For most of the 1970s, 1980s, and 1990s, the New England Patriots were a uniquely mediocre football team. They weren’t always tragically bad. They weren’t the worst team in the league. They were just persistently below average. Sure, every once in a while they would have a year that was surprising and have a winning record, but the expectation wasn’t that they should be a good football team.
The good football teams were teams like the Dolphins with coach Don Shula and the only perfect record in NFL history. Dan Marino spent the better part of almost two decades throwing the football around like a laser beam. The Bills went to four straight Super Bowls, and I know they lost four straight Super Bowls, but it takes a franchise capable of having management that understands how to win — to do something like that.
The Patriots were the football team that you would pencil in for third place in the AFC East. And fans of Boston sports didn’t really care because during this time they could pay attention to baseball, basketball and hockey. If you wanted to watch a winning Boston sports franchise during the 80s, you watched the Celtics and you were happy to do it.
I think the tide started to turn in the mid-1990s. They made a Super Bowl against the Packers in the 1996. Drew Bledsoe at quarterback and Bill Parcells at head coach put together a solid season.
Even though the Patriots lost the Super Bowl, nobody organized their worldview around the Patriots being a threat. I think that this was the beginning of the Patriots becoming something. Replacing coach Bill Parcells with Pete Carroll was not a bad choice. Pete Carroll is a good coach. He just didn’t put them over the top.
Then Robert Kraft’s front office hired Bill Belichick, drafted a skinny sixth-round quarterback out of Michigan named Tom Brady, and the next twenty years happened.
You know the next 20-year run if you have paid attention to the NFL at all. What people often forget about the beginning of that 20-year run is that, at the very beginning of the 20-year run, a lot of people were insisting that the Patriots were a fluke because they got lucky with the tuck rule and then the AFC was a soft division and then Spygate happened. They still weren’t giving the Patriots credit for being a franchise that is run by competent management that makes good decisions. It wasn’t a fluke. The Patriots actually became a well-run franchise.
The proof is what happens when Brady retires and then Bill Belichick goes to become the coach of the North Carolina Tar Heels… And somehow ends up with a girlfriend in her mid-twenties who’s running his life? (I still have a lot of questions about what’s going on there, but I’m going to leave it alone for now.)
There were a few awkward years. A franchise that didn’t have their act together would have gone into a decade of darkness. (Or several. Ask me! I’m a fan of the Dallas Cowboys.)
Instead, the Patriots hired a new competent coach in Mike Vrabel. They developed quarterback Drake Maye, incidentally, from the University of North Carolina. And at only 23 years old, Drake Maye was the quarterback for a team that came close to winning a Super Bowl. I’m not going to lie to you and tell you that the Patriots looked good in the Super Bowl because they didn’t. Drake Maye got sacked six times, and the Seahawks clobbered them.
But getting to the Super Bowl with a young quarterback shows me that institutionally, you are capable of producing greatness consistently. This is what good management looks like. You go on a good run. You win some Super Bowls. You maybe go through an awkward period sometimes, but that awkward period doesn’t last very long, and then you put it all back together and you contend for Super Bowls again. Good management should be boring, and the Kraft family is boring. They are literally known for making condiments, and also winning Super Bowls… which is one of the most American sentences that I think that anyone could ever write.
This conversation is the entire China conversation compressed into the AFC East.
For a very long time, China occupied the exact same cultural position the pre-Belichick Patriots owned. China was a punchline. They made the knock-off products. They made the slightly weird, more wrong version of the thing you already owned. Their reputation, by the way, was well earned, because I’m not gonna sit here and pretend that “Chinese knock-off” jokes came from nowhere, because for years they came from a very real place: Chinese factories that made substandard shit. And then, sometime in the late 1990s, China started getting their act together.
They built a functioning research economy, and they pointed that research economy at bleeding-edge science and technology fields, and they said, “We’d like to run the next century.”
The scorekeeper on this that I like to look at is the Australian Strategic Policy Institute. The reason I like picking the Australians is because they’re doing a bit of a dance between being a neutral party but not technically thinking through the exact mechanics of neutrality. Australia does an enormous amount of trade volume with China and also has a security relationship with the United States. Australia has every incentive to know precisely which country is good at what.
And for the last three decades, the Australian Strategic Policy Institute has maintained a critical technology tracker, and it measures the top-cited slices of global research across 64 key critical technology areas.
From 2003 to 2007, the United States led in 60 of the 64 areas. China led in three.
Now let’s fast-forward the tape.
In 2024, China led in 57. The United States only leads in seven.
We have almost completely flipped positions.
Heck, the Chinese Academy of Sciences by itself is the top performing institution and 31 of the 64.
If you need to feel good about anything buried in those numbers, here’s the good news. When it comes to natural language processing and large language models, the United States is technically still winning.
Don’t get me wrong, it’s still a 57-7 rout. If you were watching this football game, you would probably take no solace in the pick-six that your safety managed to pull off. That would just be the moment you put down the beer and said, “Well, at least we got something.”
At the end of that game, your takeaway that you brag about should not be “At least our punter is leading the league in punt yards.” If that’s the way you’re viewing things, you’re going to have a pretty bad season.
That brings us back to the models themselves. Over the last two years, open weight releases from Chinese labs, research institutions, and companies have become legitimately frontier-grade competitive models. Deepseek, Qwen, Z AI, KIMI, etc — these are not Chinese knock-offs. They are trading equal blows with U.S. made frontier models. This is Mahomes against Allen. You should not be surprised, in a game with these two elite quarterbacks, that it’s going to be close every single game.
And yet the reaction from a large number of American technologists and computer scientists has been this genuine, almost wounded surprise, as though a Chinese company shipping a competitive frontier model violates everything they understand about how AI is supposed to work.
Plot twist: the Chinese are just getting really good at it.
The American take on Chinese models is just sort of a lazy American assumption. It’s probably just an artifact of American Silicon Valley culture. The prevailing theme is that we win by spending more money on everything. We win by having the most venture capital and liquidity. We win because we can afford to lose for a long time to capture market share. This is not how reality works when you are trying to compete in a global economy.
DeepSeek humiliated that assumption when it trained competitive frontier models for a fraction of what the American labs had assumed was the cost of admission. There was no magic trick involved. It came out of a genuinely different engineering culture, one where getting the most out of the least is the definition of good work. They operate in an environment where you are expected to be surgical about where every dollar and every watt goes instead of bludgeoning the problem into submission. Efficiency is a competitive weapon.
And it doesn’t stop at software.
I will never forget watching MKBHD review the Xiaomi SU7. You have this seasoned tech and car reviewer, and you can tell that even though he is an influencer who is supposed to remain as neutral as possible, there are several moments during the filming where he genuinely has a smile on his face. His surprise was genuine. This is a person who realized in real time, as his viewers are watching the YouTube video, “Wait a second, the Chinese didn’t just catch up, they have surpassed us, and they’re doing it for half the price.”
That experience is not just a “car thing.”
China has poured money into biotech and its pharmaceutical pipeline, into materials science, into the unglamorous industrial base that every shiny consumer product sits on top of. The car is just the most tangible example that fits into American car culture, so it’s more visible, and probably a comparison that more Americans can appreciate.
For the record, none of these statements obligate me to like the Chinese government. The speech restrictions are real. The surveillance state is real. The treatment of the Uyghurs is real and… it is fucked up in ways I can’t explain in a blog post. All of those statements can be true and accurate, and it doesn’t lower China’s science, industrial, or engineering capability at all. I think Americans just want to pretend that these two ideas have to be mutually exclusive, and they are not.
China spent thirty years as the setup to a joke about knockoffs and cheap production. We got super comfortable with that punch line. And while we were making jokes, China built a research economy that leads in 57 of the 64 areas that actually matter.
At some juncture, you have to wrap your mind around China not being a fluke. You don’t win a game 57 to 7 on luck. China built a real contender in all sectors of industrial technology, engineering, sciences, and manufacturing. I know a lot of you reading this (because you’re my peers) are probably thinking, “Well, at least we’re still winning in defense technology.”
Based upon everything you’ve seen, how long do you think we can keep that competitive edge?
We are two days away from our nation’s 250th birthday. Enjoy the celebration.
Then get back to work. Because our biggest competitor is whooping our ass right now.