Chalnoth said:
Basically, the current cosmological model assumes gravity for all times. Presumably quantum gravity effects may become important at early times, but this is by no means clear. Unfortunately there just isn't any way we know today to really say how gravity might "turn on", and our one candidate theory of everything (string theory) predicts that gravity always exists.
...
Chalnoth, Steven Weinberg does not agree with you on a couple of points here.
You could watch the last 12 minutes of the talk he gave at Cern on 7 July.
He proposed an alternative TOE to string that is a good deal simpler and more straightforward--and warned string theorists in the audience that string might not be needed, ordinary "good old" QFT might work after all, running couplings, asymptotic safety.
Said "string might not turn out" to be how the world is---the fundamental theory (not merely an effective one, but fundamental and predictive out to arbitrarily high energy) might be an asym. safe development of the QFT we already have.
And in those last 12 minutes he described his own current work, which describes a way in which gravity could "turn on." At the big bang, since the G constant runs with scale, it might in fact be zero in the UV limit. The cosmological constant on the other hand might be very large in the UV limit, and so it would decline as the universe expanded. So that inflation would end.
Weinberg cited a paper by Martin Reuter and Alfio Bonanno which already explores these ideas, using the (scale/density dependent) running constants of asymsafe QG to explain early universe behavior, like turning on of gravity, like inflation (without needing a graviton!) and like graceful exit from inflation (without needing a slow-roll potential gimmick).
OK, so you may think this is weird or speculative. But it was Weinberg talking and he said there will be a conference on this very thing at Perimeter (in November). So I checked the conference website and participants included Weinberg and Martin Reuter and Renate Loll and Vincent Rivasseau and a bunch of other impressive people.
So bizarre as it seems, there is a precise mechanism being studied by which gravity can "turn on" and inflation can occur (without fantasizing a graviton field) and then automatically "turn off"---and this is being studied by good people---and the subject has become hot.
Very little has been said about this in public, which I guess is normal with new lines of research.
If anyone is curious, here is an early paper on it by Bonanno Reuter.
http://arxiv.org/abs/0803.2546
Primordial Entropy Production and Lambda-driven Inflation from Quantum Einstein Gravity
Alfio Bonanno, Martin Reuter
(Submitted on 17 Mar 2008)
"We review recent work on renormalization group (RG) improved cosmologies based upon a RG trajectory of Quantum Einstein Gravity (QEG) with realistic parameter values. In particular we argue that QEG effects can account for the entire entropy of the present Universe in the massless sector and give rise to a phase of inflationary expansion. This phase is a pure quantum effect and
requires no classical inflaton field."
For the last 12 minutes of Weinberg's talk, go here, and skip the first 58 minutes by dragging the button.
Here is the video:
http://cdsweb.cern.ch/record/1188567/
And here is a more central link if you want additional context.
http://indico.cern.ch/conferenceDisplay.py?confId=57283
The Asymptotic Safety conference (in November at Perimeter) has a webpage here:
http://www.perimeterinstitute.ca/en/Events/Asymptotic_Safety/Asymptotic_Safety_-_30_Years_Later/
The list of invited speakers is here:
http://www.perimeterinstitute.ca/en/Events/Asymptotic_Safety/Invited_Speakers/
It won't say anything in the conference description about turning on gravity, they are not trying to communicate to a broad audience. The conference is for the few experts who know about it. But we know Weinberg's current work on Asymptotic Safety is about the turning on of gravity (and the turning off of inflation) because he said so, and Weinberg will be there because he's one of the invited speakers listed (as are both Bonanno and Reuter) so one can read between a few of the lines.
======================================
EDIT TO REPLY TO NEXT POST
Hi Chalnoth, the reason Weinberg treats it as a theory of everything is that it includes the Standard Model of particle physics, or some modified version thereof. What he is talking about is QFT as a candidate theory of everything.
By enlarging what he calls "good old" QFT to include gravity (along with everything else) using asymsafe methods. (a UV fixed point). What he suggests, if you listen to his talk, is that "that is how the world is". It is all subsumed in a single mathematical framework--just that the framework is not stringy. But you can choose not to call it a TOE if you prefer.
I agree that it is interesting stuff. Everybody should listen to Weinberg's 7 July talk, or at least the last 12 minutes. It is low key but offers a remarkable change of viewpoint, considering who the speaker is.