Why science?
I’ve always been interested in science, in particular in astronomy, astrophysics, and space. One of my earliest memories - back in nursery school in New Jersey, I think - was watching one of the moon launches. I wanted that excitement to be part of my life! I never got to be an astronaut, but I did get to be part of the Planck Satellite team, and was privileged to be able to travel to the ESA Spaceport in French Guiana to watch the launch.
In between, I was lucky enough to have a supportive family, get a good education, and find inspiring teachers, mentors, and collaborators. They helped me model the universe, and helped me learn how to refine those models in the face of experimental and observational evidence. That is, they taught me to be a scientist.
Why this book?
The Random Universe shows how we can understand the world - do science - in the face of all-pervasive uncertainty. Not at all incidentally, we employ the same tools that we all use to navigate the world around us every day, even if not with such mathematical rigour. It’s all about those models of the universe - often called theories. Over the course of my career, I came to understand the scientific method as a process of constant revision and testing in the face of noise and randomness - and often the description of that randomness is exactly what we’re trying to understand. After all, our theories usually don’t tell us exactly what’s going to happen, but instead tell us how likely, how probable, different outcomes might be. Our best models depend on only a few numbers, with which we calculate the probabilities for the distribution of everything we can see in the Universe.
You come out in the book as an unrepentant Bayesian - why do you think the Bayesian/frequentist argument has gone on so long?
Bayesians are very objective about probability: given the same information (data from an experiment, a model that we are trying to check, or a set of models that we are trying to compare), the rules of probability describe the unique and rational way to use that information to come up with new probabilities - those rules tell us how we can learn from our experiences in a self-consistent way. But it is about learning and about how different probabilities mesh together. So the rules do require that we describe our prior information about the world, and this is where Bayesians get accused of subjectivity. But this does mean that we can use probability to describe the outcome of experiments: I am 95% certain, for example, that the expansion rate of the Universe is between 66.4 to 68.4 in the obscure units we use to measure this quantity, or the UK Met office is 40% certain that it will rain tomorrow.
Frequentists, however, try to limit the use of probability to the description of repeated events: the fraction of W bosons that decay into electrons at CERN, say, or the fraction of basketball games won by the Knicks in their glorious 2025-26 season. This makes it impossible to use probability to describe the actual numbers that we care about, such as whether England will win their next World Cup match, or the amount of dark matter in the Universe.
The caricature is that frequentists calculate precise numbers that no one cares about, and Bayesians calculate vague numbers to the questions we want precise answers to.
What’s next?
I am involved in a number of large experimental and observational collaborations, such as the Simons Observatory, a telescope being built in Chile to observe the Cosmic Microwave Background — light that last interacted with matter almost 14 billion years ago. It has higher sensitivity than any experiment to date, which we hope will let us refine our probabilistic models of the very early Universe, and possibly start to understand the quantum mechanical processes that took place much less than one second after the start of the Big Bang.
I am also working with the Euclid Satellite team, and just getting involved in starting to analyse data about how light from distant galaxies is deflected by more nearby mass, giving us a way to directly map the distribution of matter in the Universe on very large scales, and to compare that pattern with the predictions of our theories.
What’s exciting you at the moment?
Like many scientists, I am just starting to dip into using LLMs as a tool for my work, primarily for helping me automate programming tasks. I can explore data and try new methods much faster than ever before. The worry, of course, is that we will rely on these tools more than our colleagues and collaborators, and forget that one of the main products of any scientific endeavour is a scientific community.
More broadly, it’s an exciting time for cosmology. Different measurements of the same thing, especially the expansion rate (Hubble constant) that I mentioned earlier, are giving different answers. This is probably some sort of error in one or more of the analyses, but we are all hoping that it is actually the beginning of a crack in our Big Bang model.
These articles will always be free - but if you'd like to support my online work, consider buying a virtual coffee or taking out a membership:
Interview by Brian Clegg - See all Brian's online articles or subscribe to a weekly email free here
Comments
Post a Comment