Microsoft Claims Quantum-Computing Breakthrough—Some Physicists Are Skeptical \ stacker news ~science

pull down to refresh

Microsoft Claims Quantum-Computing Breakthrough—Some Physicists Are Skeptical www.scientificamerican.com/article/microsoft-claims-quantum-computing-breakthrough-but-some-physicists-are/

378 sats \ 22 comments \ @south_korea_ln 17h science

#890721, #890676, and others all talk about this. But be aware that the people who actually have been trying to observe Majorana fermions for years are very skeptical about this smokescreen without actual results to assess the validity of Microsoft's claims.

view all related items

110 sats \ 11 replies \ @SimpleStacker 12h

I guess you are skeptical as well?

I wonder what is the incentive that would drive companies to make premature scientific announcements. Just a temporary stock boost? Some kind of internal political dynamic?

10 sats \ 10 replies \ @south_korea_ln OP 6h

Yeah, pretty skeptical. The fact they had a paper retracted from Nature on the same topic (alleged observation of Majorana fermions, see #891138) warrants extra scrutiny into today's claims.

There is an arms race going on between the big tech companies (IBM, Google, etc) to claim quantum supremacy, so there is a lot of money at stake. Politics play a role too, for sure, as illustrated by @Cje95 eagerness to support these claims (I don't say that in a disparaging way, just an interesting observation to realize that the state also has many reasons to get your country to be the first to win the quantum-race).

To be clear, I am not saying that they have for sure not observed Majorana fermions. Just want to shed additional light from the perspective of someone who talks to people working on this kind of physics daily (a former advisor was involved in the 2018 Nature retraction), there are still many caveats to address before being fully confident about the results.

0 sats \ 9 replies \ @Cje95 6h

I wouldnt say I am supporting the claims yet I am saying though that by partnering with the US National Labs give it more credence. Given that the National Labs are huge basic science centers and they publish the majority of their work publicly its why I just dont see Microsoft being stupid enough to announce they are partnering with the National Labs knowing the results will be public.

That is actually a huge thing my Committee is having to tackle right now... Given the tech arms race and how the US has historically been super transparent with their research results we are having to readjust to the real world know if that makes sense.

0 sats \ 8 replies \ @south_korea_ln OP 6h

It's really interesting to see how you guys are on top of this. It must be very exciting to work on this with your committee now.

US has historically been super transparent with their research

That's a good ideal to work towards. However, my colleague's experience working for a Los Alamos National Lab has been one of having to sign plenty of confidentiality papers keeping him from sharing his results publicly in the way he was used to while working in Europe. But again, I understand the incentives at play, so maybe Europe should also work towards acquiring more IP while working on these topics.

0 sats \ 7 replies \ @Cje95 6h

Working on the Science, Space, and Technology Committee is amazing for me. A ton of the research is basic research and so it isnt the prettiest or news worthy but it is amazing to see how it has led to what it has.

Los Alamos (LANL) is a DOE Nat Lab and a huge NNSA (National Nuclear Security Admin) location as well. Thats going to cause that paperwork but its also one of the reasons why LANL doesnt have the top tier super computers because access is so heavily restricted same as Savannah River.

Really if I went to LANL I dont have to clearance to do much if anything. That is a huge years long effort that I am going to try and start soon.

10 sats \ 3 replies \ @south_korea_ln OP 6h

Thanks for clarifying about the specifics of LANL. Didn't know if/how they differed from other National Labs.

Let me know if you ever need some opinion on some fundamental condensed matter questions. I probably won't be of much help, but I can assure you I have no incentives to overhype any of the topics (other than the one I am trying to get accepted by the Nature editor this week~~).

view all 3 replies

0 sats \ 2 replies \ @SimpleStacker 6h

I couldn't imagine working in DC and staying sane, but working in the Science, Space, and Technology Committees is probably one of the ways to do so. Sounds super cool and interesting!

view all 2 replies

139 sats \ 2 replies \ @kepford 14h

Here we go again. Well... sats are on sale again. Thanks hype machine!

100 sats \ 1 reply \ @south_korea_ln OP 6h

Haha, maybe we should make a prediction market on this topic (@mega_dreamer): will there be conclusive proof in 2025 that Majorana fermions have or have not been observed in the Kouwenhoven experiments, from an external group not affiliated with Microsoft? or something in that sort.

130 sats \ 0 replies \ @mega_dreamer 6h

Sounds interesting, will work on creating this market.

142 sats \ 2 replies \ @Cje95 17h

I will say that partnering with the National Labs separates the claim from most we see. I would assume they will run the chips through the testing at Lawrence Livermore Nat Lab who has the fastest supercomputer in the world in El Capitan and the number 10 with Tuolumne or Oak Ridge Nat Lab that has the number 2... A curveball could be Argonne Nat Lab who has the third but I would assume the first two labs would be best prepared and able to test.

0 sats \ 1 reply \ @south_korea_ln OP 6h

I don't think having a fast classical supercomputer gives one any edge in checking if these claims are true or not.

Of course, having trustworthy institutes supporting this research provides some indication the claims are not completely bogus, but appealing to authority has not always shown to be a reliable metric to confirm extraordinary claims. I hope they will soon share more details about their experiments so that other groups can try to reproduce the physics and science can advance in concert rather than hidden behind IP from one specific company. Of course, I understand the incentives at play, so it's normal that Microsoft is trying to gain some advantage here, but until they or others can provide more evidence, I'll be remaining skeptical.

As a physicist, I of course truly hope they are right. Majorana fermions have been a theoretical object for too long. We are really living in exciting times for condensed matter physics.

0 sats \ 0 replies \ @Cje95 6h

110% quatum will be an entirely different computer structure but given we have to learn about it to build it the only way to do the calculations and more importantly the simulations esp. when it comes to materials science is going to be with El Capitan and Frontier.

Odds are with the National Lab backlog its going to be roughly a year. Everyone wants computation and the only way they could get further ahead in line I would assume would be to trade in Azure Compute power. The US DOE have launched 5 NQISRC that are specializing in the QIS research https://nqisrc.org/

116 sats \ 1 reply \ @freetx 16h

10 sats \ 0 replies \ @south_korea_ln OP 6h

The authors are indeed very careful in the actual paper. Media isn't always as nuanced when reporting on technical papers...

0 sats \ 0 replies \ @kepford 6h

People just wanna believe... as long as that's the case these silicon valley hype machines will keep it up.

0 sats \ 0 replies \ @jerrybature 14h

"Claims"