That weekend, I originally only wanted to see what Max Verstappen's first serious attempt at the Nürburgring 24 Hours would look like.

Then I kept watching, and somehow I got pulled all the way in.

I really got pulled in.

I used to mainly watch F1. My interest in GT3 and endurance racing was never that strong. Endurance racing, for me, had mostly been something I saw in short videos. I knew Le Mans, I knew Spa 24h, I knew the Nordschleife was long and absurd, and I knew everyone called it the Green Hell.

But knowing that is one thing.

Actually sitting down for a weekend and watching the live broadcast is another.

Especially this time, because Verstappen was there.

What else was there to say?

The hype was going to be off the charts.

A driver who already feels almost unreasonably strong in F1 suddenly shows up on a 25-kilometer circuit with more than 170 corners, weather you can never fully predict, nights that feel like disappearing into the mountains, and more than a hundred cars with huge speed differences. He is in a GT3 car, sharing it with teammates, trying to run 24 hours.

Just think about that combination. How could it not make people curious?

So I went online and found a stream.

Very quickly, I realized that this race asks to be watched in a completely different way from F1.

F1 is precise, modern, and almost like a high-speed sport being run inside a high-pressure laboratory. Every lap, every pit stop, every safety car window is broken down to tiny margins.

But the Nürburgring 24 Hours is not like that.

It feels more like a huge, noisy, damp automotive festival, with the smell of gasoline and grilled sausages in the air.

According to the official results, the 2026 race started with 159 cars, and 111 were classified at the finish. There were 352,000 spectators on site. The lead GT3 cars were fighting for tenths, while behind them were touring cars, cup cars, a Dacia Logan, a Subaru WRX, an Opel Corsa, and a pile of cars that make you wonder why they are allowed to share the same race as the overall-winning AMG.

But that is exactly the Nürburgring.

It does not only put the fastest people together.

It puts everyone who truly wants to prove something into the same forest.

Honestly, I went in for Max. By the end, my view of GT3 cars had changed completely.

These cars are so pure.

They do not have the alien-technology distance of F1 aerodynamics, or that feeling of a car body being as light as a sheet of paper. GT3 cars feel more like cars whose silhouettes you can still recognize from the road, but transformed into machines that breathe, roar, and bounce off the curbs.

You can hear violence in the AMG's sound, weight in the BMW's posture, and a sharp little streak of menace in a Lamborghini's taillights at night.

I was getting a bit carried away.

How carried away?

Before the race was even fully over, I impulsively bought Assetto Corsa and Assetto Corsa Competizione.

The awkward part is that I still do not have a steering wheel.

So there I was, foolishly holding the Xbox controller I usually use for Forza Horizon, running laps in the game for hours.

But even with only a controller, it was enough to get me excited. Every time I pressed the throttle and heard that pure engine sound, every upshift that sent the revs up, every time I thought I could brake a little later and then flew straight off, all of it helped me understand a little more why people are willing to pour a whole weekend, and sometimes many years, into this circuit.

Of course, what I did with a controller probably cannot be called driving.

It was education.

Nürburgring education.

Back to the race.

The two crews I cared about most were obvious: No. 3 Mercedes-AMG Team Verstappen Racing and No. 81 BMW M3 Touring 24H.

The first because of Max.

The second because of that car.

The BMW M3 Touring really is my dream car.

And this time it was not a normal M3 Touring. It was the SP-X race car built by BMW M Motorsport, officially called the BMW M3 Touring 24H. Its story is absurd too. It first grew out of a 2025 April Fools' joke, then BMW actually built it and stuffed in a lot of M4 GT3 EVO race-car components.

You think it is a joke project.

Then it finishes fifth overall.

Seriously.

How is anyone supposed to resist that?

First, car No. 3.

Car No. 3 was entered by Winward Racing under the name Mercedes-AMG Team Verstappen Racing. The drivers were Max Verstappen, Lucas Auer, Jules Gounon, and Daniel Juncadella.

It qualified fourth.

That alone was already outrageous.

This was Max's Nürburgring 24 Hours debut. Even if he is an F1 world champion, even if he had already run plenty of Nordschleife laps in simulators, the real Green Hell is not a place where you simply show up and go fast.

After the start, car No. 3 quickly entered the main storyline.

But at 16:09 local time, while Verstappen was running through the famous Pflanzgarten jump area, the wake from the car ahead suddenly reduced the downforce. As the car landed, it snapped sideways, nearly hit the barrier, briefly left the racing line and slid onto the grass. With very sharp car control, he caught it, stabilized the car, and rejoined.

It was not a disaster, but it was very close. It also served as a reminder:

After that, he began pushing forward.

At 16:28, he passed the No. 7 Lamborghini at the Sabine Schmitz corner. At 16:50, he squeezed past the No. 47 Mercedes at Kesselchen, even touching the grass. At 17:52, he passed the No. 67 Ford Mustang and then the No. 34 Aston Martin, officially taking the lead.

That stretch had a real cinematic feeling.

Someone you know from F1 was dropped into a completely different battlefield. There were slower cars, damp surfaces, traffic, and Code 60 zones that could appear at any moment. He did not have the clean attacking space you see in F1, but he still worked the car forward bit by bit.

Then the evening rain arrived, and the race changed character.

More precisely, it became what the Nürburgring is supposed to feel like.

Cars No. 3 and No. 80, both Mercedes-AMGs, made the right calls on tires and rhythm, and began taking control of the race. At five hours, car No. 3 led and car No. 80 was second. At seven hours, car No. 3 was still first. By night, the race had basically become an internal fight between the two Mercedes cars.

That part was interesting.

It was not the neat teammate parade people sometimes imagine. Media recaps mentioned that cars No. 3 and No. 80 had very tight battles around Döttinger Höhe and Tiergarten, even with contact and trips across the grass. Only later did the team start asking both cars to control the risk.

I actually liked that detail.

Because that is a very real part of endurance racing.

In theory, you are on the same side. Your shared goal is to help Mercedes win the Nürburgring again after ten years.

But you are sitting in the car. The car ahead is your teammate. Behind you are 352,000 spectators, broadcast cameras, your entire racing career, and a 24-hour race victory.

Can you really not want to win at all?

People are not machines.

That dangerous tension made the race feel more real.

Deep into the night, car No. 3 remained strong. There was one official nighttime update that I especially liked. Around 02:50, car No. 3 caught the fifth-placed No. 81 BMW M3 Touring and tried to lap it, but could not complete the pass before pitting.

That scene was almost too perfect.

My favorite driver, in the leading AMG, chasing my dream BMW M3 Touring.

How was that not tailor-made for me?

By the nineteenth hour, car No. 3 was still leading overall on 122 laps. On lap 129, it set its own fastest race lap, 8:12.818.

If the story had ended there, it would have been perfect.

An F1 world champion runs the Nürburgring 24 Hours for the first time, qualifies fourth, leads for a long stretch, survives the rain and night, fights hard with a teammate, and brings the car home to win.

If that is not motorsport wish fulfillment, what is?

Unfortunately, this is exactly where endurance racing becomes cruel.

It does not care how beautifully the first 20 hours were written.

In the twenty-first hour, everything suddenly changed.

At 11:08, car No. 3 pitted, Verstappen having completed his stint and handed over to teammate Daniel Juncadella.

At 11:32, Verstappen, having just finished his drive, gave a trackside interview. He was in a relaxed and steady mood, talking about the need for constant focus and avoiding incidents, and said he truly loved the teammate-rotation and long-distance competition of GT3 endurance racing.

Then at 11:37, right after the interview, car No. 3 was pushed into the garage.

At 11:38, the team began working around the right-rear area.

Verstappen went from relaxed to somber, silent, shaking his head in an instant. He stood in the pit lane, watching car No. 3 being slowly pushed back to the garage, brow furrowed, saying nothing for a long while, visibly holding back anger and disappointment.

At 11:48, the official update confirmed a non-contact technical issue at the right rear. The final explanation pointed to a driveshaft-related failure.

When I saw that, I genuinely froze.

That is not an exaggeration.

Because after such a long time, you start to assume it can win. Even if you know endurance racing is cruel, human emotions work like that. You watch it for more than ten hours, watch it lead, watch it survive the rain, watch it fight car No. 80, watch it thread through slower traffic in the night.

You cannot help believing that this story should have a complete ending.

Then a driveshaft tells you: sorry, there is no "should."

The broadcast cameras caught Juncadella sitting in the cockpit, lifting his helmet visor and silently wiping tears from the corner of his eye, disappointment and regret pouring out of him.

At 11:51, Juncadella got out. Car No. 3 spent a long time in the pit lane. In the 23-hour update, he said something painful, roughly that the earlier part had felt almost like a dream, and for them the race was three hours too short and three hours too long.

That line felt brutally accurate.

If the race had been three hours shorter, they might have been champions.

If they had not lost those three hours in repairs, there might still have been a story.

But the Nürburgring 24 Hours is 24 hours.

Not one minute less.

In the end, car No. 3 was repaired, returned to the track, and was classified. The provisional official result was 37th overall with 135 laps.

That was obviously not the result they wanted.

But it made me like this race even more.

Because suddenly you understand that endurance racing is not a game of whoever is fastest wins. It is a game of who can be interrupted by fate the least over 24 hours.

That may sound a bit abstract, but after watching it, you understand.

That brings us to car No. 81.

I really like the M3 Touring.

Not in the sense that I analyzed the specs and decided it was excellent.

I just like it.

A wagon body, M power, enough practicality, enough speed, and a very German kind of practical violence. It does not put desire on its face like a supercar. It feels more like someone in a hiking jacket with a helmet in the backpack, ready to head out on track at any moment.

And this No. 81 M3 Touring 24H was even more wicked.

It ran in SP-X, not as a standard homologated GT3 car. In a sense, it could have remained a good-looking marketing project, a car for fans to photograph and social media to celebrate.

But BMW did not stop there.

It was genuinely fast.

It qualified 22nd overall with an 8:18.868. That may not sound like the very front, but almost everything ahead was top-level SP9 GT3 machinery. For an experimental race car wearing a four-door wagon shell, that position was already absurd.

The race was even more absurd.

After one hour, No. 81 was still 33rd overall. That felt normal, maybe even exactly what people expected before the race: keep it steady, avoid trouble, finish.

Three hours later, it was fifth overall.

I was confused when I saw that.

How did you get up there?

And then it was not just a brief flash. Fifth hour, sixth. Seventh hour, fifth. Eleventh hour, fifth. Thirteenth hour, fifth. By the fifteenth hour, it was fourth overall. In the twenty-first hour, after No. 3 fell back, it even appeared in third overall for a while.

An M3 Touring, in the final phase of the Nürburgring 24 Hours, had touched the shadow of the overall podium.

Can you believe that?

And its pace was not merely for show. In the official results, No. 81's fastest lap was lap 123, an 8:13.580. Put that next to a field of GT3 cars, and it does not look embarrassing at all.

One of the images that stuck with me most came with 18 hours and 33 minutes remaining. It used slower traffic as a screen and pulled off a beautiful move on both the No. 3 Verstappen car and the No. 80 car that would later win, briefly reaching second overall.

That kind of moment is hard to explain through a result sheet. The sheet only tells you the position, lap count, and fastest lap.

But watching it live, the feeling was: wait, this wagon is actually fighting.

Not sightseeing.

Fighting. Fighting very steadily.

It found space in traffic, used slower cars for momentum, survived in the rain, and drove into the night with yellow lights pushing forward. When you look at its body proportions, it has a strange pressure next to the GT3 cars.

Like someone who should not be there, yet somehow stands near the front.

That is the most charming thing about it.

It is not the most standard race-car answer.

But it gave a beautiful non-standard answer.

It did not make the podium in the end, and that was a little regrettable. Media recaps also mentioned the timing of Code 60 zones, oil on the track, and a few small off-track moments that hurt its podium fight. But fifth overall and first in SP-X is already beyond success.

That is overachievement.

You think they built it for fun, and then it can actually arm wrestle the top SP9 teams. It could even keep reappearing in the same storyline as No. 3 and No. 80.

Anyway, I enjoyed it.

Congratulations, BMW.

Really, congratulations, BMW.

After the race, BMW M Motorsport chief Andreas Roos said something that made people both happy and disappointed.

He said that because of GT3 homologation restrictions, this car would not race again. But he hinted that they had some plans for it, and that we would see it somewhere. If it becomes a Nürburgring taxi, I must ride it at least once in my lifetime.

Beyond those two cars, there were many other things in this race that felt deeply Nürburgring.

For example, the No. 80 AMG.

The No. 80 Mercedes-AMG Team RAVENOL won overall. The drivers were Maro Engel, Luca Stolz, Fabian Schiller, and Maxime Martin. The official result was 156 laps in 24:05:27.664, with an average speed of 164.334 km/h across the entire race.

That car had a satisfying story too.

It was damaged in Friday qualifying and started from a relatively low position, then climbed back through the race. After No. 3 hit trouble, it took over the lead and finally ended Mercedes-AMG's Nürburgring 24 Hours victory drought dating back to 2016.

Throughout the race, the team controlled the rhythm perfectly, pit stops were efficient and clean, and the car stayed in stable condition without any major mistakes affecting the race.

Maxime Martin's part was also moving. He had been close to a Nürburgring win many times before, and this time he finally got it. Endurance racing is cruel, but occasionally it rewards those who have waited for a long time.

Then there was the No. 84 Abt Lamborghini.

It took pole position, then suffered a puncture on the first lap after body contact in traffic, fell all the way back, and spent the next 23 hours climbing back to second. It also set the fastest lap of the race, 8:08.758.

That car was like a short drama too.

Episode one: fall into the hole. The remaining 23 hours: climb out, and earn a highlight moment.

The No. 34 Walkenhorst Aston Martin met a dramatic Code 60 speed limit near the end on the long straight and could not catch the Lamborghini ahead, which was carrying an 85-second penalty. It finished third, which was painful, but still gave Aston Martin an important Nürburgring 24 Hours podium.

The No. 99 ROWE BMW M4 GT3 EVO recovered from a chaotic start and early slides to fight back to fourth overall, narrowly beating the sister No. 81 M3 Touring to the line.

That was a very BMW kind of result: a proper GT3 factory car nearly made the podium, and a wicked M3 Touring that came from an April Fools' joke fought all the way to fifth overall. BMW fans must have had the time of their lives.

And then there was Grello.

The No. 911 Manthey Porsche, that instantly recognizable yellow-green classic Grello, qualified eighth. This year marked Manthey's 30th Nürburgring 24 Hours, and from 2006 to 2009 they won the race four years in a row. That was serious dominance.

This time, they came for their eighth win, but oil on the track sent them off and out of the race early. For people who watch the Nürburgring often, that must have been a major regret.

Grello is no longer just a car.

It is a symbol inside this forest. You do not have to support Porsche, but when you see that yellow-green car, you know an old acquaintance has arrived.

This time, it could not stay.

Jules Gounon's description of the Nürburgring also stuck with me. He said this place chooses you. It is not that you choose the track, but that the track chooses you. That line feels incredibly visual here.

Then there are the locals and the grassroots feeling.

There was one car that, even without finishing, became the most romantic footnote of this year's race.

It is the No. 632 BLACK FALCON Team FANATEC.

Jimmy Broadbent, Steve Brown, Misha Charoudin, Manuel Metzger:

They are a sim racing superstar, a YouTube racing legend, the king of Nürburgring taxis, and a Nürburgring veteran who anchors the crew.

This is the most perfect crew of the internet era: from steering wheel screens to the real Nordschleife, from video comments to in-car heartbeats,

driving virtual passion into a 24-hour endurance race.

They led the AT2 class for a stretch, one step away from conquering reality from the simulator.

Unfortunately, Steve Brown crashed into the wall and retired. After 70 laps, they retired and were not classified.

But the result does not matter at all.

The meaning of this No. 632 is to tell everyone:

But its presence itself was interesting.

Because you realize the Nürburgring 24 Hours is not only a giant machine of factory teams and professional drivers. It also has room for people who grew out of simulators and grassroots car culture.

There was also Dirk Adorf: this year he was racing the No. 992 Porsche in the 24 Hours while simultaneously providing in-car commentary for the German national broadcast.

Running the Green Hell's 25-kilometer circuit, managing dense traffic, enduring the fatigue of day and night, while also passing his real-time feelings, track details, and race situation back to viewers in front of the television.

Other people drive for 24 hours fully focused. He drives and commentates at the same time. This is literally putting a living map straight into the broadcast booth.

In another race, that might feel strange.

At the Nürburgring, it somehow feels reasonable.

And then there was the No. 300 Dacia Logan, the most classic grassroots hero of the Nürburgring 24 Hours, also called the race's mascot by fans.

It is a race car built out of a cheap family hatchback and held together by grassroots passion — the most contrasting and moving presence in the entire field.

During the broadcast, the commentator said that if you have no frame of reference for this car, imagine a Chery QQ running the Nürburgring at average speeds over 100 km/h through the Green Hell for 24 hours!

The legend of the No. 300 Dacia Logan runs deeper: born a cheap family car, twice crashed and written off as a total loss, twice resurrected through global fan crowdfunding and grassroots effort. Even with failures during the race, it stubbornly finished, breaking the limits of budget and origin through pure passion, embodying the most moving grassroots romance of the Nürburgring.

On the circuit, every time this car passed the grandstands, all the spectators erupted in warm applause and cheers — genuinely moving.

In the official results, it was 107th overall with 92 laps. You cannot call it fast, but when you see it listed in the same result sheet as the overall-winning AMG, the event suddenly feels very lovable.

Oh, and there was one especially funny thing. Former F1 driver Timo Glock, driving the No. 69 Doerr McLaren, accumulated four penalty points for speeding in a Code 60 zone, which led to his Nürburgring circuit permit being temporarily suspended.

A former F1 driver losing a circuit license because of speeding — this must be quite rare in Nürburgring 24 Hours history.

These cars and stories might not trend if you pulled them out individually.

But they form the real background color of the Nürburgring 24 Hours.

This is not a race where only the front-running crews are worth watching. It is like a huge automotive ecosystem exhibition: top factory teams, professional drivers, old teams, small cars, experimental cars, internet-famous drivers, local workshop resources, camping spectators, all inside it together.

There are also very human details.

In the 19-hour official update, there was a story from Marco Seefried. He said he drove into a huge cloud of smoke and first thought something had happened. Then he smelled it and realized spectators by the track were having a generous breakfast.

That made me laugh.

That is the Nürburgring 24 Hours.

Drivers are fighting fate from inside the car, while spectators are grilling breakfast beside the forest.

On one side, there are eight-minute laps at the limit. On the other, camping chairs, beer, barbecue smoke, and cold air before dawn.

This race is bursting with life.

There was also the story of the No. 36 BMW E89 Z4. In the 23-hour official update, this car had a technical problem at night, and the team had no spare part. In the end, a local acquaintance removed a part from his own garage-kept Z4 GT3 and gave it to them.

Listen to that.

That is what home turf really looks like.

Not every problem is solved by a major manufacturer. Sometimes it depends on who you know around the circuit, whose garage still contains an old Z4, and who is willing to remove a part in the middle of the night so you can keep racing.

I have always felt that this is what makes racing culture truly fascinating.

It is not only the podium.

The podium is important, of course, but it is only the final patch of light. Beneath that light is a whole weekend held up by teams, mechanics, spectators, photographers, volunteers, circuit staff, and local people.

After watching this race, I really could not calm down for a long time.

Part of it was pure excitement.

The fuel-engine sound of GT3 cars, the lights cutting through the forest at night, the slipstream on Döttinger Höhe, tire choices in the rain, chaos caused by slower cars, and fate turning under Code 60.

All of it was electrifying.

The other part was awe.

Because when you look at car No. 3, you realize that even when you are that fast and that strong, one mechanical problem in the final three-plus hours can still take away every chance of victory.

When you look at car No. 81, you realize that a project that looks like a joke can, if built and run seriously, shatter everyone's expectations over 24 hours.

When you look at the small cars, veterans, locals, and grassroots teams, you realize racing culture is not only the few people standing in the middle with trophies.

It is a group of people collectively believing that cars are not only transportation.

Cars can be sound, speed, technology, local memory, weekend life, and a clumsy, expensive, exhausting, yet deeply moving kind of love.

I used to watch F1 and think racing was cool.

But this Nürburgring 24 Hours made my love for car culture take root a little deeper.

Not because I understand it better.

Because I want to understand it more.

If I get the chance, I really want to go to Germany, to the Nürburgring. Not just to take a photo with a sign, but to stand beside that forest and hear a car approach from far away. The sound arrives first, then the lights, then the whole car blasts past in front of you.

I want to know what the Green Hell actually smells like.

I want to see how narrow, steep, and frightening those corners from the broadcast really are.

And I want to visit the birthplace of the automobile, to see why a mechanical culture with more than a century of inheritance can still make people this obsessed today.

One last note.

This year's Nürburgring 24 Hours was honestly much more enjoyable than F1 under this year's new rules.

That may sound harsh, but that is how it felt to me. F1 is still one of the highest forms of motorsport in the world, of course. But this year, a lot of the time, it has felt too clean, too precise, too much like a product packaged by rules and strategy.

The Nürburgring 24 Hours is not like that.

It is rough, chaotic, and sometimes almost unreasonable.

But it is alive.

When I think back now, the clearest image in my mind is not a final result sheet.

It is the collapsing feeling when car No. 3 was pushed into the garage.

It is the unreasonable excitement of the No. 81 M3 Touring drilling forward through GT3 traffic.

It is the forest at night, the rain, the barbecue smoke, and the absurd cuteness of a Dacia Logan appearing in the same race.

It is the No. 80 AMG crossing the line after all that waiting finally became real.

That is what makes endurance racing powerful.

It does not only show you speed.

It shows you time.

Twenty-four hours will wear down and reveal the emotions of a car, a team, a driver, and a spectator.

At the end, what remains is very simple.

Do you still want to keep watching?

I do.

And I am already thinking about the next one.

Finally, from about 2,000 high-quality media photos, I picked some of my favorites (constantly updating...)

Every photo records a bit of the excitement that spilled out of the screen.

Thank you to all the photographers on site.

Really, thank you for bringing those 24 hours in the Green Hell to us in front of our screens.

That is all.

Thank you for reading my article. See you next time.

At the beginning of this year, places like Sushiro and Hama Sushi seemed to go from something I only occasionally saw online to places that people around me were actually visiting. Even better, there is a Hama Sushi inside Qiyi Plaza, right beside my school.

Since it had already opened next to my school, not going felt a little hard to justify.

So on May 2, 2026, I went in and had a big meal.

The Chinese title, "going to have a big conveyor-belt sushi meal", is borrowed from Crayon Shin-chan. I think it is perfect because it is not refined at all, and it does not pretend to be. It is just a very Shin-chan kind of happiness. You see conveyor-belt sushi, and you want to eat a big meal.

This meal was basically that.

The whole Hama Sushi routine is actually pretty good for going alone, or with one or two friends.

I went relatively late this time, so I got seated almost without waiting. That really added to the experience, because once conveyor-belt sushi starts with a queue, your expectation keeps rising. After you finally sit down, it is easy to order more with a bit of compensation in your head. Sitting down without waiting felt much smoother, like walking over from school and casually adding a small reward to the day.

After I sat down, the thing that really started making me lose control was the ordering tablet.

There were hundreds of items on it, sushi, sides, noodles, desserts, drinks, with all kinds of photos constantly coming at me. The key detail is that it does not let you add things to a cart first and then review everything at the end before placing the order.

You tap it, and it is ordered.

That design is powerful.

At first, you only want to see what else they have. But the menu is designed so carefully that, as you keep scrolling, thoughts like "this one looks worth trying too" keep popping up. Scroll back and forth a few more times, and your appetite gets pulled up round after round. Before you know it, you have ordered a few extra pieces of sushi.

You are not just ordering a meal.

You are constantly feeding coins into your own curiosity.

The first plate I ordered was caramel-seared salmon. That name is hard to resist. Seared, caramel, salmon. Put those three words together and it feels like the menu is actively pulling your hand toward it. After that, I ordered large-cut tuna and large-cut foie gras. One was a hit of red color; the other was the kind of thing where seeing the words foie gras makes you instinctively feel that you should try it.

Once those plates arrived, I roughly understood why conveyor-belt sushi has become so easy to like recently. It is not the kind of restaurant where you have to sit upright and evaluate everything seriously. It feels more like a small amusement park. Each plate is small, the pressure is small, and if you like something, you keep ordering. If you do not like one plate, it does not ruin the meal.

That part matters.

The caramel-seared series was the most memorable group in this meal, especially the caramel-seared salmon.

The salmon was rich in fat, and the first bite was soft and delicate. After biting into it, the experience did not end immediately. With the sushi rice, after a few more chews, the sweetness of the caramel, the aroma of the sear, and the salmon's own fatty fragrance slowly mixed together. By the end, my whole mouth was filled with that slightly toasted, rich feeling.

This was one of my favorite plates of the meal.

The large-cut foie gras was another one.

The outside of the foie gras carried a little seared aroma, and once it entered my mouth, that fatty fragrance was basically impossible to resist. Even more dangerous, the inside was extremely tender. It was not heavy in a stiff, forced way, but melted in my mouth like silken tofu, spreading that oily aroma through the whole mouth all at once.

By that point, I did not really want to describe it with professional words anymore.

Anyway, it was just happy. Umami~

The caramel-seared eel followed the same logic. Eel already has some weight to it, and after stacking caramel and searing on top, it is clearly aiming for richness. After it entered my mouth, the eel flavor and the fat under the eel skin pushed the happy mood up another notch.

Sometimes this is where conveyor-belt sushi wins. It lets you eat in a wonderfully carefree way. You can study the rice, fish, and sauce, or you can simply think that a plate looks good and tap it.

Then, a few minutes later, it slowly rolls toward you on the track.

Following that rhythm, I also ordered onsen egg udon, a tempura platter, and Japanese miso ramen.

Once these dishes arrived, the mood of the table changed immediately. Before, it was still the small happiness of plate after plate of sushi. Then suddenly, it became a proper dinner.

When the onsen egg udon first arrived, I honestly did not think there was anything especially new about it. It looked like a bowl of udon with an onsen egg, bonito flakes, and some tempura bits on top.

But once I mixed the onsen egg in, it became different.

The egg, the bonito flakes, and the small amount of broth at the bottom mixed together, and the freshness and smoothness came out all at once. The udon itself was slick and chewy. When it carried that broth and egg into the mouth, it was deeply satisfying. The tempura bits also gave little crunchy moments from time to time, so the bowl was not only soft and smooth, but had small surprises too.

Eating it was pure enjoyment.

Sushi takes care of curiosity, while noodles settle the stomach.

The tempura platter was also a very steady presence.

Fried food does not need much complicated reasoning. As long as it is hot, crisp, and fragrant, it has already done its job. Ordering tempura in a conveyor-belt sushi restaurant feels subtle too. It is not the main character, but it makes the meal feel more complete.

By the time I ordered caramel-seared inari sushi, I had truly entered the state of wanting to try everything I saw.

That is the scary part of the ordering screen.

The menu is always there, the photos are always there, and the prices and portions do not make you hesitate for very long. You only wanted to take one more look, but your hand has already tapped.

The caramel-seared inari sushi, shellfish trio, and low-temperature roast beef all came from that state. They were not things I had planned from the beginning. They appeared because, in the second half of the meal, the menu could still show me something new. My curiosity had not fully stopped, so I added a few more plates along the way.

Looking back at this meal now, I think the part most worth recording is not that any single dish was incredibly stunning.

It is that the whole process felt relaxed.

Walking over from beside school, entering Qiyi Plaza, sitting down, taking a number, ordering, and barely needing to talk to a server throughout the whole process. I just sat there quietly, looking through the endless dishes on the tablet, quietly thinking about what I should order, slowly considering which things were worth trying, and then waiting for the small plates to arrive one by one. Each plate felt like a small checkpoint, breaking an ordinary meal into many little moments of expectation.

That kind of expectation actually fits student life pretty well.

Classes, assignments, and deadlines often press down as one heavy block. But inside a conveyor-belt sushi restaurant, happiness gets sliced very small. A plate of salmon, a bowl of udon, a tempura platter, a receipt. Separately, none of them is a big event. Together, they make one evening feel very happy.

So if you ask me whether Hama Sushi is worth going to, my answer is probably yes. It is worth picking an evening when you are not in a hurry and going once.

Do not go with a very serious restaurant-review mindset.

Treat it as an adventure, experiencing that little sense of ceremony, trying ingredients you have not tried before, and feeling the happiness of ordering something you really like.

Order a few plates of sushi that interest you, order a warm bowl of noodles, and add something fried. At the end, when you look at the empty plates on the table and the receipt, you may find that the most interesting part of the meal is that it really does feel like the line in the title.

A big conveyor-belt sushi meal.

At first, it was only an April Fools' joke BMW posted on social media: an image that understood exactly what car fans would get excited about, a piece of brand mischief that looked likely to end the moment it was posted.

BMW really turned an M3 Touring into a race car capable of running the Nürburgring 24 Hours.

Not a static show car.

Not a comment-section wish.

Not an absurd April Fools' post that disappeared after the day ended.

It actually went on track. In March 2026, it already made its debut in the second round of the Nürburgring Langstrecken-Serie, winning the SPX class and finishing 13th overall. It is also set to run the main Nürburgring 24 Hours race on May 16-17.

When I saw this, my first reaction was not to look up the specs.

One phrase jumped straight into my head: evil big BMW.

In racing circles, people sometimes use "evil big BMW" to tease BMW. The joke works because it is not purely an insult, and not purely praise either. It's like knowing full well this thing doesn't play by the rules: you know it may be big, heavy, fierce, and filled with a strange kind of German engineering obsession, yet you still cannot help staring at it.

The M3 Touring 24H has exactly that flavor.

Evil. Too evil.

Before going further, let me get the name right first.

Many people may casually call it the M3 Touring GT3. That name is easy to understand, because it really does look like a GT3-ized M3 Touring: wide body, rear wing, racing stance, Nürburgring narrative, the whole package.

But the official name is BMW M3 Touring 24H.

It runs in the SPX class at the Nürburgring 24 Hours, not the official GT3 class. It borrows from the technical system of the BMW M4 GT3 EVO; the engine, chassis, and racing development logic are all related to that set of things. But it is not a customer racing car built to replace the M4 GT3 EVO.

That makes me like it even more.

If BMW had simply built another standard GT3 car, it would still be cool, but not nearly as interesting. A standard race car has the logic of a standard race car: fast, stable, maintainable, sellable to customer teams, and capable of winning races. That is a serious industrial product.

The M3 Touring 24H is different.

It feels like a half-joking sentence from a car fan comment section was heard by an engineer, and then actually carried into a wind tunnel, workshop, testing program, and race track.

That is just insane.

Think about it. A wagon is already a strange species. It does not claim the family-car role as naturally as an SUV, and it does not present itself as purely as a two-door sports car. The charm of a wagon is that it always carries a little everyday refusal to give in.

I can carry things.

I can pick people up.

I can go on long trips.

But I can also go mad.

That is why the M3 Touring is so attractive. It does not average performance and practicality into a compromise. It stuffs two desires that should conflict directly into one body. It can go to the supermarket during the day, to a mountain road at night, to a track day on the weekend, and still have a trunk full of completely unromantic daily clutter.

Sometimes I think people who like high-performance wagons all have a bit of this problem.

They say they have grown up, that they have started considering practicality, that they can no longer buy such inconvenient cars.

Then they turn around, see an M3 Touring, and their eyes become very honest.

Because it does not give you the feeling of compromise.

It gives you an almost shameless sense of completeness.

Honestly, BMW took far too long with this. The E46 M3 Touring Concept back in 2000 had already proved that M3 and Touring were not incompatible. That car was like a gun hanging on the wall, and it stayed there for more than twenty years.

Of course car fans kept thinking about it.

What if the E46 M3 Touring had gone into production? What if the E90 had a Touring? What if the F80 had one? These questions had no answers, which made them even easier to grow in the mind.

Not until 2022 did the G81 M3 Touring finally appear.

What feels best about it is that BMW did not turn it into a gentle version of the M3. It still follows the hardware logic of the M3 Competition: S58 twin-turbo inline-six, M xDrive, wagon body, M3 temperament.

It is not a grocery car with an acceleration pack.

It is a real M3 wearing a body that can carry more things.

So the later M3 Touring 24H works not because the April Fools' joke itself was clever, but because the line before it had been laid for too long.

The E30 M3 gave M3 its racing bloodline.

The E46 M3 Touring Concept left a blank space.

The G81 M3 Touring filled that blank.

The M4 GT3 EVO provided racing technology.

Then BMW lit the fuse on April Fools' Day 2025.

If it had only lit the fuse, that would not be special. Brand social media teams light fuses every day. They post an image, the comment section is lively for two days, then everyone disperses and the next round of content begins.

What is especially cheeky this time is that they did not stop at the image.

Even cheekier, the livery itself plays with the joke.

It is not just a random racing livery. You can see comments from the original social media post on the bodywork. In other words, the people who egged BMW on under the April Fools' post, wished aloud, and said "then actually build it" were not merely screenshotted by brand marketing for a report.

Their comments were put on the car.

That is exactly the "evil big BMW" flavor.

You thought you were just casually fanning the flames in the comment section. BMW turned around and carried the fire to the Nordschleife.

In March 2026, this M3 Touring 24H already made its debut in the second round of the Nürburgring Langstrecken-Serie, winning SPX and finishing 13th overall. As I write this on April 27, 2026, it has not yet run the main Nürburgring 24 Hours race on May 16-17.

So the story has not truly ended.

But it has already stepped out of the screen.

I think car fans eat this up more than anyone.

Not because car fans are easy to fool, but because everyone knows too well the disappointment of concepts that remain only as images. You see a handsome concept car, a wild design study, a visual package that understands you perfectly, and inside you know there probably will not be a next step.

It is only responsible for making you repost.

It is not responsible for truly existing.

The M3 Touring 24H is different. It really has to face the Nordschleife. It has to pass technical checks, run NLS, and deal with endurance, tires, traffic, weather, and the risks of the night. The Nürburgring is cruel. You can be very good at social media, but once you enter the track, all posture is judged again by asphalt.

So I think the most precious part of it is that small piece of reality.

Looking across the landscape, it is also not the most rational answer.

The Audi RS 4 Avant is steadier, more like a fast blade you can keep around for a long time. You do not need to adapt to it too much. It knows what it is doing: all-wheel drive, wagon, fast, useful. Its emotions are less outward, but it is very mature.

The Mercedes-AMG C 63 S E Performance Estate is more like a technical declaration of the regulatory era. Four cylinders plus an electric motor, very strong on paper, and highly engineered. But the C 63 name used to carry too much V8, sound, and rawness. It is strong, but some people will still get stuck on that one feeling inside.

The Porsche Taycan GTS Sport Turismo is another answer entirely. Electric response, low center of gravity, a beautiful long roof, thoroughly modern. It is of course fast, and fast in a very clean way. But for people still attached to engines, gear shifts, and mechanical connection, it lacks a certain visceral connection.

The M3 Touring stands in the middle.

It is not as restrained as the RS 4, not as technically aggressive as the C 63, and not as future-facing as the Taycan.

But it has something hard to replace.

Its identity is connected.

You see it and know it is an M3. You see the wagon tailgate and know it is not an ordinary M3. You see the large rear wing and racing body of the M3 Touring 24H, and your mind automatically connects M3, Touring, the Nürburgring, endurance racing, and car-fan wishes.

That is enough.

As performance cars develop, parameters alone are becoming less able to move people. Horsepower can be huge. Acceleration can be very fast. Electric cars can beat many fuel-powered performance cars into the ground. Lap time still matters, of course, but lap time increasingly feels like a private language among serious players.

What ordinary car fans truly remember is often a clear story.

An M3 Touring delayed by more than twenty years finally goes into production, and then is pushed onto an extreme stage that originally did not belong to wagons.

That story is very clear.

It is even a little childish.

But the most valuable things in car culture are often exactly this kind of childishness. Can we fit this in? Can we stuff that engine in? Can we make a wagon run the Nürburgring? Can you stop only showing me pictures and actually build it?

Very often, adults say they want rationality, but deep down this is what they most want to see.

You guys actually built it.

Those five words hit harder than a pile of specifications.

Of course, I am not saying projects like this can be copied endlessly.

If every year BMW turns another April Fools' joke into reality, and every time says they heard the fans, and every time builds a social-media highlight, the flavor will fade quickly. What makes the M3 Touring 24H work is not the rhetoric. It is the engineering investment, and the fact that it really has to face the track.

One less piece, and it collapses.

So I prefer to see it as a rare highlight.

It may not become a regular racing product, and it should not. The M4 GT3 EVO is the proper weapon in BMW's customer racing line. The M3 Touring 24H is more like a sudden bright interlude in the story of the M brand.

But this interlude matters.

It reminds me that car brands can still occasionally do things that are not so spreadsheet-shaped. Not every project has to immediately turn into sales volume. Not every answer has to be explained to a financial model. As long as the engineering is real, enthusiasm has somewhere to land.

I like this car probably because of that.

It is not rational enough.

It may even be a little unnecessary.

But in performance car culture, many of the things people cannot stop thinking about were never basic needs in the first place.

A large rear wing is not a basic need.

A wide body is not a basic need.

Sending a wagon to run the Nürburgring 24 Hours is certainly not a basic need.

But people like cars because they are often struck by exactly these non-essential things.

They make a machine more than a machine, a brand more than a brand, and make someone scrolling images in front of a screen suddenly feel that their childish love has been taken seriously once.

That is already rare.

I am putting this vertical image at the end purely out of personal preference.

If you have read this far, you probably did not click in by mistake.

You likely understand the feeling too.

A car may be impossible to buy. A Nürburgring 24 Hours race may only be watched through a screen. A pile of technical details may be forgotten in two days.

But a certain image, a certain stance, and the moment when an absurd idea actually lands can stay.

For me, the M3 Touring 24H is exactly that.

Not the most rational BMW.

But very much like the BMW in my heart that has not yet been worn smooth.

This time I did not continue making resume posters, nor did I seriously test any complex infographic. I only did a very personal little thing: I generated a set of posters for several F1 drivers I like.

And once again, I was blown away.

A few days ago, I had already written once about gpt-image-2. In that piece, I talked about why it is no longer just good at drawing images, so I will not repeat model release details, parameters, or capability boundaries today. Those things are certainly important, but for me, the truly interesting change is happening somewhere smaller.

It is the moment when you suddenly realize it can not only help you make a visual draft for work.

It can also turn your private taste into an image you genuinely want to keep.

F1 is a subject that suits posters very well.

The drivers themselves are personalities, and the teams bring their own colors, symbols, and visual language. Red Bull, McLaren, Ferrari, Mercedes. These names do not need much explanation. People who like racing see them and automatically recall a whole set of things: liveries, helmets, pit garages, start lights, radio messages, podiums, and the tension of staring at a live broadcast on a weekend night.

This is different from simply "generating a good-looking racing image."

An ordinary racing image can probably get away with being fast, flashy, and full of speed. But a driver poster cannot. It has to handle two things at once: public symbols and private emotion.

Public symbols are the parts everyone recognizes. Team colors, sponsor blocks, the sense of speed in motorsport, and the central position of the portrait. If these are lost, the image no longer feels like F1.

Private emotion is the part only you care about. Why you like this driver, why you are willing to turn his image into a standalone poster, why a certain color or posture feels right to you. These things are hard to write directly into a prompt. You can describe them, but description is never complete.

So when the generated image appears, if it happens to line up, you very clearly pause for a moment.

That pause is not simply because it "looks so much like him."

It is more like this: in your mind, there has always been a vague display cabinet containing drivers, teams, race memories, and your own aesthetic preferences. In the past, these things could only exist separately, in wallpapers, screenshots, or the feelings you carry home from a race weekend. Now gpt-image-2 suddenly gathers them into one poster.

That is incredibly satisfying.

The most easily underestimated value of AI image generation is not how much design cost it saves, nor whether an image goes from 70 points to 90 points.

It is more like helping ordinary people complete a kind of expression that used to be hard to complete.

You like F1. You like certain drivers. You want a visual collection of your own. This need is not big or grand, and it may not be worth hiring a designer to make a whole set. But it matters to you.

In the past, small needs like this were often suppressed.

Because the cost was too high. You had to find references, assemble images, adjust colors, cut out subjects, and design the layout. Once you really began, half your patience might be worn down before your enthusiasm faded.

But now it is different.

You can quickly throw that feeling to gpt-image-2 and let it give you something visible first. It may not be perfect, and it definitely still needs human judgment: whether the person is accurate, whether the team elements are appropriate, whether the text has problems, whether the scene is overly dramatic.

But it has already pushed past the hardest step.

The distance between "I want a poster like this" and "where can this poster be improved" has suddenly become shorter.

That is the strongest feeling I had this time.

When I wrote about the resume poster, what excited me more was that it had started to organize information. That was a workflow judgment. It turned a blank canvas into a draft that could be discussed, where you could point out what to change, what to delete, and what to strengthen.

This set of F1 driver posters excited me in a more personal way.

It made me realize that generative models do not only help us finish tasks faster. They also make some very small, very personal wishes, wishes that previously did not justify starting a complex process, suddenly executable.

Make a set of posters for drivers you like.

Make a series of collectible cards for a game character.

Turn a travel memory into a movie poster.

Generate a visual keepsake for a song, a car, a relationship, or a city you love.

From a business perspective, these things may not be large. From the perspective of personal life, they carry weight.

Because people naturally make room for what they love. Some buy models, some put up posters, some keep tickets, some save race screenshots. We keep these things not because they are rare, but because they anchor a part of who we are.

This is where gpt-image-2 amazed me again.

It did not simply make F1 look cooler.

It let someone who likes F1 suddenly own a set of images that could be placed into his own world.

That sounds small.

But I think it matters.

Because the most interesting moments in AI are often not when it appears all-powerful, but when it happens to catch a very specific thought of yours. You did not prepare a requirements document, you did not prepare to start a project, and you did not prepare to do anything very serious.

You just suddenly wanted to try.

And then it really gave you a surprise.

Just like this set of F1 driver posters on April 25.

I had only meant to play around.

Turns out, it took me seriously once again.

At first, I did not take it too seriously.

Over the past two years, AI image generation has produced so many astonishing moments that everyone has become a little numb. You give it a sentence, and it gives you a beautiful image. The lighting is good, the atmosphere is good, and the details are quite full. And then?

And then, often, there is no "then."

Especially when it comes to real work, problems appear. You want it to make a usable poster, and it starts inventing text. You want it to organize a pile of information, and it turns the layout into a colorful PPT cover. You want it to balance a person, resume information, technical direction, and visual style, and it often tries very hard, then falls apart spectacularly.

So in the first two days after release, I browsed what other people had made with ChatGPT Images 2.0. It was indeed impressive, but I was still only watching from across the screen.

Until today, when I tried it myself.

I gave it my own information and asked it to make a personal resume poster.

When the image appeared, I was genuinely stunned for a few seconds.

To be honest, my expectations for AI image generation used to be quite fixed.

It can make atmosphere. It can make concepts. It can make those emotionally satisfying images that look great at first glance: cyber cities, future laboratories, rainy neon nights, lonely backs, people in trench coats standing in front of giant screens.

Those images are of course good-looking.

But between that and a real design deliverable lies a very wide river.

The truly difficult part is not whether it can draw something that looks right, but whether it can place information properly. Where should the name go? How should the technical direction be presented? How should project experience be layered? How do awards and skills avoid fighting each other? Does the whole image look like a unified visual system, rather than a pile of assets forced onto a canvas?

A resume poster is exactly a stress test for this kind of scenario.

It cannot only look good.

It also has to explain things clearly.

It cannot only feel technological.

It also has to look like a real design draft that could be shown to someone.

That is the hard part. Because these were exactly the areas where AI image generation used to fail most easily. Text went wrong. Layouts became chaotic. Information hierarchy collapsed. The portrait looked as if it had been borrowed from another model. The background turned into a mess of blue light, circuit traces, HUDs, and hexagonal grids.

I am telling you, that vibe is all too familiar.

Everything looks very AI, but nowhere does it feel like a human is actually doing design.

But this image felt different.

What surprised me most was not how similar the portrait was.

Really, it was not.

The portrait matters, of course. But when I saw this image, my first reaction was instead: it is actually trying to handle information seriously.

It did not only stare at the face, nor did it just paste in the familiar technological trio of blue light, circuits, HUDs, and hexagonal grids. It placed the name, HPC, project experience, awards, and technical direction into a relatively unified layout. The whole image had clear visual hierarchy, whitespace, and a visual center of gravity.

It is not perfect.

Small text definitely still needs manual proofreading, and some details cannot be used directly as a final version. For example, whether certain descriptions are accurate, and whether certain layout choices fit the seriousness of a real resume, still need a human to close the loop.

But the point is that it is no longer the kind of image you can only post to social media with a comment like "so cool."

It feels like a draft that can be discussed.

That is a big change.

Because the most painful part of many creative tasks is not the final 20 percent of refinement, but the blankness at the beginning. You know you want something, but you cannot quite say what it should look like. You open the design software, the canvas is bright white, the mouse drifts around for a long time, and eventually you go looking for references. One hour disappears.

At this point, you should understand why I was excited.

What interests me about gpt-image-2 this time is not that it completes the design for you, but that it pushes the first visible thing in front of you.

You can say this text needs changing.

You can say this area is too crowded.

You can say project experience should stand out more, awards should not overpower the name, and the sense of HPC technology could be more restrained.

But at last you are no longer speaking into the air.

You are speaking to an image that already exists.

This is where AI tools truly begin to become useful. Not by generating a perfect final product with one click, but by pulling your idea out of the fog and turning it into something that can be seen, criticized, and modified further.

Many discussions about AI creation go too easily to extremes.

One side says designers are finished.

The other side says AI will never understand real aesthetics.

I do not particularly want to stand with either judgment.

Because what is really happening may not be that dramatic. It is more like one very annoying part of the workflow suddenly becoming thinner. It does not mean there will be no designers from now on. It means that the pain of forcing out the first draft from scratch may be reduced a lot.

You still need judgment.

You still need aesthetics.

You still need to know what information should be emphasized, what details must be removed, and which parts look high-end but are actually cheap tricks.

It is just that these abilities used to have to pass through a large blank canvas before they could take effect. Now the tool may first give you a target.

That is very different.

Back to gpt-image-2.

The official help center says ChatGPT Images 2.0 now covers all subscription tiers, and the developer documentation lists gpt-image-2 as the latest GPT Image model, with a snapshot available at gpt-image-2-2026-04-21. This information is certainly important, but simply repeating parameters and release cadence is not that interesting.

What is really interesting is that it is pushing image generation from pretty pictures toward complex visual tasks.

Posters, infographics, handouts, resumes, event materials, product images, course covers. These things were not impossible to make with AI before, but they often made you sigh halfway through. Because it was very good at generating, but not very good at organizing.

This time, for the first time, I felt quite strongly that it is starting to organize.

Of course, it is still early to say this.

One image cannot represent everything. Especially with a newly released model, it is easy to get excited, and only after calming down and testing a dozen more rounds can one draw a serious conclusion. I will probably keep trying other images later, especially harder scenarios such as multilingual text, complex infographics, and project showcase pages.

But today's resume poster is worth recording separately.

It made me realize one thing.

The most valuable direction for AI image generation may not only be helping ordinary people make a beautiful picture.

More importantly, it lets ordinary people more quickly obtain something they can begin to modify.

From blank page to first draft.

That step alone is already huge.

The best part of this kind of meal is that it does not hurry you. Once the skewers are placed beside the brazier, you cannot eat them immediately. You can only wait as the charcoal slowly draws out the fat. When the mountain-raised pork, spicy tender beef, grilled pork belly, and grilled pork skin were laid out one skewer after another, the table already had a solid, smoky warmth to it.

I have always felt that firepot barbecue suits a gathering of friends better than many forms of barbecue that arrive already cooked. It naturally slows the pace down: someone watches the fire, someone turns the skewers, someone warns that a piece is about to burn, and someone takes the first bite. Before the meat is ready, the conversation has already connected; by the time the meat is ready, the mood at the table has warmed up too.

What impressed me most during this meal was, unexpectedly, the seasoning plate. When eating barbecue, the most familiar seasoning is of course chili powder: fragrant, spicy, direct, and immediately present once you dip into it. But this time there was also soybean powder on the plate. I tried dipping grilled meat into it for the first time, and the result was surprisingly good.

Soybean powder doesn't seize the palate the way chili powder does. It is more like placing a soft layer of aroma underneath the charcoal flavor. The grilled meat has its own fat, and the charcoal carries a little roasted bitterness. After dipping it in soybean powder, the bite becomes rounder, and the aroma grows thicker. It is not simply "heavier tasting"; it is the sudden realization that barbecue can be eaten another way: less aggressive, but very lasting.

Fenjiu was poured slowly into small cups, the measures neat and just right. A plate of crisp peanuts sat at the corner of the table. No one felt the need to force conversation. We peeled a few casually, sipped amid the warmth of the charcoal, and let the conversation drift. The night was quiet, the liquor light, and the conversation drifting freely, relaxed and easy. An ordinary evening gained a few more gentle flavors.

Vegetables also fit this meal well when eaten slowly. Eggplant absorbs flavor, lotus root stays crisp, onion becomes sweet after grilling, and lettuce can wrap just-cooked meat. Under the slow heat of charcoal, the plump squid gradually shed its firmness, becoming tender and chewy. Wrapped in the clean sweetness and bite of onion, the smoky flavor seeped in just right.

Two Drinks Worth Noting​

We had Fenjiu at the table this time, and also Asahi beer. Putting them into the same barbecue meal was actually quite interesting: one is a representative of light-aroma Chinese baijiu, while the other is a beer built around a dry, clean finish.

Fenjiu is a classic representative of light-aroma baijiu, closely tied to Xinghua Village in Fenyang, Shanxi. The feeling of light-aroma baijiu is usually not heavy or stacked, but clean and crisp. At a barbecue table, drinking it slowly from small cups means it does not overpower the skewers or seasonings. Especially in a meal carrying the smell of charcoal, a baijiu that is too sweet or too thick can easily steal the scene. Fenjiu's cleaner style pairs well with the richness of grilled meat.

Asahi beer takes another route. Super Dry was born in 1987 and centers on a Karakuchi style: dry, crisp, and clean on the finish. Its fit with barbecue is also straightforward. Skewers have fat; chili powder and soybean powder enrich the flavors in the mouth. At that moment, a sip of cold beer with a clean finish can gently carry away the oil and spice from the previous bite, making you naturally want the next skewer.

So the drinks were not there to make the night extravagant. They added ease to the table. Fenjiu suits slow, small toasts. Asahi suits cleansing the palate after grilled meat. Together, they matched two rhythms in a gathering of friends: one for slow conversation, one for continuing to eat.

An Ordinary Night​

What I remember is not just what we ate and drank. More specifically, it is the red glow in the charcoal, the roasted aroma slowly forming on the edges of the skewers, the unexpected delight of dipping into soybean powder for the first time, and the ease of sitting beside friends without having to search for a topic.

Very often, the things worth keeping in life do not need to be grand. A barbecue, a few skewers, a plate of soybean powder, a little drink, and several people willing to sit together and wait for the meat to cook are already enough to become a real footnote to an evening.

The more I think about it, the more I feel that many people have been deeply deceived by the word "passion." It is as if something is not worth doing unless you genuinely love it from the bottom of your heart. That is not true. Looking at reality plainly, anything that can support you, help you stand firm, and let you accumulate ability already has value. Survive first. Become stronger first. Then talk about freer choices.

If you really want to make money, do not begin by asking other people, "How do I make money?" Behind that question, you should first ask something else: why would they be willing to tell me? What they are telling me, is it a way to make money, or a way to turn me into their income?

In recent years I have become increasingly wary of one kind of sales pitch: on the surface, it teaches you how to enter the game; underneath, it monetizes your anxiety. There is a signal that is not absolute, but very useful, when judging whether an industry is starting to decline: are more and more people inside that industry beginning to teach others how to make money?

When an industry is truly still rising, most people keep putting their energy into doing the work, because doing the work directly is more profitable. Only when some people find that the easy money on the ground is no longer so easy to earn do they start dismantling their experience and packaging it into courses, communities, consulting, and methodologies for a second round of monetization. To put it bluntly, you may well be the final wave they plan to earn from before exiting.

So a person needs to be brave, and also strong. Bravery means daring to judge for yourself, instead of being easily led by other people's success gospel. Strength means being able to distinguish a real opportunity from bait someone laid out while planning their own exit.

On April 7, I spent some time reading LangChain's article, The Anatomy of an Agent Harness.

At first, I was drawn to the article by the word "harness."

It has been appearing more and more often in the Agent world recently, but it is also a slightly awkward word. If you translate it directly as "tack" or "restraint gear," it sounds strange. If you translate it as "framework," it feels too light, as if it is only a few layers of code wrapping. "Exoskeleton" may be a bit closer, but it is still not exact.

In any case, it points to this thing:

the whole system outside the model that lets an Agent actually do work.

My strongest feeling after reading it was that, in the past, we may have stared too much at the model itself.

Which model is smarter.

Which model reasons better.

Which model writes code more aggressively.

These things are of course important. You cannot take a very weak model and insist that engineering alone can make up for it. That is unrealistic.

But if you have really used Claude Code, Codex, Cursor, or built Agents yourself with LangGraph, the OpenAI Agents SDK, or CrewAI, you quickly run into an annoying problem.

A smart model does not equal a reliable system.

It can think of the next step, but that does not mean it knows when to stop.

It can call tools, but that does not mean it knows which tools must not be used carelessly.

It can write files, but that does not mean it knows how to roll back a bad change.

It can read context, but that does not mean it will not tie itself in knots after a dozen turns.

It is like putting a very smart person into a car with no steering wheel, no brakes, and no dashboard.

The brain is good.

The car is not.

What struck me most in the LangChain article was that it made this problem very concrete. It was not vaguely saying that Agents need engineering. It took the harness apart and showed you planning, filesystem, subagents, stateful middleware, context engineering, tool permission, and evaluation loop.

None of these words look exciting.

But every one of them is critical.

Take planning, for example.

Many people may think a plan is just asking the model to write a todo list. Sounds ordinary, right?

But think about what most often goes wrong when an Agent does a complex task. It is not that it completely cannot do the work. It is that it drifts while doing it. It starts out fixing a bug, then halfway through begins refactoring the whole project. It starts out writing an article, then halfway through adds a pile of background material barely related to the topic.

At that moment, a plan is not decoration.

A plan is a rope.

It is not there to restrict the model for the sake of restriction. It is there so that every so often the model can look back and see what it is actually doing.

The filesystem is the same.

We used to think the longer the context window, the better. The longer the window, the more the model remembers, and the stronger the Agent becomes.

That is only half true.

Long context is useful, but you cannot stuff everything into it. Code repositories, logs, drafts, materials, previous attempts, intermediate results. Once everything goes into the context, the model looks as if it knows everything, but in reality its attention becomes scattered.

The filesystem is another brain.

Write down what should be remembered. Read it back when it should be read. You do not need to stuff the whole world into the prompt. You only need the model to know where things are, and when to retrieve them.

This has also become one of my strongest feelings recently.

An Agent is not a model that is better at chatting.

An Agent is a model placed into an environment.

Once the environment becomes complex, the problem changes.

The early ReAct pattern had already exposed this issue. The model thinks one step, acts, observes, and then continues thinking. Back then it was still a very simple loop of Thought, Action, and Observation, almost like hand-building a small operating loop inside a prompt.

Then AutoGPT had its moment. For the first time, many people saw a model decomposing tasks, searching, writing files, and continuing to iterate by itself. It felt astonishing.

And very soon, people saw the other side.

It gets lost.

It falls into loops.

It works very hard in a large circle, and finally hands over something no one knows how to evaluate.

That moment already proved that a model's ability to think is not enough. You still have to give it state, permissions, tool boundaries, error recovery, and places where humans can intervene.

Back to the harness.

I think what LangChain's article really wanted to say was not, "We have invented another new concept."

It was more like a reminder that a large part of Agent capability lives outside the model.

There was an interesting example in the article. They said that with the model unchanged, changing the harness moved a coding Agent from Top 30 to Top 5 on Terminal Bench 2.0.

That result should not be mythologized. A benchmark is always only a slice.

But it is enough to make the point.

The same model can perform very differently when placed inside different systems.

This is actually similar to people.

A smart person without schedules, documents, collaboration tools, review mechanisms, or reminders about priority can also become chaotic. Conversely, someone of more modest ability can still produce steadily if they have good processes, notes, tools, and feedback systems.

Agents are the same.

The model is the brain.

The harness is the work habit.

At this point, I suddenly understood why products like Claude Code are so hard to put down.

It does not only give you a model that writes code better. It gives the model a very concrete work setting. It can read a repo, edit files, run tests, inspect errors, confirm dangerous operations with you, and keep moving through a task.

Behind those experiences is the harness.

The OpenAI Agents SDK is moving in the same direction. It brings concepts like agent, handoff, guardrail, session, and tracing into the SDK. Anthropic's MCP is, in a sense, trying to standardize the way models connect to external tools. LangGraph feels more like a controllable Agent runtime for developers, so that state, branching, human intervention, and durable execution do not all have to be handmade from scratch.

You can see it: everyone says they are talking about Agents.

But the real competition increasingly looks like a competition over whose shell is better.

This matters to ordinary developers too.

If you are only playing around, pick whatever product you like. If Claude Code feels good, use Claude Code. If Cursor suits you, use Cursor. If the OpenAI Agents SDK is convenient, use the OpenAI Agents SDK.

But if you really want to put an Agent into your own business, you cannot only ask whether the model is strong.

You also have to ask where its memory lives.

Who manages tool permissions.

How failure recovers.

Whether logs can be replayed.

Whether humans can step in at critical points.

Whether this workflow can be carried over when you change models later.

This question will only become more important.

Because after using Agents for a long time, the most valuable thing may not be any single answer, but the working style accumulated around it. It knows how you write code, how you organize materials, your team's process, which commands are dangerous, and which files should not be touched.

If all of that sinks into a closed-source product, you may feel very comfortable in the short term.

In the long term, you may feel uneasy.

This is not to say closed-source products are bad.

I use them too, and in many cases they are indeed good.

But we need to know what we are handing over.

We used to say that model vendors sell intelligence. Now it seems that, in the future, they may sell an entire work environment. The longer you work inside it, the more it understands you, and the harder it becomes to leave.

What a time to be alive, my friends.

So my mindset when looking at Agents is no longer quite the same.

When I saw a new Agent demo before, I would first ask what model it used.

Now I first ask what its harness is.

Does it have a planning system?

Does it have external memory?

Does it have tool permissions?

Does it have observability?

Does it have rollback and evaluation?

Is it a smart brain that can chat, or a small system that can truly do work reliably?

That is the difference.

After reading the article on April 7, an image kept running through my head.

A model stands at the center of the stage, and everyone is watching how smart it is.

But lighting, sound, teleprompter, backstage coordination, stage machinery, and emergency brakes are all hidden in the dark.

The audience sees the actor.

What really decides whether the performance can finish smoothly is the whole theater.

Agents are the same.

We will of course keep chasing stronger models.

But perhaps from now on, we also need to look carefully at the theater beneath the model's feet.

"Paper is dead" is of course a deliberately provocative phrase. But what it really pierces is not whether papers have value, but a deeper question: when AI is systematically rewriting the research workflow, what should researchers use to prove their value? And what should count as the real deliverable of future research?

After listening to the full talk, my strongest feeling was a question I kept turning over: has the AI era made research easier, or harder?

When Papers Are No Longer The Only Endpoint Of Research​

For a long time, papers were almost the most typical representation of research output. Whether a study was considered complete often depended on whether it could be written into a paper and published at a sufficiently good conference or journal. Papers are of course important. But if the endpoint of research is completely equated with publication, that paradigm is now being rapidly challenged by the AI era.

Professor Xu repeatedly emphasized that research outputs with real vitality in the future should not only be papers. They should also include tools, APIs, benchmarks, ground truth, and other infrastructure that Agents can call. In other words, the value of research is no longer reflected only in citations. It is also reflected in whether the work truly solves real needs, whether it can be called, reused, extended, and eventually enter a long-running technical ecosystem.

This is, in essence, pulling research back from "writing it out" to "building it out."

Around this point, the talk also mentioned an evaluation framework worth remembering: ROSE. A good research work should solve a real need, be original, be significant, and retain a certain elegance. Compared with asking "where was this paper published," this standard is obviously harder, but also closer to research itself.

Because truly good research does not necessarily stop at a paper. It is more likely to continue becoming tools, systems, and methods, and to keep living inside later work.

The Stronger Agents Become, The More They Need A Regulatory Kernel​

The second main thread of the talk was system governance in the Agent era.

Today, when many people talk about Agents, they focus more on the expanding boundary of capability: Agents can understand tasks, decompose workflows, call tools, and in some scenarios even show signs of near-autonomous action. But the more this is true, the harder it is to avoid one question: if Agents become increasingly powerful, who ensures that their final behavior remains safe, reliable, and controllable?

Professor Xu's answer was a "regulatory kernel."

This concept is very interesting. It is not simply adding a few rules to a model. Rather, it tries to build a constraint structure between the large language model and the external world, somewhat like an operating system. The large model is responsible for open-ended semantic output, for proposing plans and creating possibilities. The regulatory kernel maps those open semantics into deterministic instructions that can be inspected, constrained, and executed, and then uses boundary checks and rule interception to ensure that the system does not cross lines or lose control.

In one sentence: the LLM thinks; the kernel guards.

Behind this lies a crucial cognitive shift. A mature Agent system cannot be built only by stacking "smarter" models. It must have a verifiable and governable underlying mechanism. System design in the AI era is not only a question of capability. It is also a question of boundaries.

What AI Replaces Is Craft, Not Vision​

The point I agreed with most in the entire talk was that it did not treat AI simply as an "efficiency tool." It went further and asked: when tools become stronger and stronger, what remains irreplaceable in a researcher?

Professor Xu's judgment was clear: AI will indeed flatten many differences at the skill level, but it will not flatten research ability itself. What automation more easily replaces is repetitive labor, churning out shallow papers, and routine work built only on proficiency. What remains difficult to replace is insight into problems, judgment about directions, the ability to distinguish reliable information from falsehood, and the ability to move a vague problem toward a verifiable result.

This means that research training in the AI era has changed.

In the future, more important than "can you use tools" will be: can you propose a question worth studying? Can you communicate with AI at a high level? Can you judge which results are genuinely effective and which are merely hallucinations that look like answers? Can you push a piece of work from an idea into an outcome that can truly be reused, communicated, and deployed?

The tools are getting stronger, but the abilities that decide the ceiling are becoming more fundamental.

The Q&A Made The Researcher's "Self-cultivation" Clearer​

Several points from the Q&A after the talk are worth recording separately.

Someone asked whether AI would make the gap between people larger, especially whether those already good at organizing, judging, and leading would pull further ahead with AI support.

Professor Xu's answer was not pessimistic. He did not think research ability would become permanently locked in the hands of a few "naturally gifted" people. Large models may well amplify efficiency gaps. But research thinking itself can be trained. How to sense the value of a problem, how to judge whether information is reliable, and how to organize one's research process are not mysterious talents that cannot be learned. They are abilities that can be built through long-term training.

He also mentioned that when reading papers, the most important thing is not only to see what conclusion the authors reached, but to ask: how did they get onto this path? What a mature researcher truly needs to cultivate is not only a store of knowledge, but research taste and judgment. You do not only need to "understand a paper." You need to see how a problem was discovered, defined, advanced, and solved.

This is very important. AI can help you summarize papers faster, extract information, and generate ideas. But if you lack judgment, you can easily be led around by content that is fluent on the surface but empty underneath. The stronger the tools become, the scarcer judgment becomes.

In The AI Era, Education Should Not Mainly Train "People Who Can Solve Problems On Tests"​

The talk also included a discussion about education that left a deep impression on me.

With the help of AI, knowledge acquisition and skill learning will become increasingly democratized. Many things that once required long accumulation are now becoming easier to reach and easier to call upon. But this does not mean education will become lighter. On the contrary, it may become harder.

Because when knowledge itself becomes easier to obtain, what becomes scarce is no longer how much content you remember, but how you judge what is worth learning, what is worth doing, and what is worth believing. What truly matters in the future may not be only teaching students to master knowledge, but cultivating judgment, the ability to choose, critical thinking, and the ability to form a sense of problems around real interests.

To put it more directly, AI can become an executor, but only if you first are a person with a sense of direction. If someone has no problem consciousness, stronger tools will only let them spin in place faster.

Closing​

So what "Paper is dead" really means is not that papers are useless, but that the era in which papers alone serve as the endpoint of research is ending.

Real "hardcore research" in the AI era asks researchers, on one hand, to produce outputs closer to real-world needs, leaving deliverables that can be reused, called, and incorporated into systems. On the other hand, it also asks researchers to cultivate again those most basic and most crucial abilities: problem consciousness, judgment, research taste, and the ability to work with new tools.

The tool revolution has already happened. What will really decide how far a researcher can go next may not be whether they can use AI, but whether, in the AI era, they can still see which problems are truly worth solving.

Appendix: Original Meeting Notes​

The following preserves my note-style record from the talk for reference.

Summary​

This session centered on changes to research paradigms in the AI era. It discussed the reshaping of research deliverables, the regulatory kernel in the Agent era, and core questions around research value, AI tool use, and educational models.

1. Reshaping Research Paradigms And Deliverables​

The core value of research should not be limited to paper publication and citations. It should also be reflected in whether the work can be called by AI Agents and generate real social impact.

Future research deliverables should include infrastructure such as tools, APIs, benchmarks, or ground truth that Agents can call, ensuring the long-term vitality and ecological position of the work.

The "ROSE rule" was proposed as a new standard for evaluating research: solve a real need (R), possess originality (O), have significance (S), and demonstrate elegance (E).

2. Regulatory Kernel In The Agent Era: OS System​

To address uncertainty, uncontrollability, and potential risks in Agent systems, the talk proposed building a "regulatory kernel" (OS) as the Agent's "rein."

This kernel maps semantic output from large language models into deterministic instruction sets, combines them with rules for boundary checking and interception, and forms a neuro-symbolic system.

The goal is for the LLM to create possibilities and formulate plans, while the regulatory kernel guards the boundaries and ensures safe, reliable execution.

3. Key Issues And Views​

Research value and "water papers": the value of hardcore research lies in deep insight into problems and real solutions, not simple repetitive labor. Automated low-quality paper systems lack cognitive gain, while high-quality research processes themselves benefit personal ability.

Use of AI tools: AI tools such as large models can effectively flatten skill-level gaps among researchers with different capabilities. But core "research thinking," such as how to propose valuable questions, how to communicate effectively with AI, and how to judge truth from falsehood, still requires training and cultivation.

Evolution of education: facing AI's impact on knowledge acquisition, education should shift toward cultivating students' core judgment, critical thinking, and cross-domain integration ability, rather than simply transferring knowledge. The role of the classroom should become guiding students to solve problems and engage in deep exchange.

To-do​

1. Open-source project release

Qiang Xu will soon release the regulatory kernel system (OS) developed by his team and invite the community to try it and provide feedback.

Q&A Notes​

Liu Weifeng: In the first year or two, my feeling was different from the last two years. AI seemed to flatten the research ability of mid-level and senior people, but recently it feels like the gap has widened. People who can use large AI models well are pulling ahead, and many of them have strong organizational and management talent, abilities they have had since childhood. These abilities let them make better use of AI. This gap seems hard to catch up with and may become fixed. How does Professor Xu view this?

Xu Qiang: I have a different opinion. The Industrial Revolution flattened differences in physical strength, and large models flatten differences in human skills, but I do not think they flatten research ability. The research way of thinking can be trained: how to sense the value of a problem, how to communicate effectively with AI, and how to judge whether information is true or false. If you consciously train research thinking, it will not be flattened by AI. So I disagree that research ability in the AI era is strongly tied to leadership and organizational talent. Xu also mentioned in his book that the gap will not necessarily become larger; research has patterns and can be trained.

There are too many papers to read, so you must select carefully. When reading papers, the main thing is to consider how the author arrived at this path, not just the research content itself. Improve your research taste and judgment, and then you can improve the efficiency and effectiveness of using AI for research.

What kind of model to use: use the best AI tools, and pay for the strongest models to improve yourself.

School education: with AI, knowledge and skills become more equalized, but judgment and the ability to choose still belong more strongly to experienced people, such as senior students and teachers.

Children's education: every era has its own needs. It should not be viewed through the lens of people like us who reek of that old-fashioned, preachy attitude. Knowledge and skills themselves are becoming more equalized, so education now needs to change in this respect. Children must have things they are interested in and love. If they love something, they will naturally have discoveries, needs, and ideas, and at that point they can use AI to work on them, so that even the hard work feels worth it because it serves what they love. People with ideas and ability will not be replaced. But if someone has no interests and no ideas, then AI really may replace them to a large extent.

That afternoon found its rhythm the moment we walked into the simulator shop. Once the triple screens, steering wheel, and pedals were set up, sitting down felt automatic. We drove from F1 to GT3 cars at the Nordschleife in Assetto Corsa, then switched to rally racing, and somehow even Euro Truck Simulator did not escape us. When people who all like cars stay together, happiness often needs very little setup. Circuits, mistakes, restarts, and teasing one another were already enough to fill the afternoon.

After the simulator session, we went to the market on the way and then returned to my friend's home. In the afternoon we were still battling the circuits; by evening we were busy around the kitchen. That switch was actually quite fun. I made garlic butter shrimp, steamed sea bass, stir-fried clams, and bitter melon with eggs. As several of us helped here and there and waited for the dishes to come out, the smell in the room slowly shifted from the freshness of spring outside into the unmistakable warmth of home cooking.

When we finally sat down at the table, a gathering where we bought groceries and cooked for ourselves did feel different from eating out. The steamed sea bass had a calm freshness. The butter shrimp carried a rich aroma. A large bowl of clams immediately made the table feel lively. The bitter melon with eggs balanced the meal with a home-cooked flavor. The photos also show a few other simple side dishes, but what I remember most from that night is that feeling of finally letting go after everyone had made the dishes one by one and could finally sit down.

By the time dinner started, two one-liter cans of beer were already on the table. We didn't drink that much. It was just enough, from the first bite of food onward, to push the whole evening into a comfortable place: faces warming a little, people relaxing a little. Later, my friend brought out some of the model cars he collects. Some had hoods that could open; some even had removable body shells. They were so detailed that you could not help leaning closer and looking slowly.

What stays in the memory is not just which games we played or which dishes we cooked, but how naturally everything connected that day. We talked about racing in the afternoon, cooked in the evening, and looked at model cars after dinner. Interest and ordinary life were not separated; they flowed into one full day. A gathering like this does not need to be grand, but it can make a day feel thoroughly lived.

That is the strongest and most dangerous part of the method of spiritual victory. It does not make a person truly stronger; it only makes them feel, for a moment, less embarrassed. It wraps failure in a nicer explanation, gives incompetence a respectable excuse, and packages retreat as the posture of "I never really cared anyway." On the surface, it looks like self-comfort. In reality, it is avoidance.

This has not disappeared at all today. You do badly on an exam, but instead of thinking through what you failed to understand, you say the grading system was flawed to begin with. You miss an opportunity, but instead of admitting you were not ready, you say you never wanted it anyway. You fail to do something well, but instead of reviewing where you went wrong, you first blame the environment, luck, or other people's lack of taste. The biggest benefit is that you feel better immediately. The biggest cost is that you stop right there.

Of course people need some self-comfort. Everyone has moments when they cannot hold it together. No one can be calm, clear-headed, and ready to review immediately after every failure. The problem is not comforting yourself. The problem is whether you turn that comfort into a long-term way of living. Taking a breath for a while is fine. What is truly frightening is that, after breathing in that story long enough, you begin to believe the words you invented.

The more I think about it, the more I feel that the method of spiritual victory ruins a person not by making them look ridiculous, but by taking away the desire to change. Once you can always explain away the past, it becomes very hard to force yourself to face the most painful question: where exactly am I lacking? That question hurts, but only after asking it can a person move forward. If you keep yourself upright only through stubborn words and self-anesthesia, it may look like you are protecting your dignity, but in fact you are slowly eroding the possibility of becoming stronger.

So what deserves the most vigilance in Ah Q is not that he is "funny", but that he resembles a common part of human nature. When people are hurt, they want to save face. When they fail, they first want to preserve their self-image. That is normal. But if a person is always willing to defend only that small inner dignity and never willing to face the loss in reality, then in the end that dignity is all they have left.

A truly strong person is not someone who never feels pain, never loses face, and never fails. A truly strong person is someone who does not rush to excuse themselves after failure. You can be sad. You can feel unwilling. You can even give yourself a little time to recover. But in the end, you still have to come back and honestly admit: this time, I did not do well. Only after that can improvement, growth, and reversal begin.