black hole memory



If you have an hour, Malcolm Perry explains to you what the black hole information loss problem is and how to solve it eventually (BMS and all that).

probabilities

According to 538 the probability for Hillary to win Iowa is currently 79%.
The probability for Ted Cruz to win in Iowa is 52%.

The probability that Britain will leave the EU is around 33% according to betting markets.

And the probability of rain tomorrow in Nassau, Bahamas is 10% according to wunderground.

The question is what "probability" means for each individual case above; I guess The Matrix has 100 different programs labeled "Nassau, January 21, 2016" in its archive and only 10 of them with rain ... but for some reason the programming director recently likes to pick one of the rainy versions (*).


(*) added later: It's raining.

asymptotic safety again

I wrote previously about this approach to quantum gravity and in some sense it is a reason this blog still exists.
Recently, Frank Saueressig et al. were able to answer an important question and argue that their "result vanquishes the longstanding criticism that asymptotic safety will not survive once a "proper perturbative counterterm" is included in the projection space."

There were some comments about it on Jacques Distler's blog, who posted about a different paper.
Let me admit that I understand less than half of what he wrote about soft gravitons and photons, but it does seem to me now that asymptotic safety is in better shape than he thought previously.

Something deeply hidden had to be behind things.

One of the web pages I read on a regular basis is called brainpickings.
Nothing earth shattering, but I like the sentiment expressed e.g. in this short piece about Albert.