Wired: ex-director of Google will build the perfect military machine with artificial intelligenceThe former CEO of Google wants to introduce artificial intelligence into the US armed forces, Wired reports.
Eric Schmidt is going to equip the American army with a machine learning system that will allow for virtual side-testing and testing of equipment using computer models.
Will KnightThe former CEO of Google set out to equip the US armed forces with advanced artificial intelligence to counter China.
Will our world be safer from this?Before going to the battlefield, expensive military equipment like the same tanks is thoroughly tested.
A startup called Istari, the brainchild of former Google CEO and Alphabet CEO Eric Schmidt, believes that at least part of this work will be more convenient to do in the metaverse.
The Istari program uses machine learning for virtual assembly and testing of combat vehicles from computer models of individual components that, as a rule, do not intersect on digital drawing boards — for example, chassis and engines. It will seem hackneyed, but Schmidt says it will bring much-needed technological innovations to American military engineering. "The Istari team makes models and simulations more convenient to use in the image and likeness of the Internet," he says. "In the future, this will make it possible to create flexible software for physical systems — it's terribly interesting."
The company embodies Schmidt's unique position as a kind of link between the technology industry and the Pentagon. Virtual copies (also known as "digital doubles") are already being used in production and will help the Pentagon accelerate the development of new equipment. Istari is just one of the stages of an extensive project during which Schmidt hopes to introduce Silicon Valley technologies and ideas into the US armed forces.
It all started with the shock that Schmidt experienced in 2016, when he first saw firsthand the level of Pentagon technology. At that time, he was still the head of Alphabet, but he had already accepted the invitation of the Obama-era Defense Secretary Ashton Carter to head the new Council on Defense Innovations with an eye to the modernization of the armed forces. Schmidt's new position began with an introductory tour of laboratories and bases in the United States.
"I had to travel around a lot with Eric and see at what level commercial technology is," recalls Will Roper, then Assistant Secretary of the Air Force for Technology, later co—founder and CEO of Istari. — It turned out that the Ministry of Defense develops software in the same way as it was done in the 1970s and 80s.x". He turned out to be just one of many Pentagon bosses who were impressed by Schmidt's "diagnosis" and his willingness to at least try to solve these problems.
Schmidt became CEO of Google in 2001, when the search engine had several hundred employees and barely made a profit. He left Alphabet in 2017, having built a large and highly profitable company with an extensive portfolio of advanced projects, which included artificial intelligence, self-driving cars and quantum computers.
Today Schmidt saw a new opportunity for a technological leap for the sake of domination — this time for the US government in the rivalry of world powers. With his understanding of the Pentagon's technological goals, he had a unique opportunity to help achieve them. But his industry connections raise the question of how America can reconcile the interests of the government and the private sector. U.S. military power has long relied on advanced technology, but some fear that military AI is fraught with new risks.
The people are good, the system is unimportant
Speaking on Zoom from his office in New York, Schmidt outlined the grandiose plans of the advanced Ministry of Defense, which will deftly use technology from companies like Istari. In his cheerful orange sweater of the finest wool, he casually talks about a large-scale "reset" of the US armed forces.
"Let's say we're going to build the best combat system in the world," Schmidt says, outlining a radical restructuring of the most powerful army on earth. "We will start by creating a technology company." He explains his interpretation of the "Internet of deadly things": "There will be a large number of inexpensive devices — extremely mobile and easily accessible. These devices (or even drones) will carry sensors and weapons and will unite into a single network."
According to Schmidt, the problem of today's Pentagon is not money, talent or determination. He described the modern US armed forces as follows: "the people are good, the system is unimportant." It is a product of a bygone era, in which large, slow and expensive projects dominated (for example, the same aircraft carriers), and bureaucracy prevented decisions from being made promptly. Independent research and congressional hearings have shown that it takes years to choose and purchase software from the Ministry of Defense, and it may become obsolete by the time it is installed. Schmidt believes that this is a huge problem for the United States, because digitalization, software and networking can revolutionize the armed forces.
Schmidt believes that the path to progress for the Pentagon was indicated by Ukraine's response to the Russian military campaign. The Ukrainian armed forces managed to resist the superior power largely due to operational actions and the introduction of technologies from the private sector — in particular, arming commercial drones, repurposing failed communication systems right on the battlefield, as well as setting up three-dimensional printing of spare parts and developing useful software for military tasks such as payroll in months, and not in years.
As an illustration of the difficulties faced by the US armed forces, Schmidt offers another thought experiment. "Imagine that you and I want to solve the Ukrainian problem, and the Defense Ministry allocated us $ 100 million and gave us six months," he says. "And six months later, someone really came up with a new device, tool or a whole method that will help Ukrainians win." Is the problem solved? It wasn't there! "Everything I've just told you is illegal: procurement rules prohibit the Pentagon from handing out money without a thorough check, and it takes a disproportionate amount of time," Schmidt explains.
New weapons
According to Schmidt, the most acute technical problem of the Pentagon is in the field of artificial intelligence (AI). "From time to time, a new weapon or a new technology appears that radically changes the situation," he says. — In the 1930s, Einstein wrote a letter to Roosevelt and said that there was a technology that would change the war - called nuclear weapons. And so it turned out. Today I would single out AI-based autonomy and decentralized distributed systems."
Not without Schmidt's help, a similar opinion has taken root in the Pentagon over the past decade: its leaders agree that AI will revolutionize military technology, intelligence gathering and software. In the early 2010s, the Pentagon began to look closely at technologies that would help maintain an advantage over the growing Chinese army. The Council for Defense Sciences (the highest advisory body on technology) concluded that the future of military confrontation and conflicts lies precisely with AI-based autonomy.
But AI technology is developing mainly in the private sector. The best tools that can prove to be a decisive help for the military — for example, algorithms for identifying enemy equipment or specific people on video, or tools for studying "superhuman" strategies — are created either by technology giants like Google, Amazon and Apple, or thematic startups.
"The big problem that the US military will face in the future is how to adapt commercial technologies to the demands of the army faster than competitors," says Paul Scharre, vice president of the Center for a New American Security and author of the book "Four Battlefields: Military Power in the Era of Artificial Intelligence." In it, Sharre notes that the Pentagon's share of global spending on research and development has fallen from 36% in 1960 to just 4% today.
The US Department of Defense mainly works with the private sector through large defense contractors specializing in the creation of labor-intensive and expensive equipment, rather than flexible software. The Pentagon's contracts with tech giants of the caliber of Amazon, Apple and Microsoft are gaining momentum, but at the same time they cause complaints. The analysis of drone footage using Google AI in the framework of the Maven project ("Expert") caused a protest by employees, and the company terminated the contract. Since then, Google has expanded its work in the military sphere — although its rules still exclude certain categories of weapons.
Sharre believes that people like Schmidt with their influence within the industry will be able to bridge this gap. Under the threat of change, tech giants sometimes successfully reinvent themselves. And industry representatives can help the Pentagon to fight bureaucracy and become more attractive as a partner for startups — the most important generator of new ideas. "Otherwise, we are still trying to build an army of the XXI century with the bureaucracy of the XX," he complains.
U-turn towards China
Schmidt concluded that the technology industry is obliged to help the Pentagon, and the government, in turn, Silicon Valley. In 2019, he headed the US National Security Commission on Artificial Intelligence, which Congress established to study the impact of this technology on the security and competitiveness of the country. Her final report for 2021 focuses on the AI rivalry between the US and China. In particular, there is a warning that this technology can contribute to strengthening authoritarian values. In order to preserve the purity of the AI source in the United States, the government was urged to work more actively with the private sector and provide funding, data and computing power to both public and private projects.
At an event last fall, Schmidt admitted that working for the commission had changed his life and opened his eyes to the Chinese threat. "We are facing an extremely serious challenge: our opponent is focused on his goals and has a good idea of what he is doing," he said. Since then, the commission has been disbanded, but Schmidt himself now works in a similar department, studying the consequences of advances in biotechnology. And he also funded a new independent analytical center called the Project of Special Competitive Research to complete the recommendations of the commission. The project not only deals with artificial intelligence technologies, but is also modeled after the Cold War-era anti-Russian initiative created by Nelson Rockefeller under the leadership of Henry Kissinger.
Last year, the project released a series of reports in which it called on the government to finance areas crucial for the further growth and competitiveness of the United States — including nuclear fusion, quantum computing and communications, as well as genome editing. The reports were published against the background of growing support for large-scale government intervention in the field of technology. The law on chips adopted last year is dictated by concerns about China, and it will allocate 280 billion dollars for research and production of semiconductor devices in the United States.
However, it will not be easy to establish closer cooperation between the government and industry. In 2017, when Schmidt was still working at the Defense Innovation Council, one official expressed concern, later withdrawn, about possible conflicts of interest — both personally to him and to other Silicon Valley executives. Schmidt still owns about $5 billion worth of Alphabet shares, is listed among the investors of the start-up military contractor Rebellion Defense, and through various investment firms has connections with other companies cooperating with the government.
"It's hard to name another CEO with comparable influence in the field of national security technology," says Jack Poulson, who tracks relationships between individuals, corporations, nonprofit organizations and governments through the nonprofit Tech Inquiry. He notes that Schmidt participates in the work of several technology companies at once, and all of them work in areas that, according to the Special Competitive Research Project, deserve expanded government funding.
Perhaps Schmidt's work demonstrates not only the value of cooperation between government and the private sector, but also the need for greater transparency and accountability as it expands. Schmidt's spokeswoman Melissa Stavenhagen says he always discloses all the necessary information in full. "With his experience working under democratic and republican administrations, he is well aware of the importance of this problem," says Stavenhagen.
During our conversation on Zoom, it seemed from time to time that Schmidt was disappointed by the systemic crisis in the US government and its approach to technology. In 2016, when he first joined the Pentagon, he did not expect that he would find a new vocation. "I decided that I would do this for about a year to help," he recalls. However, in the end it became his second career. Whatever success the Pentagon has achieved in the development of artificial intelligence — and no matter how his dreams affect the world around him — Schmidt will surely play a key role in this.
