The Last Physicist

For those coming here from reddit or similar, I’ll come clean. I had high expectations for this book. As a physicist myself, a book about a physicist using the scientific method to put some hard rules to magic and delve into the mechanics had me almost giddy with excitement. So I dove into this book, and, I’ll be frank, came away disappointed.

The book follows the MC, Will. In another fun coincidence, Will works at Fermilab in Chicago. Fun fact, I still have a Fermilab login, and every six months they hassle me to do the most painful password reset, despite the fact that all I ever use them for these days is access the internal wiki from the Dark Energy Survey, of which I am a member. Initially, this was great. Another connection between me and the MC.

The author has Will talking about code, the training dataset used for his AI, and his particle physics work. As an aside, I am also a software engineer (and worked as one for a handful of years), and I currently do work in AI and machine learning in the renewable energy domain. And that’s how I know that the author, Dominic, is probably not that familiar with AI, nor coding, nor physics. Which is totally fine. Expecting people to only write about subjects they are a subject matter expert in is ridiculous. But, for any other coders out there, you probably feel my pain when you watch CSI and and you hear “I’ll create a GUI interface using Visual Basic, see if I can track an IP address. Instead, you get something like this:

He injected the Mirrored IDs into the neuromantic portion of his counter-attack, switching ports as the multi-vectored branches of arcane lightning finally connected with the enemy.

Like sure, I don’t expect someone to have all of the knowledge, but better in my mind to keep it simple but correct, instead of jumping down the rabbit hole and saying things which are incorrect. Dominic, if you ever read this, I’m always happy to act as a consult, and I know book two is still being developed (and I will be reading it). As an example, the premise of the black hole trigger is super interesting, why ruin it by saying that if black holes became small enough, they would turn into white holes? That’s not how Hawking radiation works, and it’s not even needed as a plot device.

But anyway, lets put that to the side, because once Will goes isekai mode, the inaccuracies of modern physics and code should be a minimal issue. I just wanted that scientific exploration of a magic system. Here, Will is helped by his AI companion, and I thought that was great. It’s a nice solution to crunching the number to have a supercomputer in your head, and I was all for it.

Okay, so let’s just break this down;.

Things I liked:

• The AI had some good banter. Sassy AIs are great.
• Initially using time dilation to accelerate projectiles is a great use of power.
• Initial fun exploration of the magic system.

Things that I wish were done a bit better:

• Progression isn’t earned. The MC literally starts his reincarnation per se with incredibly OP powers that trivialises all challenges.
• Progression seems made up and stats don’t make sense. The power scaling of the book is explicitly “Each level is twice as powerful as the prior one.” And yet, right at the start of the book (literally less than 5% of the way in), Will hits something with the energy the energy of 400 kilograms of TNT. By the end of the book, when he is canonically more than a million times stronger, he’s still hitting things this hard (judging by the craters and impact on the environment). His power seems effectively as weak or strong as the plot needs. On top of this, if you are going to have power scales be exponential, but then constantly have Will defeat creatures which are explicitly stated to be a thousand to a million times stronger than him… the power scale sort of stops having any meaning?
• Will barely uses the scientific method. All investigation is done by the prompting of the AI, and when will makes a breakthrough, it is always because of some intuition, or feeling, or some vague sense, and never by rational inquiry. If the whole point of the book is having someone come in and break down a magic system to exploit it, why is that pretty much absent from the book? I would have loved to see some mathematical relationships, or quantifying various aspects, Will getting the AI to run optimisation problems on how to maximse magical power as a function of mana usage, etc.

And, I think more than any of this (spoilers below):

Look, I really wanted to love this book. I really wanted a novel where someone exploits the hell out of a fantasy magical system using modern knowledge and scientific methods. But this doesn’t scratch that itch at all for me.

That all said, it is fun, it was a good read, and I will be reading book two.

Most of my quibbles above are me coming in with expectations that are probably higher than your average reader, simply because of my background.

I hope that book two does delve more into that scientific approach to exploring the magical system, because that is what I’m looking for most of all in this series.