The Palantir you see in the media is not the real Palantir.
Peter Thiel, co-founder of Palantir and godfather of Silicon Valley
For most Chinese investors, "Palantir" is not a familiar name compared to tech giants like Apple, Google, Tesla, Amazon, and Facebook.
However, in today's European and American markets, Palantir is even more popular than these star companies.
Palantir's latest market value is about $437 billion, equivalent to about 31.3 trillion RMB.
Although its market value still lags behind the "Seven Sisters of Silicon Valley", the uniqueness of its value is enough to make it stand out.
From the perspective of many media, Palantir is regarded as an AI technology company. While this is not wrong, it is not enough to get to the essence of its business model -
As a technology company with half of its revenue from the consumer side (C-end) and half from the government side (G-end), it is an atypical commercial enterprise and should be called the "Skynet version of the military-industrial complex".
Palantir really broke through the circle on the Russia-Ukraine battlefield. Beneath the surface of "making a fortune" from the war, its ability to influence sovereign-level power is actually the key to its superiority over other technology companies.
As a key promoter and core founder of Palantir, Peter Thiel, the godfather of Silicon Valley, has more influence in the U.S. political field than any other Silicon Valley entrepreneur. He is the mentor of J.D. Vance, and even Elon Musk, who is keen on power games, can only look up to him.
It is still an extremely difficult task to fully describe and analyze Palantir today.
Therefore, the Palantir we see in the media is far from the real one.
Palantir's original idea came from the Paypal Mafia: financial anti-fraud.
But after 2003, it received investment from In-Q-Tel (CIA's venture capital) for counterterrorism intelligence.
Before Palantir, intelligence work was very primitive.
It was divided into the following types:
1. Large-scale invasive methods. For example, to prevent bombers, everyone was required to undergo a 20-minute strip search at the airport. Peter Thiel said that the bomber actually succeeded because he killed everyone's time.
For example, due to loan fraud, no one was allowed to take out loans.
2. The traditional intelligence model, a hierarchical system. Due to the self-expansion of bureaucratic institutions, the tendency to spend more money and do less work, it became more difficult and costly to collect intelligence. For example, the discovery of Dogecoin and the U.S. Department of Defense's $1 trillion budget.
3. The black-box group model: Hiring algorithm experts to bypass supervision and engage in black market transactions, which led to a lot of corruption. Moreover, the algorithms were static and easily became ineffective the next time, so the data could not be reused and scaled up.
For example, the war in Afghanistan had no supervision, no results, and cost trillions of dollars.
4. U.S. institutions still adopted the cost-plus procurement model. Since they were all old-school enterprises, ineffective software and overlapping and redundant hardware were inevitable results. The 1980s-style interface was very common.
The products used in the Afghanistan war were regarded as junk by soldiers, inferior to Palantir.
Palantir's technology adheres to three major principles:
1. Technology must be connected to real-world actions; otherwise, data and models are meaningless.
2. Organizations must accept complexity to survive.
3. Technology should simplify complexity as much as possible.
The first approach is to do more with more: Spend more money to invade people's privacy on a large scale.
The second approach is to do less with less: Simply cancel the loan business to avoid fraud; or simply ban people from flying to prevent terrorists.
Palantir does more with less:
It reorganizes intelligence with software, enabling existing intelligence data sets to be efficiently coordinated to help managers make accurate decisions in a short time.
Ontology: Decision-making is the essence of an organization. To make judgments and take actions based on data, data such as vehicles, personnel, phone numbers, and travel routes need to be organized into a system and aggregated on an intelligent platform.
The commercialization process: Since most large institutions control huge budgets, and the U.S. government spends nearly 40% of GDP, about $12 trillion, it is a very valuable but extremely difficult market to enter. Marc Andreessen said: The United States is not a free market. It's easy to open a restaurant, but most institutions are monopolies.
For Palantir, a Silicon Valley startup with only a few dozen people aiming to serve large institutions like the U.S. government, the entry threshold is unimaginable no matter how good the product is:
1. If serving institutions, most startups will experience a very long period of zero revenue. Few people have the patience for this. Most companies serve the consumer side, like Google and Facebook. In the past 30 years, only three startups have truly served the defense industry: Elon Musk, Peter Thiel, and Anduril, which he invested in.
Peter Thiel has provided extremely long-term support for this vision.
2. In addition to tangible barriers such as certifications, the intangible barriers are also extremely high. Positive feedback from frontline agents and soldiers cannot promote procurement, and administrative orders from top generals are not enough to change the situation. Bureaucratic institutions can delay indefinitely and suffocate startups in the paperwork process for various reasons.
Against the advice of consultants, Palantir sued the Department of the Army and won a great victory, completely entering the largest market of the U.S. Department of Defense.
The two-way push method: Applying continuous pressure from the top down and from the bottom up until the bureaucratic institutions break down. Hiring a large number of lobbying firms to promote legislation, and engineers working directly with users, with extremely high deployment efficiency.
3. In 2016, Peter Thiel supported Donald Trump: Kingmakers can get their due rewards.
4. What really matters is the cultural change. The Russia-Ukraine war in 2022 was a real turning point. Before that, the Silicon Valley culture was to stay away from the military industry. After that, it was too late. When people realized that U.S. military hegemony was no longer enough to win the war in Ukraine, the culture of the tech community changed, and the whole society shifted towards accelerating U.S. technological dominance.
Whether it's the AI explosion or the rapid growth of Palantir's revenue, it is extremely important compared to the technological changes of its competitors. It is a watershed in social culture.
In the 2024 election, the establishment of Dogecoin actually helped the entire U.S. government system become more coordinated, with system data connected in one line. After the data was connected, Palantir's revenue grew even faster.
In the G-end market, Palantir, as a software service provider, actually has no competitors. You can think of it as SpaceX after NASA. It seems that everyone can build rockets, but after SpaceX broke through the competitive barriers, it took the lead and grew at an amazing speed.
The characteristics of the software industry itself are decisive: Users have excellent stickiness, and Palantir's efficiency is far higher than that of any competitor. Most competitors need to wait for licenses, which will give Palantir a 5 - 7-year time window, during which Palantir will gain even deeper competitive advantages.
Elon Musk said: The U.S. G-end service is like a group of cows on the grassland. A few people are milking them, but there is a 7-foot-high fence outside. Once you open the fence, it's all about competition.
But who would have the patience to pry open the fence for 15 years? Peter Thiel still remains silent.
This article is from the WeChat official account "Jinduan Research Institute". The author is Mu Zhi. It is published by 36Kr with permission.