Chinese companies have a resounding technological and supply chain lead globally. The overhang of a TikTok ban has largely shadowed the prospect of these technologies from entering the US market, even though there are willing buyers on the US side and willing providers on the Chinese side.
The US has made some meaningful steps to indigenize capacity and production in some of these components, like the government owning stakes in both MP Materials and Intel. But real products that can be sold and deployed at scale will still take years. Meanwhile, the Chinese alternatives are ready to go today.
If licensing is the mechanism of this template, the next question is what does the perfect licensee look like?
US Ownership with a Golden Share
That looks to be a fresh new US entity, super-majority-owned by US investors with the US government also at the seat of the table with a “golden share”. (Yes, the memetic nature of Chimerica is reaching an all-time-high!)
The Wall Street Journal report revealed that TikTok’s new US entity, which will be the licensee of the algorithms, will be 80% owned by a consortium of new American investors, including Oracle, Silverlake, and a16z. The remaining 20% is assumed to be held by ByteDance and other previous existing investors, like Susquehanna. The board of this new entity will be mostly American with one seat reserved for a representative of the US government.
The final deal may look a bit different from what is leaked now. The participants of the consortium may change, with more investors clamoring (and lobbying) to get into the deal. The ownership percentage may change slightly in the end. But I don’t see these details changing that much materially. A super majority of American ownership, probably 75% or more, is part of the template. Direct US government participation is another key element. The Chinese side holds a minority, passive stake.
This is the formula that will likely be applied should, say, BYD want to make a jump into the US market, or CATL want to step up its expansion to supply more US carmakers with its batteries, or Hesai want to ship more LiDARs to US robotics and physical AI companies. There are two tricks to executing this part of the template well. One, be fine with settling for a minority stake and giving up a chunk of commercial upside for the "privilege" of selling to US consumers and companies. Two, find the perfect mix of American investors and corporations, who are very (very!) politically aligned with the White House, to anchor the new US entity as the licensee and rid you of any political culpability if things go wrong.
Solving For A Political Problem
What this template ultimately shows is that the problem that the TikTok saga symbolizes has always been a political problem, dressed up as a national security problem, dressed up further as a technology problem.
When I first started analyzing and writing about TikTok five years ago, at the first specter of a ban initiated by the first Trump administration, I had mistakenly and naively thought it was a technology problem that could have a technology solution. That’s why I wasted much digital ink looking at ways to use technology to (dis)trust and verify the risks that TikTok presented, and ways ByteDance could use open source to foster trust on a technical and operational level.
When the focus seems to have shifted to a more national security and legal lens, I wasted more digital ink dissecting TikTok’s credibility problems and ways that those problems could be addressed, again, with technology and more air-tight data compliance and legal processes. Since those analyses were published, a law was passed just to ban TikTok and the Supreme Court upheld that law with a 9-0 vote on national security grounds, just to be completely ignored.
As the TikTok saga (hopefully) draws to an end, it is apparent that all that effort was a giant waste of time. This was a purely political problem all along, so it could only be solved by a political solution. Technology doesn't matter. National security doesn’t matter. The law doesn’t matter. And what makes a good political solution? Identify the most powerful figures with the deepest well of political will, then bind them all into a tangled web of conflicting but attractive interests and incentives that is hard to refuse.
After all, “no conflict, no interest.” No company embodies this phrase more than TikTok. Whoever follows the TikTok template will need to do exactly that: maximize interest, maximize conflict.
