Microsoft's Big Win in Quantum Computing Was an 'Error' After All (wired.com) 34
In a 2018 paper, researchers said they found evidence of an elusive theorized particle. A closer look now suggests otherwise. From a report: In March 2018, Dutch physicist and Microsoft employee Leo Kouwenhoven published headline-grabbing new evidence that he had observed an elusive particle called a Majorana fermion. Microsoft hoped to harness Majorana particles to build a quantum computer, which promises unprecedented power by tapping quirky physics. Rivals IBM and Google had already built impressive prototypes using more established technology. Kouwenhoven's discovery buoyed Microsoft's chance to catch up. The company's director of quantum computing business development, Julie Love, told the BBC that Microsoft would have a commercial quantum computer "within five years." Three years later, Microsoft's 2018 physics fillip has fizzled. Late last month, Kouwenhoven and his 21 coauthors released a new paper including more data from their experiments. It concludes that they did not find the prized particle after all. An attached note from the authors said the original paper, in the prestigious journal Nature, would be retracted, citing "technical errors."
Two physicists in the field say extra data Kouwenhoven's group provided them after they questioned the 2018 results shows the team had originally excluded data points that undermined its news-making claims. "I don't know for sure what was in their heads," says Sergey Frolov, a professor at the University of Pittsburgh, "but they skipped some data that contradicts directly what was in the paper. From the fuller data, there's no doubt that there's no Majorana." The 2018 paper claimed to show firmer evidence for Majorana particles than a 2012 study with more ambiguous results that nevertheless won fame for Kouwenhoven and his lab at Delft Technical University. That project was partly funded by Microsoft, and the company hired Kouwenhoven to work on Majoranas in 2016. The 2018 paper reported seeing telltale signatures of the Majorana particles, termed "zero-bias peaks," in electric current passing through a tiny, supercold wire of semiconductor. One chart in the paper showed dots tracing a plateau at exactly the electrical conductance value that theory predicted. Frolov says he saw multiple problems in the unpublished data, including data points that strayed from the line but were omitted from the published paper. If included, those data points suggested Majorana particles could not be present. Observations flagged by Frolov are visible in the charts in the new paper released last month, but the text does not explain why they were previously excluded. It acknowledges that trying to experimentally validate specific theoretical predictions "has the potential to lead to confirmation bias and effectively yield false-positive evidence."
Two physicists in the field say extra data Kouwenhoven's group provided them after they questioned the 2018 results shows the team had originally excluded data points that undermined its news-making claims. "I don't know for sure what was in their heads," says Sergey Frolov, a professor at the University of Pittsburgh, "but they skipped some data that contradicts directly what was in the paper. From the fuller data, there's no doubt that there's no Majorana." The 2018 paper claimed to show firmer evidence for Majorana particles than a 2012 study with more ambiguous results that nevertheless won fame for Kouwenhoven and his lab at Delft Technical University. That project was partly funded by Microsoft, and the company hired Kouwenhoven to work on Majoranas in 2016. The 2018 paper reported seeing telltale signatures of the Majorana particles, termed "zero-bias peaks," in electric current passing through a tiny, supercold wire of semiconductor. One chart in the paper showed dots tracing a plateau at exactly the electrical conductance value that theory predicted. Frolov says he saw multiple problems in the unpublished data, including data points that strayed from the line but were omitted from the published paper. If included, those data points suggested Majorana particles could not be present. Observations flagged by Frolov are visible in the charts in the new paper released last month, but the text does not explain why they were previously excluded. It acknowledges that trying to experimentally validate specific theoretical predictions "has the potential to lead to confirmation bias and effectively yield false-positive evidence."
Re:Don't be Fluffy. Hiding contradcitory results (Score:4, Insightful)
Not necessarily. I've run into lots of researchers who were seriously taught that any data point more than two standard deviations from what you thought it should be is an outlier and should be removed as part of good "data cleaning" practice.
It could be fraud, but stupidity is almost always a more likely explanation.
Re: Don't be Fluffy. Hiding contradcitory results (Score:1)
Or just , science is hard, shit happens.
Re: (Score:2)
Excluding inconvenient data isn't "shit happens." It's shit that was made to happen. It's not necessarily malicious though. The state of education on the basic tools of science is quite poor.
Re: (Score:3, Insightful)
Sometimes a piece of equipment is simply calibrated incorrectly, and the data point is just wrong. Figuring out why it was wrong could take an enormous amount of effort to no actual useful end.
Furthermore, statistical tools assume certain kinds of randomness are the primary causes of the observed distributions (i.e. calculating a standard deviation is silly on a bimodal distribution; using a chi square test on a gaussian is silly, too, etc.).
Non-randomness that causes a datapoint to be far off makes can i
Re: (Score:1)
Not relevant here. Read the summary FFS
Re: (Score:2)
Re: (Score:2)
Yup. That's the attitude all right. Fortunately it's science, so if you don't figure out your mistakes someone else eventually will. Quite embarrassing though.
Re: (Score:2)
Yup. That's the attitude all right. Fortunately it's science, so if you don't figure out your mistakes someone else eventually will. Quite embarrassing though.
That sounds nice standing at a whiteboard or a keyboard, but what you are saying often just does not matter.
A scientist improves his techniques a certain way because he has enough reason to believe it will give better results, not necessarily usually to solve a specific data point.
Likewise, veteran software engineers do not generally refactor code to solve one specific bug (data point) in the giant pile of bugs. They refactor because an area of code needs improvement, and expect a number of bugs (data poin
Re: (Score:2)
I have no idea what you're talking about, but you're definitely not a scientist.
Re: (Score:2)
Nice bluster. Stick with your first four words and stop there. What is confusing you is I know what I am talking about.
Re: (Score:1)
Then it was fraud.
Stupidity may the reason they committed fraud.
Re: (Score:2)
Although the GP is obviously trolling, I wouldn't dismiss the case so quickly. The data plot in question is not a simple XY graph. The data form a color-coded "band". In order to take data out of it, someone carefully cut a portion of the band from the middle section and pasted the band back together so that it looks continuous again. Also, the right part of the band was cropped out completely. The result shows a continuous band, that proves the paper's point, thereas the discontinuous band rejects it. Even
Re: (Score:2)
That does sound pretty damning. Will have to keep an eye on retraction watch to see how it plays out.
Re: (Score:1)
Read the summary you moron.
Majuriana Particles ? (Score:1)
I wouldn't be surprised if the researchers had a few too many marihuana particles before deciding that it would be a great idea and publishing the paper.
It's a chiffon (Score:2)
If You think it's Marjorana but it's not, it's Chiffon.
https://youtu.be/LLrTPrp-fW8 [youtu.be]
Quantum Clippy (Score:5, Funny)
"It looks like you are trying to kill a cat. Would you like probable help with that?"
Math hard (Score:2)
Welcome to the 20s, where the truth is never true (Score:5, Interesting)
Re: (Score:2)
'Never'?
That seems like an overstatement.
Publish or Perish (Score:1)
Re: (Score:2)
Quality is a slippery concept when it comes to papers. You can have one paper that is panned for one conference's referees yet another conference's referees think it is the bees' knees. Then there are the reviewers who look to see if you referenced their papers in yours. If you don't, they don't like your paper. And each discipline has various "correct" lines of thinking. Tell them something that contradicts one of those lines and no publishing for you.
Over time I expect it all evens out, but you won't know
Re: (Score:3)
Over time I expect it all evens out, but you won't know that until years later.
Personally, I've become quite fond of Thomas Kuhn's explanation for this. Basically, he argues that this is just how science works and you have to take the good with the bad. A consensus forms and until that consensus becomes clearly falsified, it actually makes sense to cling to it because that consensus creates a certain worldview—a paradigm—that the science can organize around and form a coherent argot to describe.
If you look at the sciences that have run into trouble recently—psycholog
But it was a big win! (Score:5, Funny)
You changed the result by looking at it!
You don't say (Score:2)
No One Knows What Happened to Majorana (Score:2)
The brilliant Italian physicist Ettore Majorana disappeared while taking a boat trip from Palermo to Naples on 25 March 1938, he was 32. He did two strange things just before his disappearance - withdrawing all his money, and then sending a letter to a colleague saying he was going to disappear. Where he actually disappeared is unknown, there are conflicting reports of his being seen on the journey. No motivations, or earlier signs, are known to account for any rash or strange behavior.
The major theory is s
Ah, 2021 - When Truth is in Science is back (Score:2)
Wonder what else we will "discover" is in error?
- Bigfoot does not exist?
- NASA actually landed men on the moon?
- Wearing face masks reduces the spread of air born disease?
- Vaccines actually reduce disease symptoms and spreading?
- Their was voter fraud in the USA? (And by the orange party!)
Oh
Next up, the Cortana Fermion (Score:1)
... which will produce a quantum computer with unparalleled AI capabilities, running nearly 90% slower than its competition, guaranteeing the wrong answer to at least 50% of questions put to it, and instantly being exploited by ransomware gangs who will demand 50 BTC for the score of last week's football game.
Science (Score:1)
But wait, it was "science." It must not be questioned. It is impossible for it to be wrong. Off with their heads!