The Official Cricketweb Science Thread!

[RossTaylorsBox]

I think we're disagreeing on our definition of "nerds".

 

[StephenZA]

How did this dross end up in the science thread, this seems far more politics than science.

 

[RossTaylorsBox]

You've made multiple posts about the academic publishing industry.

 

[StephenZA]

You've made multiple posts about the academic publishing industry.

Yours is about Epstein... my posts on the academic publishing has everything to do with science, and the ability of academia to research that science. Epstein has nothing to do with science or academia.

 

[RossTaylorsBox]

Dude, the source of funding for scientific projects is 100% to do with science, in particular rich "donors" like Epstein. The guy literally offered to fund projects and threw parties at his mansion hoping he could convince someone to help him seed humanity with his DNA. The funding problem in general is especially pronounced at Media Lab where it's become more about gimmicks than rigour, it's an absolute blight on scientific advancement. I'm not surprised Ito is involved. I am surprised that Stallman finally got pinged for being a gross human.

 

[StephenZA]

Dude, the source of funding for scientific projects is 100% to do with science, in particular rich "donors" like Epstein. The guy literally offered to fund projects and threw parties at his mansion hoping he could convince someone to help him seed humanity with his DNA. The funding problem in general is especially pronounced at Media Lab where it's become more about gimmicks than rigour, it's an absolute blight on scientific advancement. I'm not surprised Ito is involved. I am surprised that Stallman finally got pinged for being a gross human.

Linking to a post such as this: wired.com_the-problem-with-rich-people-funding-science is talking about funding problems in science, and the problems with becoming involved with people that have money. Jumping on a few individuals who have had contact and done things with Epstein is not talking about science or academia.

Anyway whatever, I thought that the comment would have been better in one of the other news threads, you not gonna change my mind about that. Nuance is important.

 

[RossTaylorsBox]

You're absolutely naive if you think this issue is restricted to "a few individuals" and presuming that it doesn't negatively influence science.

 

[StephenZA]

You're absolutely naive if you think this issue is restricted to "a few individuals" and presuming that it doesn't negatively influence science.

Did you even look at the article I linked to?

 

[StephenZA]

https://svs.gsfc.nasa.gov/13326

In April, astronomers wowed the world with the first real-life picture of a black hole. But that blurry, still image of the supermassive monster in the galaxy M87 doesn’t really convey just how wildly a black hole’s immense gravity distorts its surroundings. Now, images from computer simulations highlight in more detail how a black hole warps spacetime like a fun house mirror — and how that affects the appearance of its glowing accretion disk of infalling material.

 

[SillyCowCorner1]

I'm amazed at the image one would get as they get just past the Schwarzschild radius, looking back at us...that person would see the universe as a circle concentrated at a 'pole'.

Reading this the first time in Kip Thorne's book left me spellbound.

 

[Spark]

Google claiming that it's hit quantum supremacy in its quantum computing experiments.

 

[vcs]

What does this mean?

 

[StephenZA]

What does this mean?

https://www.nature.com/articles/d41586-019-03213-z

 

[Spark]

What does this mean?

It means that — in the fairly small subset of problems that quantum computers can actually do with any reasonable efficiency — they’re claiming superior performance to a classical computer.

 

[DriveClub]

It means that — in the fairly small subset of problems that quantum computers can actually do with any reasonable efficiency — they’re claiming superior performance to a classical computer.

And the small subset of problems that quantum computers can actually do are simple problems that can be easily verified.

 

[StephenZA]

It means that — in the fairly small subset of problems that quantum computers can actually do with any reasonable efficiency — they’re claiming superior performance to a classical computer.

Not to quibble. But it is more than just superior performance, it is claims of efficiency that classic computers could never achieve, even theoretically.

What my real question is, is the practical application of quantum computers and the ability to program them to model complex interactions. In the nature article, they mention releasing experimental results on the error correction, now that would be fascinating to see.

Edit: Article on the argument over quantum supremacy actually means.
https://www.quantamagazine.org/google-and-ibm-clash-over-quantum-supremacy-claim-20191023/

Insert when nerds fight meme!

 

[vcs]

I will have to read the actual papers to understand the Big-O difference in Google's Quantum computing algorithm vs. the IBM classical algorithm (not that I'm likely to understand the Math or the technical details LOL). That article claims it solves a problem that is exponential for classical computers in polynomial time, which is the main takeaway (not 10000 years vs. 2.5 days vs. 200 seconds or whatever, which depends on the actual problem instance).

This comes as a surprise to me. A paper I read by Scott Aaronson some time back said that it is not clear whether quantum computing moves any of the non-polynomial problems (that is, problems where conventional algorithms cannot scale) to the set of polynomial-time problems. Even in the famous example of factoring semi-primes, it is only conjectured that no classical polynomial-time algorithm exists. Moreover, it is not at all clear whether any of the NP-complete problems, which are the fundamental "litmus test" problems for evaluating the limits of computability, become tractable with quantum computing. Aaronson doesn't think they do.

https://www.cs.virginia.edu/~robins/The_Limits_of_Quantum_Computers.pdf

 

[weldone]

This is just massive if true, right? It's scary how much power this one organisation is accumulating.

 

[vcs]

I'd prefer to look at it as a cool scientific result that significantly advances the field. Since they did it in close collaboration with a University physicist, I am sure they'd have to release the code etc. on a public license. May not particularly matter though, if only Google is capable of putting together the quantum hardware to run it on.

IBM Research has definitely been working in this field for many years now, and I'm sure there is a lot of academic research on it as well, so I don't think they can monopolize it.

 

[StephenZA]

I'd prefer to look at it as a cool scientific result that significantly advances the field. Since they did it in close collaboration with a University physicist, I am sure they'd have to release the code etc. on a public license. May not particularly matter though, if only Google is capable of putting together the quantum hardware to run it on.

IBM Research has definitely been working in this field for many years now, and I'm sure there is a lot of academic research on it as well, so I don't think they can monopolize it.

If google funded it directly... not really; it is only if funding for the project came direct from government that there is a requirement to publish publicly. IBM in the past have been pretty good in publishing research, uncertain regards google, think they keep lots pretty well hidden.