Most of the news I’m hearing today about the current wormhole publicity stunt is that physicists who could do something about it are instead blaming any problem on journalists and defending the stunt as some sort of progress forward.
I’ve been wondering what the future for this kind of thing looks like, got a partial answer by looking at this presentation today by the director of Fermilab. On page 67 she explains
Future experiments with better QC and with QCs connected through quantum networks, such as those under development at Fermilab, could provide better insight through better resolution and adding non-trivial spatial separation of the two systems.
So, next generation wormhole publicity stunts will involve, beyond going from 9 qubits to more, putting two quantum computers in two places and connecting them by a quantum network. The press reports will explain that physicists not only created a wormhole on a chip, but created a wormhole connecting two different labs.
I started looking for more information about these next-generation wormhole publicity stunts, and found instead something I hadn’t been aware of, an older such stunt, described in Towards Quantum Gravity in the Lab on Quantum Processors, which got attention last spring not at Quanta, but in the much lower profile Discover Magazine, where one reads:
The team developed quantum software that could reproduce wormhole inspired teleportation on both quantum computers and then characterized the results. “We have designed and carried out “wormhole-inspired” many-body teleportation experiments on IBM and Quantinuum quantum processors and we observe a signal consistent with the predictions,” say Shapoval and co.
One reason for the lack of significant attention to this publicity stunt as opposed to the current one surely is the decision of the authors to claim not “wormhole teleportation” but “wormhole-inspired teleportation”.
The past is the past, but it looks like the field of quantum gravity research is from now on going to be dominated by these wormhole publicity stunts, using more qubits and more quantum computers. This kind of research project is nearly ideal: you can get lots of funding from conventional sources like DOE, or even better, funding from and access to equipment at large tech companies like Google and IBM. You can convince the director of your lab or institute that you’re doing research of significance comparable to the discovery and testing of general relativity 100 years ago and your work will be vindicated by cover stories in Nature and all over the rest of the media.
Back in 1996, in The End of Science, John Horgan worried that this kind of science would end up in a “speculative post-empirical mode”, and quantum gravity theorists have for years now worried about accusations of not being connected to experiment. The solution to this problem is now clear: no one will take your wormholes seriously if they’re just on paper, so the thing to do is to get them realized in an algorithm that you run on the most twenty-first century experimental hardware available, a quantum computer in a tech company lab.
Update: There was a hoax comment posted here last night, supposedly from Natalie Wolchover, which had me fooled for a while. Whoever was doing this seems to have been making excellent use of ChatGPT, together with manipulating other aspects of how the comment was posted that helped fool me.
In my opinion, worse than the overhyped wormhole was the overhyped 2021 Fermilab presentation of muon g-2 results as a 4.6σ anomaly, done ignoring one lattice computation that was in agreement with the measured value. Now ≈three lattice computations agree, so the g-2 anomaly is likely over, but the Fermilab director (in the recent presentation you link) still writes: “Muon g-2: Complete data production, analysis, theory to achieve 5σ”.
Furthermore, I don’t understand the claim, in the same presentation: “U.S. is universally acknowledged as the world leader in neutrino science for decades to come”. Main results come from Japan, and their future HyperKamiokande can be faster/better than DUNE.
Thanks for the shout-out Peter. The phrase with which I describe physics involving strings, multiverses and other untestable stuff is “ironic science,” which should be considered more akin to philosophy or science fiction than science. This wormhole story seems like a fascinating example of it. The details of the “experiment” described in Nature are obscure to me, but here is how I understand it: The authors postulate that entanglement is carried out via wormholes. Then they perform quantum computations, which supposedly exploit entanglement. To emphasize the wormhole angle, they simulate a wormhole with their computations, and on that basis they claim to have actually created a wormhole, because entanglement entails wormholes. Is that right? If so, it’s a spectacular case of question begging, in the sense of assuming what you set out to prove. And by the way, I am not an AI bot, this is the real, fully sentient John Horgan.
I still think the best way to explain the problem is that the physics of a quantum computer has nothing to do with gravity. You can’t do an experiment that has nothing at all to do with certain physical degrees of freedom, and claim that you are studying those other degrees of freedom experimentally.
What’s going on here is people muddying the waters because it sounds good, bringing in highly speculative and ill-defined theories (ER=EPR) about relations between the quantum computer physics and geometry of some space (which has nothing to do with our space-time), together with mixing up a simulation of a model and the physics being simulated itself.
Relevant to your writings, what I wanted to point out is that many of the people involved in this have been getting beaten over the head for years about embodying the “post-empirical” part of the danger you were pointing to. What’s going on here is basically a fraudulent attempt to claim that they have solved this problem, have found a way to experimentally investigate quantum gravity on a quantum computer, despite the fact that there is no relevant interaction of the experiment with gravitational degrees of freedom.
Complicating the discussion is the fact that there really are toy models of quantum gravity for which it’s currently intractable to calculate what they predict, and there really is some hope that a 100- or 200-qubit quantum computer could help with that. Indeed, because of the extreme simplicity of the toy models (typically just an array of qubits subject to some time-independent Hamiltonian), this might even be one of the first applications of quantum computers to become technically feasible (for some definition of “application”).
Of course, as shouldn’t need to be said but often does, even if this program succeeds, the question will remain of what if anything those toy models say about the real universe. Calculating the predictions of the toy models can only ever be one part of an answer to that question.
As you rightfully pointed out the authors, at least the last author, love public attention. For instance, this experiment got so much media attention and it was covered everywhere as if they invented teleportation for the first time ever (look at the altmetric score) https://journals.aps.org/prxquantum/abstract/10.1103/PRXQuantum.1.020317
However, teleportation experiments at telecom wavelengths have been around for decades, for instance: https://www.nature.com/articles/nature01376
What they did last year has improved the state fidelities etc. but the result is far from what is made out to be. For instance, they also speculate that their work would open the path towards integrating these teleportation systems with quantum memories, but even then that was done some 8 years ago: https://www.nature.com/articles/nphoton.2014.215
John Horgan here repeats their mantra “entanglement entails wormholes.” This is simply false. The EPR paper in 1935 had nothing whatever to do with a wormhole, it was a Minkowski spacetime calculation. Entanglement happens quite happily in flat spacetime, as has been experimentally proved. It could perhaps occur in traversable wormholes IF any were to exist in the real universe, which is highly doubtful. In any case the claimed equivalence represented in an “equation” of the form ER = EPR is not true. Don’t fall for the propaganda!
Thanks guys, really helpful. Just one question: Are the particles in the 9-qubit Google computer, when they carry out a computation, entangled with each other? Is that why the Nature authors can claim that they have created an actual wormhole, and not just a simulation, because according to the ER-EPR conjecture entanglement entails a wormhole?
My problem with ER=EPR is not that it is untrue, but that it’s meaningless, as about “not even wrong” as “not even wrong” gets. It’s basically a conjecture that “maybe there is a duality between conventional QM and some unknown theory”, but doesn’t tell you what the unknown theory is (it’s not conventional GR). What’s going on here is a study of some toy model infinitely far from anything physical, but with some ER=EPR hoped for properties. The authors do completely standard calculations in the toy model on a classical computer. Then, they do again the simplest such one possible on a quantum computer, purely as a (very successful) publicity stunt.
The qubits in any quantum computer are entangled, that’s the basis on which it works. It’s also true that entanglement is everywhere in any quantum system. What’s hard to do is not to create entanglement but to isolate the qubits so they don’t entangle with the environment. See my comment to Ellis about ER=EPR: if you believe it, a quantum computer is a collections of wormholes (where the word “wormhole” is meaningless, since you have no idea what the dual theory is in which they supposedly exist).
The “wormhole” they are claiming is a name for something in the 2d gravity toy model dual to an SYK toy model. I have to advise you to not even try and understand exactly what this means and what this “wormhole” is, if you think you understand it I’ll bet you’re mistaken…. Such things have no relationship at all to physical reality, they’re objects in toy models. Also nothing to do with quantum computers, which come into play here purely as a way to fool the rubes (such as the IAS and FNAL directors).
George Ellis, don’t worry, I haven’t fallen for the ER-EPR “propaganda.” I’m just trying to understand how the Nature authors could possibly claim that their crude little simulation contains an actual wormhole. The only justification–and again, I don’t believe it, I’m just trying to figure out what the authors believe–is that they assume the ER-EPR conjecture is correct, which implies that a real wormhole might underpin their quantum-computer simulation of a wormhole. Peter Woit, thanks for your clarifications, and cautions. I don’t know if I’m going to write about this strange episode, but I enjoy trying to make sense of it, for my own sake if no one else’s. Physics just gets weirder and weirder.
By the way, even if one is a “true believer” in ER=EPR (as the authors of the paper supposedly are), this particular result is nowhere near the landmark for wormholes, nor a first one created. A quick google search says this about a world record for controlled entanglement:
“A team led by physicists Prof. Harald Weinfurter from LMU and Prof. Christoph Becher from Saarland University have now coupled two atomic quantum memories over a 33-kilometer-long fiber optic connection. This is the longest distance so far that anyone has ever managed entanglement via a telecom fiber. (Jul 7, 2022)”
That must be one hell of a wormhole, and moreover, inside a fiber optic cable! The media should have covered *that* wormhole, not this tiny one inside Google’s processor… 🙂
However you frame it, the media outlets have been had. 🙂
Funnily, assuming one would take ER = EPR seriously, one would believe that the ‘real’ entanglement in the quantum computer is dual to some wormhole in some spacetime, whereas the hypothesized 2D wormhole is dual to entanglement in the simulated model.
So I guess an equally valid headline would have been: “Physicists simulate entanglement using Google’s wormholes”
Quanta might have been better off putting the qualifier in front of the ‘Physicists’ rather than the ‘wormhole’ when amending the title.
I don’t think that’s how wormholes work. A wormhole takes matter directly from the input to the output without going through the physical space in between.
Similarly, quantum teleportation takes the input quantum state directly to the output quantum state without this state being present anywhere in the physical space in between.
So clearly, by Occam’s razor, teleportation must work by transporting the quantum state through a wormhole. 😉