But plenty of questions still remain about the technology’s basic usefulness as anything but a pathbreaking physics and math experiment, meaning companies working on quantum need to periodically update the public on exactly what it is they’re doing.
IBM is one of the biggest corporations in the field, and this morning it rolled out the red carpet for its new Quantum System Two — a modular quantum computer powered by a new IBM-made chip that the company says is far, far more stable than anything seen before in a quantum system. (Stability, of course, is paramount for a technology that rapidly loses its utility when it’s jostled, or warmed, or even used at all.)
The company’s new chip is called the “Heron,” the latest in a long line of bird-themed nicknames for their quantum line. I met with Dario Gil, IBM senior vice president and director of research, and Jay Gambetta, the company’s vice president of quantum computing, several weeks ago to see the System Two in person at their laboratory in Yorktown Heights, New York. There, they described in detail how the Heron chip is the culmination of years of effort toward improving “error correction,” or the process of fighting decoherence.
Gambetta described its development as a long-term project aimed, quite self-consciously, at pushing the field forward and cementing IBM’s status at its forefront. (For a detailed technical explanation of IBM’s strides with the Heron, see this Ars Technica explainer and conversation with Gambetta.)
One major problem in the quantum-computing ecosystem is that the miraculous applications that some promise — the end of cryptography, instant development of miracle drugs, climate change reversal — are still far, far in the future.
That means companies in the private sector that want to make the case for quantum investment while keeping expectations in check have to find a delicate balance between the technology’s current-day uses and how they might realistically build a path to a quantum-powered tomorrow.
Right now that means using quantum computers mostly for research, like a recent quantum experiment at the University of Washington that demonstrated how quantum could potentially be uniquely useful in answering questions about how matter interacts at the atomic level.
“You’ve probably heard of quantum supremacy, or quantum advantage,” said Jay Gambetta, IBM’s vice president of quantum computing, during our tour of the Yorktown Heights laboratory, referring to the idea that quantum computers will someday have uses where they clearly surpass the power of classical computers. “All of this gets complicated, comparing apples and oranges. We said, let’s just simplify it and focus on scientists and what they’re able to explore [with quantum computers].”
Sergio Gago Huerta, head of quantum at Moody’s and author of the Quantum Pirates newsletter, is one of the industry watchers who are optimistic about IBM’s proposed “road map” for more usable quantum computers — which culminates in 2033 with its proposal for the world’s first quantum supercomputer, nicknamed “Blue Jay.” Still, he’s looking for more detail on how it plans to get there.
“In general, I think it is a great step towards achieving quantum advantage in the future,” Gago Huerta said. “But we believe that many of the computational problems we face may be difficult to encode in parallelized systems like those… The updated roadmap gives us a fantastic baseline to work with, although I wished it had more information.”
For his part, Gil described the discovery and application of quantum computing as “like gravity,” an inevitable outcome where the only salient question is who might perfect it first — something he says the U.S. has the potential to do with the appropriate support from Washington.
“We need a moment of vision and leadership, for someone to say, we’re going to do it, and I want it by this day, and with this data, and we’re spending this many dollars,” Gil said. “We’re in a moment in quantum where that level of ambition is needed, to say that we’re going to do it and no one company or one institution will be doing it alone.”
Some tech watchdogs are worried a program meant to put AI experts in congressional offices could be a front for industry lobbying.
POLITICO’s Brendan Bordelon reported yesterday on the “rapid response cohort” of AI experts placed in Washington by the American Association for the Advancement of Science, a nonprofit funded by tech giants including Microsoft, OpenAI and Google. Those companies say their involvement is hands-off and the fellows are chosen solely by the nonprofit, but watchdogs are worried that the funding means the fellows’ advice to lawmakers might be tantamount to advocacy.
Sarah Myers West, former senior adviser on AI policy at the Federal Trade Commission and managing director at the AI Now Institute, emphasized to Brendan that “Tech firms hold an unprecedented amount of financial capital and have a long track record of using it to attempt to tilt the playing field in their favor.” Craig Mundie, an adviser to OpenAI and Microsoft who largely conceived and stood up the project, defended it, saying “When you’re at the bleeding edge of anything and you want to make good policy, you want to talk to the people who actually are doing the stuff — not people that are opining on what people might be doing.”
The United Kingdom’s House of Lords is warning governments to proceed with caution when it comes to using autonomous weapons.
In a report published Friday, the lawmakers make a case that world governments’ attempts to establish democratic oversight of AI should pay special attention when the technology is weaponized.
“The AI Safety Summit was a welcome initiative, but it did not cover defence,” they write. Quoting that summit’s Bletchley Declaration, they continue “The Government must include AI in [automated weapons systems] in its proclaimed desire to ‘work together in an inclusive manner to ensure human-centric, trustworthy and responsible AI that is safe’ and to support ‘the good of all through existing international fora and other relevant initiatives.’”
The authors emphasize the nuclear command and control system as a key part of this effort, and recommend further clarifying all procedures in defense systems to mandate human control whenever AI is used.
Stay in touch with the whole team: Ben Schreckinger ([email protected]); Derek Robertson ([email protected]); Mohar Chatterjee ([email protected]); Steve Heuser ([email protected]); Nate Robson ([email protected]) and Daniella Cheslow ([email protected]).
If you’ve had this newsletter forwarded to you, you can sign up and read our mission statement at the links provided.