Closing remarks, by David Gross 271
that it is determined by supersymmetry, people who believe in anthropic arguments
would just say“OK, that is not a parameter that scans”. So, my problem with the
predictability of anthropic arguments is that it they are incredibly imprecise and
easy to squirm out of. It is true, as Steven has emphasized to me, that in science
you do not get a choice to decide what. it. is that scieIice can do or cannot do:
Nature decides. True, but we have made incredible progress in science by pushing
predictability, and increasingly precise predictability: we do not just estimate that
the cosmological constant lies in some range; we calculate the gyromagnetic ratio of
the electron with incredible precision. This ever increasingly precise predictability
is onc of our most special and important scientific tools. I would like not to have
to give it up. But it is not up to me, as Steven correctly points out.
The other philosophical aspect of the anthropic principle that bothers me enor-
mously is that, unlike other principles in physics, it thrives on ignorance. Other
physical principles in physics get stronger the more we know, yet the anthropic
principle get stronger the less we know. Once we know something, especially if we
know it precisely, it disappears completely from the realm of anthropic arguments.
Anthropic arguments might be necessary, but they fly in the face of the success of
physics over the last few centuries. Hopefully they will not be required.
On a less philosophical note I find that the arguments that lead, in string theory,
to a multitude (a landscape) of possible universes to be very shaky. The so-called
vacua that define the landscape are not vacua at all but metastable states. As
such they are really not well defined in a truly precise sense. To define these we
would have to know how they were populated, where they come from. To answer
this question we are driven to understand eventually the full cosmological history
of such states, including the big bang. And although there are very clever and
somewhat convincing arguments for many metastable states, there is to date not a
single consistent stringy description of cosmology all the way back to the beginning.
Could it be that most or all of the landscape is ruled out since it does not fit into
a consistent cosmology--we do not know!
Most important, we do not know what string theory really is. When we say
“The theory leads to ...”., we do not really know what we are talking about. We
have many, often totally different, ways of describing approximate solutions to string
theory; but what is string theory? We do not know the basic formulation of the
“theory”, to which all of these different dual descriptions are approximate. I am
beginning to wonder whether we might be coming to the conclusion that string
thcory is inherently incomplete. Originally, many of us believed that string theory
represented a very dramatic break with our previous notions of quantum field the-
ory. That was good. We probably needed something that was a serious break with
quantum field theory( QFT) to solve the problems of quantum gravity, cosmological
singularities, etc. But now we have learned that string theory is not that much of
a break with QFT. Tn fact, our best definitions of string theory are QFTs. Maybe,
as Nathan (Seiberg) and Juan (Maldacena) remarked, any QFT is equivalent to a
