Updates from 2024

Another year ends. Here are some updates from my end and also some thoughts. 

• I’ve become director of IDEAL for this year (and will take my sabbatical the following year as my reward). I’m happy to note that Avrim Blum and I are hosting Idan Attias as a postdoc at the institute. We’ve started working on some interesting problems on learning-augmented algorithms, and he’s working on many other interesting things with various people at UIC and TTI-C. 

• My student Jake Maranzatto finished his Ph.D. this year and has started a postdoc with Sennur Ulukus at the University of Maryland. In his dissertation, he solved a variety of interesting problems at the intersection of combinatorics and information theory. You can read it here. 

Hooding Jake at the UIC commencement. 

• Two progenitors of the deep learning revolution won the 2024 Nobel Prize in Physics, and scientists employing machine learning won the 2024 Nobel Prize in Chemistry. I understand the chemistry prize, but it took me a while to come up with a theory of what happened with the physics prize. The best explanation I can come up with is that physics, which is considered less exciting these days, is attempting to share in the credit for some of the most groundbreaking advances in computer science.  

• On a related note, AI continues to upheave many fields, including math, and I’ll be discussing some of that at the JMM next week in Seattle, where AI is the overarching theme. Deepmind's latest specially-trained software can already achieve silver-medal performance on the IMO. How long will it be until a future such system makes real progress on research problems? (I'm guessing not that long.)

• I am quite excited to say that I’ve agreed to become the Editor-in-Chief of a new top-level diamond-access journal published by the AMR at the intersection of mathematics and data science, machine learning, and artificial intelligence. Stay tuned for official info on this new venture soon. 

• Culminating a decade-long project, the brain of a fruit fly was completely mapped. This was done with the help of AI, and this result might also help advance AI by giving insights into (natural) neural networks.  I have a feeling there will probably be interesting things to post about AI every year.

Neurons in the fruit fly brain, by Philipp Schlegel

• Donald Trump has been elected President for a second time. There is, of course, always much to say about him, but I'll stick to academia/science here. There, I hope to see Title VI better enforced against universities (including my own) with respect to antisemitism, a reduction of DEI initiatives at the NSF and other funding agencies, and a renegotiation (downward) of overhead rates on grants. I’m, of course, not holding my breath for getting all three. And I’m not even listing the elimination of federally-backed student loans because that would take an act of Congress, and they don't seem likely to act.

Wishing everyone a happy New Year's celebration, and then it's on to 2025!

On The Beginning of Infinity

Here's my review of "The Beginning of Infinity," a book that touches on many of my interests. Its author advocates a philosophy that's much closer than that espoused by the author of the previous book I blogged about.

---

The Beginning of Infinity: Explanations That Transform the World by David Deutsch
My rating: 5 of 5 stars

The Beginning of Infinity presents a strong thesis about the importance of explanatory knowledge and its relationship to limitless progress. I’d heard Deutsch talk about this topic before, and I had assumed this would be yet another typical nonfiction book belaboring the same points. A friend whose tastes I respect recommended I read it anyway, so I gave it a try. I’m very glad I did!

The books thesis in clearly laid out, and the arguments for it are very persuasive. Deutsch confronts many possible philosophical objection, and attempts to obliterate each one by using very general arguments. He convincingly takes on inductivism, empiricism, justificationism, relativism, instrumentalism, and of course post-modernism, among many other philosophies. In fact, his arguments against other philosophies are probably stronger than his argument for his own thesis, but I think that still fits with the main worldview of the book: that progress requires replacing mistaken ideas.

Deutsch reinforced many of my views on some topics (eg the existence of objective values) and convinced me of some others (on why political compromise is bad). Even in parts of the book where I found him less convincing (eg in his defense of the many-worlds interpretation of quantum mechanics or his views on AI), I feel I still learned something and have more to ponder.

I also appreciated Deutsch’s fearlessness. For example, he doesn’t buy into environmental sustainability and is willing to buck most academics on sacrosanct topics. You know where Deutsch stands and why. Go read this book if you have any interest in the philosophy of science and of human progress.

View all my reviews