Credit – Getty Images/Science Photo Library
WWhen Lawrence Gasman was looking for a PhD topic in the 1970s, computer labs were already buzzing with smart people proposing smart studies in artificial intelligence. “But the problem was we had nothing to run them on,” he says. “The necessary processors simply did not exist.”
It took half a century for computing power to catch up to AI’s potential. Thanks to powerful chips like GPUs from California-based Nvidia, generative artificial intelligence, or generation AI, is revolutionizing the way we work, study and consume entertainment, empowering people to create customized articles, images, videos and to make videos. music in an instant. The technology has spawned a slew of competing consumer apps that offer improved voice recognition, graphic design, and even encryption.
Now AI is poised to get another boost from a radically new form of computing: quantum. “Quantum could potentially do some really remarkable things with AI,” says Gasman, founder of Inside Quantum Technology.
Instead of relying on the binary ‘bits’ of traditional computers (switches referred to as 1s and 0s), quantum uses multivariant ‘qubits’ that exist in a certain percentage of both states simultaneously, similar to a coin spinning in the air . The result is exponentially increased computing power and an improved ability to intuitively mimic natural processes that rarely conform to a binary form.
While the consumer-facing applications of generation AI have made its impact broader and more immediate, quantum has been more focused on industry, meaning several recent milestones have slipped under the radar. However, they could potentially boost the AI revolution.
“Generative AI is one of the best things to happen to quantum computing,” said Raj Hazra, CEO of Colorado-based quantum startup Quantinuum. “And quantum computing is one of the best things that could happen with the rise of generative AI. They are two perfect partners.”
Ultimately, AI relies on the ability to process massive piles of information, and that’s where quantum excels. In December, IBM unveiled its latest processor, called Heron, which features 133 qubits, the company’s best-ever error reduction and the ability to be linked together within its first modular quantum computer, System Two. In addition, IBM unveiled another chip, Condor, which has 1,121 superconducting qubits arranged in a honeycomb pattern. These are advances that mean “we’re now moving into what I like to call ‘quantum utility,’ where quantum is used as a tool,” Jay Gambetta, vice president of IBM Quantum, tells TIME.
Because qubits are incredibly delicate subatomic particles, they don’t always behave the same way. This means that quantum depends both on increasing the total number of qubits to ‘control’ their calculations and on increasing the reliability of each individual. Different technologies used to create a quantum effect prioritize different sides of this equation, making direct comparisons very difficult and reinforcing the secretive nature of the technology..
IBM uses superconducting qubits, which must be cooled to near absolute zero to reduce thermal noise, maintain quantum coherence and minimize interactions with the environment. However, Quantinuum uses alternative ‘trapped-ion’ technology that holds ions (charged atoms) in a vacuum using magnetic fields. This technology does not require cooling, although it is believed to be more difficult to scale up. However, Quantanium claimed in April that it had achieved 99.9% reliability of its qubits.
“The trapped ion approach is miles ahead of everyone else,” says Hazra. Gambetta, in turn, argues that the superconducting quantum has advantages in terms of scale, speed of quantum interactions, and leveraging existing semiconductor and microwave technology to make faster progress.
For impartial observers, the jury is still out, as the array of competing, non-linear metrics makes it impossible to tell who is actually ahead in this race. “They are very different approaches, both of which are promising,” said Scott Likens, Global AI and Innovation Technology Lead for business consultancy PwC. “We still don’t see a clear winner, but it is exciting.”
What Gambetta and Hazra agree on is that quantum has the potential to produce truly amazing hybrid results when combined with AI. “I would like to see quantum before AI and AI before quantum,” says Gambetta. “The synergies between them, and the advancements in technology in general, make a lot of sense.”
Hazra agrees, saying that “generative AI needs the power of quantum computing to make fundamental progress.” For Hazra, the Fourth Industrial Revolution will be led by generative AI, but supported by the power of quantum computing. “The workload of AI and the computing infrastructure of quantum computing are both necessary.”
It’s a vision shared across the Pacific in China, where quantum investment is estimated at around $25 billion, dwarfing the rest of the world. China’s top quantum expert, Prof. Pan Jianwei, has developed a Jiuzhang quantum computer that he claims can perform certain types of AI-related calculations about 180 million times faster than the world’s best supercomputer.
In a paper published last May in the peer-reviewed journal Physical Review Letters, Jiuzhang processed more than 2,000 examples of two widely used AI-related algorithms – Monte Carlo and simulated annealing – which would take the world’s fastest classical supercomputer five years, in less than a second. . In October, Pan unveiled Jiuzhang 3.0, which he claims was 10 quadrillion times faster at solving certain problems than a classical supercomputer.
Jiuzhang uses a third form of quantum technology – light or photons – and Pan is widely hailed as China’s quantum king. Pan, a professor of physics at the University of Science and Technology of China, launched Micius, the world’s first quantum communications satellite, in 2016, which a year later beamed entangled photons between Earth for the world’s first quantum-secure video call.
Micius is considered quantum’s “Sputnik” moment, prompting U.S. policymakers to pour hundreds of millions of dollars into quantum information science through the National Quantum Initiative. Bills such as the Innovation and Competition Act of 2021 have provided $1.5 billion for communications research, including quantum technology. The Biden administration’s proposed 2024 budget includes $25 billion for “emerging technologies,” including AI and quantum. Ultimately, quantum’s awesome computing power will soon make all existing cryptography obsolete, creating a security migraine for governments and businesses around the world.
Quantum’s potential to boost AI also applies to the simmering technology competition among the world’s superpowers. In 2021, the US Department of Commerce added eight Chinese quantum computing organizations to its Entity List, arguing that they “support the military modernization of the People’s Liberation Army” and adopt US technologies to “develop counter-stealth and counter-submarine applications, and the ability to break the encryption.”
These restrictions join a series of measures aimed at China’s AI ambitions, including last year blocking Nvida from selling AI chips to Chinese companies. The question is whether the competition between the world’s two largest economies is hindering overall progress in AI and quantum – or pushing each nation to accelerate these technologies. The answer could have far-reaching consequences.
“AI can accelerate quantum computing, and quantum computing can accelerate AI,” Google CEO Sundar Pichai told the MIT Technology Review in 2019. “And collectively I think this is what we need to solve some of the most persistent problems over time. problems we face, such as climate change.”
Yet both the US and China must overcome the same hurdle: talent. While only a few universities around the world offer quantum physics or mechanics, specific courses on quantum computing are even rarer, let alone expertise on the different specialties within them. “Normally, the most valuable and scarce resource becomes the basis of your competitive advantage,” says Hazra. “And right now in quantum it is people with that knowledge.”
Write to Charlie Campbell at charlie.campbell@time.com.