How bad does the world have to get for Palantir to be happy? In the sunny American technology sector, the software company stands out for its ominous warnings about global instability. Artificial intelligence could be the crisis it has been looking for.
These are high times for cynics. AI is touted as dystopia in the making. In meetings with technology companies, I keep hearing the phrase “Oppenheimer moment” – a reference to Robert Oppenheimer, the physicist who led the creation of the atomic bomb.
It’s still not clear exactly how generative AI could destroy us, or what fortunes it will create. But all this talk of great power is turning out to be extremely useful for the value of a handful of companies. This week, AI chipmaker Nvidia’s market cap briefly hit the trillion dollar mark. AI startups like Character.ai and Anthropic continue to raise money even as funding dries up elsewhere. The share price of Palantir has more than doubled in five months.
This week, Palantir made its new AI platform widely available. The tool can generate conversation responses using the kind of large language models, or LLMs, that power chatbots like ChatGPT. Because it’s based on customers’ specific data, it should prevent hallucinations – the false answers that plague other chatbots. A demo on YouTube shows how it might work on the battlefield, helps identify an enemy tank and offers suggestions for ways to target it. The company says Ukrainian troops are already using some of the initial features.
Palantir isn’t the only software company racing to show how generative AI can be used for something more productive than college essay writing. IBM also announced a new AI platform called Watsonx. But IBM’s share price has fallen this year.
Palantir seems to outperform most real world applications. “You need a core set of technologies that will allow you to bring these LLMs to your enterprise — to work on your data,” said Shyam Sankar, chief technology officer. “And then you need a really strong governance control layer that allows you to develop trust in AI.”
It helps that AI-style riddles and existential threats are Palantir’s trading stock. It’s hard to think of a company that talks more about disasters. Last year it warned that the world was underestimating the threat of a nuclear attack, pegging it at about 20 to 30 percent. Silicon Valley co-founder and investor Peter Thiel is known for making his own eerie statements about global destruction. In 2008, he described what he called the re-emergence of an apocalyptic dimension in the modern world. While the 20th century had been “great and terrible,” he wrote, the 21st century promised to be more of both. Fears around AI fit perfectly into that worldview.
What does Palantir actually do? The $31 billion company has been described as the technical answer to management consulting. It was created in the aftermath of the 9/11 attacks to build software that could be used by intelligence agencies to counter terrorism before being expanded to other government departments and businesses. The software sifts through data, collects information, finds patterns and presents them in ways intended to be useful and easy to understand. It says its services have helped BP cut production costs by about 60 percent. It is also the frontrunner for a new seven-year NHS contract worth up to £480 million.
However, in recent years, Palantir has also been a somewhat cautionary tale about what can happen when a company known for working in the shadows finds itself in the spotlight. Named after the dark, prescient crystal balls in the Lord of the Rings trilogy, it has long made a virtue of secrecy. The idea of an eerily omniscient tech company was irresistible to the media. In 2018, a Bloomberg article claimed the company knew “everything about you.” A few years later, The New York Times asked if it saw “too much.”
Chief executive Alex Karp has done a good job of keeping Palantir’s eccentricities at the forefront. He is known to love German philosophy and has described his desire to work with creative, “crazy” people. When I was in the company’s Denver office, I noticed a display case with a boring office suit with a sign saying “Emergency Break”.
Some of the glamor disappeared when the company went public at the end of 2020. Suddenly it was subjected to the banalities of quarterly earnings and investors wanting profits that meet generally accepted accounting principles. The inconvenient truth is that Palantir has never reported an annual profit in 20 years. This is expected to be the first year it breaks that spell.
Interest in AI has given the company back its mystique. Add renewed profitability and the result is stock price turmoil. It helps that the zeitgeist is catching up with Palantir’s way of thinking. If AI really sends us to oblivion, don’t expect Palantir to be surprised.
elaine.moore@ft.com
