Kip Thorne: The Man Who Imagined Wormholes and Schooled Hawking (2007)
discovermagazine.com
discovermagazine.com
One thing is strange though, they're conceding bets on outcomes that are far from conclusive/decided. For example, the bet about the black hole & information paradox. There are many different ways both of them could be wrong/right. One such possible answer resides in string theory. A few string theorists put forth the idea that black holes may not have a singularity at all (thus, no paradox). Let me explain:
Massive enough stars collapse into neutron stars when they die. A neutron star is made out of neutron degenerate matter. Basically, when a star dies, gravity rips apart/compresses/implodes everything down to neutrons. The gravity is so strong that helium, hydrogen etc. cannot exist within the star, it immediately gets broken down into more fundamental components. It gets ripped apart and turned into this "neutron degenerate matter". This is because neutrons have enough energy or "strength" to resist gravity. This is why a neutron star is stable. But if a star has enough mass (more gravity), it's possible for even neutrons to break down and succumb to gravity. Some people have hypothesized that they can break down into quark stars but we have yet to see them. Instead, we currently believe they form singularities (a black hole) since there is nothing more fundamental to hold back the force of gravity.
Massive enough stars collapse into black holes/singularities because after neutrons/quarks, there is nothing more fundamental to break down into. But in string theory, there is. In string theory, strings are the fundamental component of the universe. A few (not all, it's pretty fringe/new) string theorists have hypothesized that a black hole could be simply a "fuzzball". It's basically a star made out of strings. It would behave and have the exact same properties that current black holes have (light cannot escape, no limit on size, extreme gravity, etc) but it resolves 2 of the paradoxes -- no singularity (no invoking infinity & breaking QM) and no information paradox.
See here: https://en.wikipedia.org/wiki/Fuzzball_%28string_theory%29
If string theory is correct, and fuzzballs are the true description of black holes, the bet itself is wrong. It would be like betting whether the sun was made out of cheese or wax. The answer would be neither.
Hawking's explanation for his position was that if black holes didn't actually exist much of his research would be incorrect, but at least he'd have the consolation of winning the bet. Stephen Hawking (1997). Black Holes (VHS). New River Media
"The bet was very nonpolitically correct: He gave me a subscription to Penthouse magazine when I won."
The problem exists classically as well as in QFTs: if I drop a trillion particles of energy e into a black hole or a two trillion particles of energy e/2 with pairs stacked up on top of each other instead, how do I tell the difference?
With no-hair, I can't.
If the horizon persists into future timelike infinity, I don't really care about this -- the information distinguishing between these two cases is hidden behind the horizon and I can shrug it off. The same is true if there is a remnant rather than a complete final evaporation.
But if the horizon goes away altogether, I may need to distinguish between the trillion and two trillion particle cases; that's especially true if I care about unitarity, as I should when using a QFT.
The main problem here is the singularity, and the inevitable collision of the trillion or two trillion particles with it. If I can abolish the singularity so that these particles evolve normally within the horizon rather than being squashed into a singularity, my information loss problem goes away, and I don't care what happens to the horizon at all. A useful high energy theory of gravity should solve this, whether it does it by changing spacetime, or fundamental particles, or produces a working renormalization (one could conceivably defang the relevant infinities in principle, analogously to how asymptotic safety, but not necessarily using ASG's nonperturbative renormalization), or some combination, or something else entirely.
The retreating horizon is another problem, but any approach there has to deal with at least stellar (and potentially smaller) mass black holes and also supermassive black holes. The former may evaporate entirely in finite time, and one has to be careful about tricks that prevent that somehow because the latter may have arbitrarily flat spacetime at the horizon. These features tend to break remnant and RG approaches.