2024 Nobel Prize in Physics Awarded for Applications of Statistical Physics to Machine Learning!

The 2024 Nobel Prize in Physics has been awarded to physicist John Hopfield and computer scientist Geoffrey Hinton.

The official prize citation is "for foundational discoveries and inventions that enable machine learning with artificial neural networks". However, I strongly prefer the phrasing along the lines of "for the statistical-physics basis of neural networks", which is how mathematical physicist Barry Simon described it.

Here is the information, press release, and other materials from The Nobel Foundation.

Naturally, I am strongly in favor of more Nobel Prizes being awarded for foundational interdisciplinary work, as that is the world in which I live. Unsurprisingly, many traditional physicists are arguing against and doing their common practice of staking territorial claim (as we also often see on the job market and in other arenas). This is an old battle, and we can look forward to a later phase when somebody gets a physics Nobel Prize for work in networks. That said, I am grateful to see people arguing about science, rather than about other ideological things!

2021 Abel Prize: László Lovász and Avi Wigderson

The 2021 Abel Prize goes to to mathematician László Lovász and computer scientist Avi Wigderson "for their foundational contributions to theoretical computer science and discrete mathematics, and their leading role in shaping them into central fields of modern mathematics."

Tales from the ArXiv: Was the AI Eaten by a Grue?

This paper, called NAIL: A General Interactive Fiction Agent, describes an autonomous agent for general parser-based interactive-fiction games.

(In my quick glance, I noticed a couple of slightly misleading statements about Infocom and their text-adventure games. For example, one figure implicitly seems to suggest that Trinity came out in the early 1980s, but it actually came out in 1986. I know, I know: I am being very picky.)

I hope that they didn't get eaten by a grue.

Or, given that I mentioned Trinity, maybe they reached into the beehive for a third time?

A Paper that Rejects Itself

What a conundrum! :P

Amazing. Train a network to classify papers (accept/reject). Then run the network on the paper describing the network, and it classifies the paper as a strong reject. This is why we can't have nice paper classifiers. (h/t @hardmaru) https://t.co/OC2baCI1Za pic.twitter.com/3N9wsKQ374

— Chethan Pandarinath (@chethan) December 21, 2018

(Tip of the cap to Karen Daniels.)

Print and Play Cards: Women in Science

These print-and-play cards of Women in Science include some nice artwork and some very familiar names. My favorites are the mathematicians, of course.

A couple of years ago, I blogged about playing cards of Women in Computing.

A Big Pile of ... Linear Algebra

That's right. In many cases, machine learning is a big pile of ... linear algebra.

Conway's Game of Life In Real Life (on an Ocellated Lizard)

Wow! This is amazing!

Here is the blurb on the Facebook post that goes with the Physics Today article (though I added the hyperlink): The ocellated lizard develops an intricate, ever-changing pattern of black and green spots when it matures. Now researchers have determined that the patterns on the animals' backs update according to a well-defined algorithm: Over a period of a month or so, a given scale will change color—from green to black or black to green—with a probability that depends on the colors of the scales around it. In essence, the reptile is the embodiment of a cellular automaton, a type of discretized model made popular by John Conway’s Game of Life and used to simulate the spread of wildfires, the firing of neurons, and other phenomena.

Physics Today's article is about a recent article in Nature called "A living mesoscopic cellular automaton made of skin scales".

Graph Alignment: A Major Open Problem

Graph Alignment is a major open problem in mathematics and computer science. pic.twitter.com/Q8EJDSYLge

— Mason Porter (@masonporter) April 20, 2017

A Movie About Ada Lovelace

Ada Lovelace's story (or at least a Hollywood version of it) looks like it's going to make the big screen.

(Tip of the cap to the Mathematical Association of America.)

Pokémon Go and the Traveling Salesperson Problem

I should have posted this article about Pokémon Go and the Traveling Salesperson Problem a couple of weeks ago when I first saw this blurb.

I thought about it again today when I posted the link as comments on posts by a couple of Facebook friends (who are both also former undergrad students of mine from Somerville College).

"Curry's Paradoxical Combinator": A 3rd Level Computer Scientist Spell

"Curry's paradoxical combinator" (see this page) is something that occurs in lambda calculus.

It sounds more like a spell, like Mordenkainen's Disjunction (or perhaps even Tasha's Uncontrollable Hideous Laughter).

Curry's paradoxical combinator is a 3rd level Computer Scientist spell. It requires an intelligence score of at least 13.

The fact that the last name is "Curry" (think Tim Curry) also adds to the effect.

(Tip of the cap to Quentin Miller.)

Tales from the ArXiv: "Other Naive Models"

This paper's abstract has the following final sentence: "Furthermore our transition probability model is up to 20% more accurate than other naive models."

Notice the phrasing "other naive models", which contrasts rather sharply with the phrasing "other, naive models".

Theorem: There Exist Arbitrarily Long Songs of Complexity O(1)

As discussed in a paper by Donald Knuth, the proof is due to Casey and the Sunshine Band.

A New Quasipolynomial Time Algorithm for Graph Isomorphisms

In case you haven't been paying attention, you may be interested in reading Jeremy Kun's post about the seminar covering the announced quasipolynomial algorithm for graph isomorphism. Once the preprint comes out and it is vetted, I hope the result does turn out to be genuine. Exciting times!

RIP UGCS (1989–2016)

Well, from the e-mail today (forwarded to my current account from a chain that started with my UGCS account), it looks like (barring a miracle) Caltech's UGCS cluster is going to bite the dust in a few months. As Blake Jones wrote, "Alas, poor envy! I knew it well." To quote Steve Ginzburg, it's the "End of an era". Indeed it is.

UGCS has served the Caltech undergraduate and alumni communities for 26 years, but technology has moved on. Forwarding will still occur, but I assume that in a few months my "awesome" undergraduate website (from the days of "Oh, cool! I can put up a public page on this thing called the Web!") will go away. (Maybe it will be visible on an internet archive.)

Recap of Somerville College's Ada Lovelace Bicentenary

Here is a recap from the Ada Lovelace Bicentenary that we held at Somerville College last Friday. Planning started in 2013, and last Friday was the big day.

"Women in Computing" Playing Cards

You can buy (or freely download) these Notable Women in Computer Science playing cards.

(Tip of the cap to the Eight of Spades, who was our opening speaker in Somerville College's event today to celebrate women in computer science.)

Computer Scientists Versus Computer Engineers

Yup, SMBC has illustrated the difference very nicely indeed.

Being a student at Caltech was like this. (Bring back the Lloyd Coke Machine!)

Tales from the ArXiv: "Finding a Mate With No Social Skills"

Well, the title of this paper says it all, doesn't it?

In fact, there is a wonderful ambiguity in the article title: it can either refer to finding a mate without using any social skills or, more amusingly, to finding a mate who doesn't have any social skills.

If it refers to the latter and people circulate this study widely enough, maybe people like me can finally have some hope? :)

Update (4/30/15): Here is my post on this topic for the Improbable Research blog.

Excursion to Bletchley Park: 2015 Edition

Every few years, we take a bunch of mathematics and computer students from Somerville on an excursion to Bletchley Park (home of the U.K. codebreaking during World War II). Today was another one of those treks.

As usual, I took some pictures from today's trip.

And here are my photos from our 2012 excursion.