Nilesh Jasani's Reviews > Existential Physics: A Scientist's Guide to Life's Biggest Questions

Existential Physics is a rare book I read twice back-to-back. This was not because it was difficult to understand (actually, some of the technical and philosophical discussions are) but because I wanted to ponder more over so many new points the author had put forward.

Ms. Hossenfelder is an accomplished, empirical experimentalist scientist who is also a gifted writer. She is clear about what she believes in and why. More importantly, she is clear about what she does not believe in, cannot believe in, would not believe in, and where she has no way of knowing whether to believe in or not. Her conceptual clarity may not convert others who do not share the same views, but they would surely spark many worthwhile new thoughts.

The author is not exactly a fan of theoretical physicists. If the previous books had not clarified this clearly, it comes out far more forcefully in Existential Physics. The armchair physicists are frequently put in the same category as the prophets, sages, charlatans, and others from time immemorial that claimed to have figured out how the world works. The path chosen to come down heavily on this group is through the notion of scientific versus ascientific theories. The author makes it clear that just like most religious assertions, theories like inflation, loop quantum, string, etc., are speculative without any basis in known or even knowable science. The book is about how much we really know and what we can know. It does the first part extremely well. But it isn't easy to agree with the author's binning of a lot of good theoretical work despite their speculative nature.

Some propositions and assertions, often misnamed as theories, are clearly wrong. In taking a liberal view, the author not only shies away from picking the most obvious wrong ones, but she also refuses to acknowledge that some of the theoretical physicists' suppositions are significantly less wrong compared to made-up assertions from other fields. Historic, non-rational theories are issued as given; they are unmodifiable and non-mathematic. To equate them with mathematical, modifiable forms with current theoretical physicists' is needlessly extreme in many cases.

As the author says herself, we need hypotheses for science to progress. From a given circle of relatively proven "science," one path of progress is by observing and fitting more data (and, in the process, expanding our equations/theories). This is the path of experimentalists like the author. The other path is by first hypothesizing what further progress could be and then looking for data for confirmation or repudiation.

The second method, by definition, will start speculatively. It may often have to turn bombastic, too, like in search of better math (or alternate math) that can deal with the infinities/singularities that recur in current mathematical equations (of science) but somehow get resolved without problems in reality as we experience. For example, the Navier-Stokes equation could be more resolvable using completely outlandish math, like Fermat's last theorm. If such math is proposed Today, it may not appear to have any utility at the outset until it suddenly begins solving unsolvable equations using current methods.

This hypothesis-first path of theoretical scientists has worked spectacularly for centuries. Given the state of Today's data, tools, and science, it might not be possible for humans sitting behind a desk to produce the correct theories out of thin air like Einstein before, but proposals that make some sense could be put forward by machines too. If they are not falsifiable in their first versions, many of these proposals could be modified/amended in their equations to become more verifiable over time. This reviewer is with the author in seeing a diminished role for theoretical innovations in the periods ahead, but to equate their ascientific theories with other ascientific theories from the non-physics realm, like the way the author does, is extreme.

The book makes some excellent points on fundamental versus emergent theories. The author asserts that there are no strongly emergent theories. All emergent properties, like the structure of DNA or waves of an ocean (let alone consciousness or love the mother has for her child), emerge from more fundamental laws of particle physics/relativity, even if we may not be able to provide the full calculation linkages Today. As much as the author claims that she is not a reductionist - ontological or theoretical - her assertions against strong emergence appear ascientific. If Today's fundamental theories - along with 26 constants, 25 particles, four forces, and the current set of equations - cannot precisely explain why the H-O-H bond in a water molecule is 104.5 degrees and not 125 degrees or predicate why proteins take certain shapes but not others, there might be something in emergent properties not completely available in fundamental theories as we know them.

In other words, we are supposed to take it for granted that all higher-level emergent properties come out of the same lower-level equations even when it is not mathematically possible to achieve it currently. The reviewer, like the author, believes that there is no mind or consciousness or free will below the emergent properties of underlying forces and matter, but this is another ascientific assertion, in a way. This is not much different from multiverses the author argues against because they are ascientific and arbitrary.

The larger point is that almost all of us are prone to fall for unprovable assertions without realizing them. The author is far more obviously ascientific in her views on artificial general intelligence.

The book is not only good in addressing many philosophical issues but also in explaining many scientific concepts. One example is the discussions on why it might not always be possible to predict completely deterministic systems. The author explains this through quantum uncertainty and then the chaotic nature of calculations first, but soon moves beyond. Godel's incompleteness theorem and the halting problem are used to show how notwithstanding the rest, there are always problems that cannot be solved in finite steps using any mathematical system. There are many other problems that encounter singularities and infinities in the math we operate, even while they resolve beautifully in real life.

Here are some other small nuggets of information:

-- Is the world mathematical, or is math just the language we use for its description? The second is demonstrably true, while the first is an ascientific assumption. It is unlikely we figured out what the world is or even what the best description language of the world is on our first try

-- All laws are time irreversible, except the collapse of the wave function and the information lost at the black hole.

-- Quantum collapse is something that happens when a wave function interacts with any sufficiently impactful macro matter. There are non-zero chances of quantum equations being emergent and not fundamental.

-- Time irreversibility means that from any fully described system, we can deterministically determine both the full past and present.

-- Today's theoretical world is full of speculative conjectures on items 10-15 orders different from our best ability to observe - like inflation, protein decay, black hole evaporation, etc. One should realize that these ascientific theories are hopeless in many ways.

-- Quantum entanglement is observed non-locally but has to be enacted locally.

-- General relativity does not limit the speed of light but merely states that acceleration from below the speed of light level to above requires infinite energy.

-- No cloning means you cannot ever copy a neuron completely faithfully

Overall, the book has many extreme conclusions for almost any type of reader. The author stands her ground well. As a result, she will make her readers think deeply to understand what they honestly disagree with.