Avi Loeb and Freeman Dyson on the future of the universe

This morning I was delighted to receive an email from Avi Loeb, who’s paper on the (far) future of astronomy we discussed yesterday. Avi shared with me a conversation he had by email with another noted theorist, Freeman Dyson.

The premise of Avi’s paper is that about a trillion years from now, all extragalactic light sources will cease to be visible due to the accelerating expansion of the universe. Future astronomers would therefore be stuck looking only within their own galaxy. Not only would the absence of extragalactic sources make for a (apparently) lonely universe, but it would deprive future astronomers of the tools that we have used to arrive at our current understanding of cosmology, such as extragalactic supernovae and the cosmic microwave background (CMB) radiation. Despite this, Avi suggests an observational signature by which these astronomers could still derive the standard cosmological model: hypervelocity stars.

In their correspondence, Freeman, in his role as an “incurable optimist,” suggests a means by which future civilizations could avoid this fate. He proposes that a civilization could harness the gravitational energy of a group of galaxies in some way as to pull together colossal collections of galaxies before they were thrown apart from each other by cosmological expansion. Perhaps this will remind you of another one of his well-known ideas for civilization-scale engineering, the Dyson Sphere. Freeman suggests searching for such anomalous overdensities of galaxies to detect such “cosmic engineering.”

Avi notes two possible observational signatures of such engineering: redshift surveys and the Sachs-Wolfe effect. Redshift surveys such as SDSS make 3D maps of the large scale structure of the local universe by measuring the distance to many galaxies. By the Sachs-Wolfe effect, if some advanced civilization pulled together an cluster of galaxies far bigger than “normal” superclusters, then we could detect it from the energy lost by CMB photons as they climb out of the cluster’s enormous gravitational potential well. Signatures of such anomalous overdensities are not actually observed in either type of survey, but Freeman notes that given the trillion-year timescale of the problem, “we have plenty of time to start doing it ourselves.”

Their wonderful senses of humor make the full correspondence a pleasure to read. The conversation is reproduced in its entirety below with permission from Freeman and Avi – we hope you will find the ideas they discuss as fascinating as we did.

Avi Loeb:

Dear Freeman,

In 2002 I wrote a paper about the long-term future of our Universe (Phys.
Rev. D65, 047301, 2002). Back then, you asked me in an e-mail to let you
know if I have any interesting follow-up thoughts on the subject.

Attached is a short paper that I had just submitted for publication on the
same subject. I would love to hear any comments you might have about it.

With best wishes,

Freeman Dyson:

Dear Avi,

Thank you for sending the paper. I found nothing wrong with it.
It presents a dismal picture of the future awaiting our descendants.
Since I am an incurable optimist, I raise the question, how much this
future could be changed by a large-scale intervention of intelligent life.
A very rough estimate indicates that large-scale “cosmic engineering”
could be feasible. Using available gravitational energy as the motive
power, roughly ten percent of the mass from one percent of the observable
universe could be collected within a volume small enough to remain
permanently bound together gravitationally. The collection could be done
in a single Hubble time and could then be maintained with small active
adjustments. So our descendants would stay in communication with a
hundred million galaxies instead of only one. It is also possible that
some of our more advanced colleagues elsewhere in the universe already
began this process. We should look out for evidence of large-scale
coordination of gamma-ray bursts or other phenomena indicating high
velocity movement of large masses. It would be interesting to examine
such possibilities in detail. Thankyou for the suggestion. Yours ever,
Freeman Dyson.

Avi Loeb:


Your underlying assumption is that intelligent beings prefer as much
company as possible. I can only say that as I get older I prefer to stay
away from other people as much as possible, since I have the feeling that
I heard it all. Extrapolating to the distant future, I am entirely
complacent with us being surrounded by vacuum and protected by an event
horizon. Aside from the benefit of not having any distractions, this
will reduce the risk of a hostile invasion by another civilization.


Freeman Dyson:

Each to his own taste. History without hostile invasions would be very
boring. FD.

Avi Loeb:

Putting prejudice aside, we already have relevant data for testing your
proposition about “cosmic engineering” on very large scales. SDSS provided
us with a map of the distribution of galaxies out to a redshift of z=0.3
(and SDSS III is now reaching farther out). The biggest bound systems in
the survey are clusters of galaxies, containing at most ~10^{15} solar
masses or ~1000 galaxies each. The existence of clusters is fully
consistent with the initial conditions we detect in the microwave
background at redshift z=1000. Since the initial conditions are Gaussian,
it should be easy to identify “cosmic engineering” in the form of a rare
overdense region (supercluster) containing many more than ~1000 galaxies.
A region compact enough to bind a million galaxies against cosmic
acceleration would have imprinted an anomalously large Sachs-Wolfe effect
on the microwave background or would have induced unusually high peculiar
velocities. We do not see anomalies of this magnitude. The biggest
supercluster in our vicinity is the Shapley supercluster, but it is
expected to be desolved in the future by the cosmic acceleration according
to the calculation in

Munoz, J, & Loeb, A. “The Density Contrast of the Shapley Supercluster”,
MNRAS, 391, 1341

Of course, it is possible that the Shapley supercluster is still “under
construction”, or that “cosmic engineering” operates on much smaller


Freeman Dyson:

That is disappointing. On the other hand, if our colleagues have been
too lazy to do the job, we have plenty of time to start doing it
ourselves. FD.

About Nathan Sanders

I am a graduate student in astronomy at Harvard University. I’m particularly interested in how galaxies have changed over time and the chemical evolution history of the universe. I’m currently working with Alicia Soderberg on observations of supernovae and their host galaxies; investigating how massive stars explode and enrich the interstellar medium. I graduated from Michigan State University in May, 2010.

1 Comment

  1. My favorite part:
    “Your underlying assumption is that intelligent beings prefer as much
    company as possible. I can only say that as I get older I prefer to stay
    away from other people as much as possible, since I have the feeling
    that I heard it all.”
    Thanks for the post, Nathan!



  1. Astronomical methods for the year 1 trillion « astrobites - [...] Update: you can read Avi Loeb and Freeman Dyson’s discussion of this issue in our latest post. [...]
  2. WMAP’s Closing Comments: ΛCDM Stands Strong | astrobites - [...] get harder in the future; a trillion years from now, the Universe will have expanded so much that cosmology won’t…
  3. Hacedores de mundos | Naukas | Cuaderno de Cultura Científica - […] antepasados de dicha civilización se las ingeniaron en su momento para acumular todas las galaxias que pudieron antes de…

Leave a Reply