I wanted to comment on all of his points but I gave up, some simply do not even permit counter arguments and these are definitely not definite proofs. Definite possibilities maybe.
He argues that if someone could develop simulation technology, they would.
Not necessarily, but if we assume that this is correct, combine it with a non-conditional - someone could - and you might have a point. On closer inspection though, the claim that someone could would be based on the laws that govern our universe, which would just be an argument in favor of simulations inside the universe we inhabit and here's my problem with that:
I find it problematic that people use data collected from their own experience in this reality to come to conclusions about a hypothetical base/parent universe.
If they are evaluating the likelihood of an advanced post-human civilization that is both capable and willing to run a complex simulation, and using some empirical estimation conclude that it is probable and then proceed to argue that such civilizations produce huge numbers of simulations, then you might argue that we are more likely to be simulated than not. And Nick Bostrom makes a claim of that sort, though more formally stated as a disjunctive argument.
One way you would justify such a claim would be by appealing to the principle of indifference and assigning equal probabilities to all individual outcomes, but we know for sure that we aren't part of any potential simulations that are running inside the universe we inhabit and that we can observe, that is, the ones that branch out from the node that is our own universe.
So you cannot rightfully count them towards the probability of our being simulated.
What could be done instead is to evaluate the probability of there being simulations outside of our world, which also requires the value of the probability of there being any other worlds. But even then, how do we make assumptions about a world separate from our own when the laws that govern it might bear little resemblance to our own?
And on what basis do we trust our knowledge of this universe if it could very well be designed and modified by the simulators and even shut down at any point.
Perhaps I'm missing some important fact that's obvious to the people who subscribe to these beliefs.
Then he mentions the anthropic principle which as far as I understand it doesn't really have the same conclusion as the one he drew.
The conclusion usually isn't that of the existence of an intelligent designer.
It's often formulated vaguely as - It isn't remarkable that we observe the fine-tuned universe since if we are to observe anything, we must observe precisely the conditions that enabled us to evolve. The conditions that aren't compatible with the existence of observers simply aren't observed.
Regardless, I don't see how the simulation hypothesis could be a solution since we'd still be left with the necessity of explaining the base reality of the root of the simulation tree unless perhaps the simulation at some level isn't designed by an evolved intelligent civilization but occurred randomly as a consequence of a random assemblance of a Boltzman brain or some such entity, however absurd it may seem.
Then he argues that we can model everything using our mathematical knowledge. Perhaps the universe is mathematical but our knowledge of it is incomplete and some might go so far as to say that it's necessarily going to remain incomplete. Or take Roger Penrose as a modern example of a mathematician who is highly skeptical of the computability of various phenomena such as consciousness which he bases on Godel's incompleteness theorem. In any case, we certainly do not understand the mechanism underlying conscious experience and, at least from my limited perspective, I find it highly doubtful that we ever will.
As far as Fermi's paradox is concerned, I wouldn't say that the claim that this hypothesis might explain a phenomenon is enough to show its validity, especially since there are simpler solutions.
Now, that's all speculation and perhaps all that should be said is - is it really consequential?