Anthropic vs. Nvidia? The Ugly Truth of the Big Tech CEO Clash at Davos | AI Deep Dive
Anthropic vs. Nvidia? The Ugly Truth of the Big Tech CEO Clash at Davos | AI Deep Dive
Right now, all eyes are on the 'Davos Forum'. The Davos Forum used to be a place for genteel discussions about things like climate change and poverty, right? But this year, the atmosphere is completely different. It's turned into a giant tech octagon.
Big tech CEOs like Elon Musk, Jensen Huang, and Satya Nadella gathered in one place, and instead of exchanging pleasantries, they engaged in a war of words that was nothing short of a brawl, just without actual weapons. On the surface, they talked about the rosy future of AI, but underneath, there was a scramble to protect their own interests and a palpable fear of a potential bubble burst.
The moment when Anthropic's CEO, Dario Amodei, openly took a shot at Nvidia was a real highlight. After all, Anthropic is a key customer that buys an enormous number of GPUs from Nvidia, right? Despite this, Amodei had this to say about Nvidia selling chips to China:
"An AI data center is like a country full of geniuses. Exporting chips is like handing over that entire nation of geniuses to China and giving them control."
He launched a fierce criticism, touching not just on technology but on international politics and trade issues.
And what about Microsoft's Satya Nadella? He revealed his blatant intentions by calling data centers 'token factories'. "Folks, if more people don't start using AI, this will all become a bubble and we'll all go down together," he said, in a comment that almost sounded like he was begging for users. He framed it as a warning against AI being concentrated in certain wealthy nations, but in the end, it reeked of a desperate plea: "Please use our services more."
Nvidia's Jensen Huang was no different. "We are not investing enough yet. We need to pour more money into this for the industry to work," he cried out. On the surface, he talks about creating jobs, but underneath lies the anxiety that they need to make the pie bigger before the bubble bursts.
What does all this mean? It's proof that the AI market is intensely competitive and, at the same time, built on a foundation of instability that could crumble at any moment.
But this brutal atmosphere isn't just limited to Davos. According to a domestic article I referenced, the Korean business world is on high alert. An analysis of the New Year's addresses from Korea's top 10 conglomerates revealed the most frequently mentioned keyword. What do you think it was? Yes, it was AI. A keyword that wasn't even in the top 10 last year skyrocketed to first place this year, mentioned a total of 44 times. SK has declared it will completely rebuild its business model around AI, and Samsung Electronics is also championing 'AX' (AI Transformation) innovation.
Now, this is where we need to be level-headed. Behind the brawl at Davos and the desperate declarations of Korean conglomerates, there lies one fundamental question.
"Who is actually trying to make real money?"
"And who can actually make real money?"
AI startups are popping up like mushrooms after rain. Legendary developers are starting companies, and trillions of won in investment are pouring in. But when you look under the hood, it's often hard to tell if they're trying to build a real business or just have fun doing interesting research with investor money.
So today, I've brought a very interesting analytical framework. It's a 5-level scale to measure the ambition of current AI companies. Using this scale proposed by TechCrunch, I will help you distinguish the real players from the fake ones in the market. Investors and those preparing to enter the AI industry, you'll want to perk up your ears and listen closely from now on.
First, let me explain what these five levels are. This isn't about how much money they're actually making, but rather a division based on how strong their intention to make money is.
Level 5 is the "We're already raking in fortunes every day, thank you" stage. Giants like OpenAI, Google, and Anthropic belong here. They already dominate the market and have solid, functioning revenue models.
Level 4 is the "We have a very specific, step-by-step plan to become the richest on the planet" stage. They might not be raking in as much money as Level 5 yet, but they have a clear business roadmap and are sprinting like mad.
Level 3 is the "Our idea is really great, we'll probably make money someday, right?" stage. They have good technology, but the concrete picture of how to monetize it is still blurry, and while they have many product ideas, the release dates are TBD.
Level 2 is the "We at least have a rough outline of a plan" stage. They are still in their infancy or are more focused on the technology itself rather than the business.
Finally, Level 1 is the "True wealth is loving oneself" stage. Sounds absurd, right? These are the pure academics whose goal is the research itself, rather than making money. Investors also tell them, "Alright, you just focus on research, we'll provide the money." But what's scary about them is that they have the potential to jump to Level 5 at any moment.
Alright, now let's use this scale to bluntly dissect four of the hottest AI labs right now. These are the ones that are the talk of Silicon Valley recently.
First up is Humans&. They've received a lot of attention recently. Their vision is to create communication and collaboration tools that go beyond simple computation with next-gen AI models. They say they'll create post-software to replace Slack or Google Docs.
But the problem is, despite the media praise, it's impossible to tell how they plan to make money. They say they'll build a product, but they're cagey about what exactly they'll sell. They're a classic Level 3 company: all talk about "doing something great" but with no visible business model. The technology is interesting, but from an investor's perspective, it makes you scratch your head.
Second is Thinking Machines Lab. Hopes were high for this one, as it was founded by Mira Murati, the former CTO of OpenAI. Who is Mira Murati? She was at the core of OpenAI, wasn't she? So until early 2026, everyone saw this place as a Level 4, a company with a solid roadmap. After all, they received a seed investment of over 2 billion dollars, that's more than 2 trillion won.
But something shocking happened in the last two weeks. Co-founder and CTO Barrett Zoph left, and five key personnel followed him out the door. Half of the executive team was gone within a year of its founding. The departing members said they were "concerned about the company's direction." What does this mean? It means that while they pretended to be Level 4 on the outside, a look inside revealed they only had a flimsy plan at a Level 2 or 3. They left thinking, "We thought we were joining a world-class lab, but it turns out there's no plan?" This is a very dangerous signal to the market.
Third, the comeback kid: World Labs. This company was founded by the godmother of AI research, Professor Fei-Fei Li. Since it was created by an academic giant, everyone thought, "Oh, that must be a research lab run by professors. It's probably a Level 2 at best." They were skeptical even when it received $230 million in funding.
But what a surprise! In just one year, it showed a completely different side. It didn't just stop at research; it commercialized a world-generating model that creates 3D spaces and released it as an actual product. Real demand is exploding in the gaming and special effects industries. They've accomplished something that other major labs haven't yet. We thought she was a scholar, but it turns out she's a real entrepreneur. World Labs is now a solid Level 4, and perhaps the strongest candidate to enter Level 5 soon.
Finally, we have Safe Superintelligence (SSI). This is the company founded by OpenAI's former chief scientist, Ilya Sutskever. Just look at the name. Safe Superintelligence. It doesn't smell of money at all, does it? They even rejected an acquisition offer from Meta and have no product launch plans, stating they will focus solely on superintelligence research. And yet, they've pulled in nearly 3 billion dollars, close to 4 trillion won, in funding. This is a classic Level 1.
But don't let your guard down. In a recent interview, Ilya Sutskever hinted, "If the research takes longer than expected, or if a very powerful AI can make an impact on the world, we might change direction." The AI world moves incredibly fast. This means that depending on their research results, a pure research lab could transform into a commercial monster overnight.
Everyone, why did I tell you this long story?
The AI market right now is pure chaos. At the Davos Forum, big tech CEOs are grabbing each other by the collar, screaming "Give us the money before the bubble bursts!", Korean conglomerates are chanting 'AI' for their survival, and startups are walking a dangerous tightrope between ambition and reality.
If you're an investor, or someone who wants to work in this industry, you need to be able to see these levels. Don't be fooled by the flashy investment figures or the big names of the founders.
You must distinguish who has a real business model, who is all flash and no substance, and who is sharpening their knives for the future.
The bubble will burst someday. What Satya Nadella said at Davos—"If you guys don't use more of this, we're toast"—wasn't an empty threat. When the current hype cools down, only the Level 4 and 5 companies that are ready to make real money will survive, and the rest will disappear into history.
It's a new beginning once again. I will continue to relentlessly dig into the unseen side of the AI market and the true faces of big tech and startups, and spoon-feed it to you.
In the next video, I'll introduce some hidden champion companies that I'm watching, who are quietly building 'technology that makes money' amidst all this chaos
Watch the Video
This post is based on our YouTube video. Watch it for more details!
Originally published on YouTube: 2/16/2026