Tuesday, January 09, 2018

Me, elsewhere

Beginning 2018, I will no longer write for Ethan Siegel’s Forbes collection “Starts With a Bang.” Instead, I will write a semi-regular column for Quanta Magazine, the first of which -- about asymptotically safe gravity -- appeared yesterday.

In contrast to Forbes, Quanta Magazine keeps the copyright, which means that the articles I write for them will not be mirrored on this blog. You actually have to go over to their site to read them. But if you are interested in the foundations of physics, take my word that subscribing to Quanta Magazine is well worth your time, not so much because of me, but because their staff writers have so-far done an awesome job to cover relevant topics without succumbing to hype.

I also wrote a review of Jim Baggott’s book “Origins: The Scientific Story of Creation” which appeared in the January issue of Physics World. I much enjoyed Baggott’s writing and promptly bought another one of his books. Physics World  doesn’t want me to repost the review in text, but you can read the PDF here.

Finally, I wrote a contribution to the proceedings of a philosophy workshop I attended last year. In this paper, I summarize my misgivings with arguments from finetuning. You can now find it on the arXiv.

If you want to stay up to date on my writing, follow me on Twitter or on Facebook.


John Fredsted said...

I completely agree, Quanta Magazine does an excellent job. I have 'subscribed' to it for quite some time. Looking forward to read your articles as well there.

Phillip Helbig said...

"If you want to stay up to date on my writing, follow me on Twitter or on Facebook."

I have accounts on neither. I realize that one can access some content without an account, but nevertheless the signal-to-noise ratio in both of these media seems rather low. Life is short! I still think that the world of blogs is woefully inefficient compared to the old web (calling it web 1.0 is so web 2.0!) of email, usenet, and so on. I can just manage via RSS feeds, but with Firefox having changed their philosophy so that LiveClick will stop working, I have to waste time finding a new solution there. (At the same time, I am still using the same newsreader I was using 25 years ago, and it still works fine, and my work at customization and so on pays of because it is stable.)

tl;dr: Please post as much as you can here. If you can't post it, post a link to something freely available if you can. If you can't do that, still post a link and we can decide whether we want to subscribe or not. I do subscribe to things (mostly print), but it is a hard decision to make unless one knows in advance what one will get. I've also given up reading a couple of blogs which moved to Slate or some other subscription service, not because I have to pay (though that still leaves the question whether it is worth the money), but because the web sites are so full of ads and subscribers get so many emails and so on that I decided that it wasn't worth the trouble.

David Schroeder said...

Both Quanta and Starts-With-a-Bang have been my favorite destinations, along with your Backreaction blog. Very glad that you are continuing to contribute to at least one of them. With a nice hot peppermint-mocha coffee in hand, I'll be trekking over to Quanta to read your article on asymptotically safe gravity, in a moment.

Sabine Hossenfelder said...


I beg to differ. The signal-to-noise ration on twitter and facebook is as high or small as you make it. Both give you the option to set up lists and in that way function pretty much like an RSS feed. It's up to you who you add to that feed.

Communication on twitter is a complete disaster, even after the raise to 280 characters. I find it extremely cumbersome to figure out who is commenting on what and hence have pretty much turned of most notifications.

Communication on facebook works quite well, actually much better than on blogger because the comment sections are better readable (as with allowing answers to comments and also edits).

Having said that, it's not my intention to talk you into using facebook, just that I personally think it has its uses. If you want to avoid that, it's not difficult to convert a twitter-feed to rss, see eg here.

Phillip Helbig said...

RSS is good for checking up on things one wants to read (as is a sensible newsreader), but a problem with blogs etc is that one has to sort through new ones by hand. This is not a real problem, though, since I have enough to read already, though I fear that I might be missing something good. (With usenet, I automatically see every new group, and can subscribe if I want, and unsubscribe if I don't want to follow it.) An RSS feed shows me what I haven't read, and a preview function lets me quickly check what I haven't read, but as far as I know there is no way to hide a complete thread (many blogs, including this one, don't even have threaded comments), so on the whole blogs are less efficient than usenet. (On usenet I can of course use my own, customized editor, so that is more efficient as well.)

However, even when I look at the Twitter feeds of people who sometimes post interesting stuff, most of it is noise. OK, maybe there are some high signal-to-noise feeds, but again there is the time required to find them. And even if retreated things are interesting, I need to see them only once (another advantage of usenet; if I've read a post which was posted to more than one group, it is marked as read in all of them).

John Fredsted said...

Sabine: Thanks for your lucid article at Quanta Magazine. I find it highly interesting, and it is novel to me, that asymptotically safe gravity, if that indeed turns out to be correct, restricts the number of fundamental particles (as well as the mass splitting of the top-bottom doublet).

Uncle Al said...

Congratulations on your new public platform!

Physical theory is arrogant for "knowing" where the answer is. "Accepted" theory is forever empirically sterile. It cannot be corrected by publishing more of it while boasting of output volume.

"The R&D Function" Harvard Business Review 61(6) 195 (1983)
... Clever research is fruitful. Anything else is managerial bafflegab.

Udi Fuchs said...

Doesn't asymptotically safe gravity imply that entropy should scale with volume? This seems to contradict with Bekenstein-Hawking entropy, which scales with surface area.

I think that trying to find a UV fixed-point to 4 dimensional gravity is interesting. But the motivation is to gain deeper understanding of why it doesn't exist.

Sabine Hossenfelder said...


Yes, I think that's right. There is a nice, recent, paper summarizing this argument, which is here. Though one should more carefully say that it violates the microcanonical interpretation of the BH entropy.

Kakaz said...

@Philip I've read a lot, and subscribe many blogs and internet media platforms by RSS. Tiny Tiny RSS platform works very well for this purpose. Of course it is probably not so easy. It consists of two elements: server and client working on android mobile platform: just application you download and install on your phone or tablet.
Server part had to be installed on cloud, your own or hosted computer. Then it periodically fetch RSS feeds into MySQL database, and ... You may read blogs you subscribe! It works very well. Try it!

Matthew Rapaport said...

Well this is too bad. I will miss you but perhaps once in a while you can summarize for us on g+.

Willard Mittelman said...

Your new arXiv article is great. There may be some problems, however, with the fourth paragraph on p. 6. For one thing, the lower-case "lambda" at the end of the paragraph needs to be upper-case, since you're clearly referring to the actual measured cosmological constant (cc). Also, your "typical" finetuning argument there begins with "We don't know lambda;" but it makes more sense to say "We don't know lambda-prime" (where lambda-prime is the "bare" cc of general relativity), since we do have some information about lambda itself, as the first paragraph on p. 6 makes clear. (Maybe this is a typo? In any case, the "lambda-prime" near the end of the paragraph should be changed to "lambda," if I'm correct that the first lambda should be lambda-prime.) It's lambda-prime that we really don't know anything about, other than that it's a free constant; and it therefore seems reasonable to take the bare cc to have a uniform probability distribution. And in that case, the finetuning argument here seems to be a strong and viable argument.

Sabine Hossenfelder said...


Thanks, I'll check the lambda's. I also noticed some other typos and will have to replace it anyway. You cannot use a uniform probability distribution because that's not normalizable, so think again.

Phillip Helbig said...

At least in astronomical cosmology, the convention is that lambda = Lambda/3H^2. (Sometimes Lambda has dimensions of inverse time squared, sometimes of inverse length squared, regardless of whether c is set to 1. It's always good to mention the dimensions of ones variables.

Phillip Helbig said...

"Tiny Tiny RSS platform works very well for this purpose. Of course it is probably not so easy. It consists of two elements: server and client working on android mobile platform: just application you download and install on your phone or tablet.
Server part had to be installed on cloud, your own or hosted computer. Then it periodically fetch RSS feeds into MySQL database, and"

The only android mobile platform I have is an eBook reader (tolino), and android is the underlying OS but this is not accessible to the user. In any case (though it has a minimal web browser) I don't use it for reading blogs (or only very rarely). A MySQL database seems like overkill for this.

Remember, we are talking about duplicating the functionality of a 1980s newsreader written in a few thousand, at most, lines of code. This sort of confirms my point that blogs are not as efficient.

Willard Mittelman said...


Thanks for your reply. Perhaps one could use a Gaussian distribution sharply peaked at zero, motivated by the proposal of Ng and van Dam that Lee Smolin discusses in arXiv:0904.4841 (p. 14). In that case, the finetuning argument seems (to me, at least) to be a good argument.

Art Brown said...

Typos watch: In "Screams for Explanation", reference 4, should "Porter, W" be "Williams, P"? Thanks for the articles.

bud rap said...


I was reading your post on superfluid dark matter and it went dark.

sniffnoy said...

Question: Is the constant you call lambda even *meaningful*?

I seem to recall seeing this argument from Sarah Kavassalis somewhere, arguing why the "mismatch" between lambda and Lambda (to use your terms) was not the problem people claim it is, back when she did relativity. The argument was that lambda isn't even meaningful, because it's a constant energy [density] outside the context of general relativity. General relativity (in which I'm including also theories that attempt to encompass it) is the only thing that provides an "absolute zero" for energy; outside that context, all energies only make sense up to an additive constant. Since the constant lambda is neither itself a difference nor being compared with anything, to suppose that Lambda must be equal to it, as many do, isn't just wrong but is in fact a type error!

I may be mistaken here (if so please credit the mistake to me rather than Dr. Kavassalis :P ); perhaps lambda is in fact an energy difference and there is no type error, and the only error is the one that you point out. I don't work on any of this sort of thing so this is not obvious to me. But I seemed to recall seeing this argument and thought it was worth pointing out.

Sabine Hossenfelder said...


Thanks... Sorry about that. I should never have changed my bibliography to last name first.

Sabine Hossenfelder said...


How about you read my paper to answer that question.

neo said...

Sabine re:

"Even still, physicists can already put the ideas behind asymptotic safety to the test. If gravity is asymptotically safe — that is, if the theory is well behaved at high energies — then that restricts the number of fundamental particles that can exist. This constraint puts asymptotically safe gravity at odds with some of the pursued approaches to grand unification. For example, the simplest version of supersymmetry"

"But not only does asymptotic safety provide a link between testable low energies and inaccessible high energies — as the above examples demonstrate — the approach is also not necessarily in conflict with other ways of quantizing gravity. That’s because the extrapolation central to asymptotic safety does not rule out that a more fundamental description of space-time — for example, with strings or networks — emerges at high energies. Far from being disappointing, asymptotic safety might allow us to finally connect the known universe to the quantum behavior of space-time."

Doesn't symptotic safety rule out string/M theory, as it rules out supersymmetry given the assumptions behind predicting higgs mass 126 and top quark mass assumes no SUSY?

is symptotic safety gravity compatible or an implication of loop quantum gravity?


Sabine Hossenfelder said...


ASG is an extrapolation from low to high energies. It is in principle possible that at higher energies some new degrees of freedom come into play which will change the extrapolation. That's why it is, at least in principle, compatible with other approaches to quantum gravity. An argument that has been made against it being compatible with string theory is that ASG doesn't allow for the microcanonic interpretation of the black hole entropy, and then you have to ask how well-established that is in string theory.

ASG is not an implication of loop quantum gravity. It's believed to be related to causal dynamical triangulations. Best,


neo said...

ASG that correctly predicts the higgs mass and top quark mass assumes no new physics from fermi to planck scale, and is not compatible with additional particles such as SUSY nor extra dimensions. i can post papers if you want but S&W assumption in predicting 126 higgs, rules out SUSY. ruling out SUSy rules out string theory. and 126 was what LHC observed

doesn't this version of ASG rule out string theory?

btw in your future quanta articles where do you want to discuss the issue you raise?

Sabine Hossenfelder said...


There are many variants of SUSY and the only one I am aware of having been shown to be incompatible with ASG is the MSSM, which is meanwhile ruled out by LHC data anyway. If you have other references, I'd be interested to have a look.

You are welcome to discuss this on facebook or twitter. As I said above, I prefer facebook.