Nobel Prize in Physics (& Computer Science?) - Computerphile

The 2024 Nobel Prize in Physics is awarded to John Hopfield and Geoffrey Hinton “for foundational discoveries and inventions that enable machine learning with artificial neural networks”. This video features Juan Garrahan, Phil Moriarty and Mike Pound... More links and info below ↓ ↓ ↓

This video was filmed and edited by Sean Riley.

[jarlsparkley]

The 60 symbols version is one second longer on my end. what kind of content are you holding out on us that you’re giving away to the physics nerds?

[volodyadykun]

YouTube should've implemented shared video long long time ago

[tomburns]

John Hopfield wasn't the first to describe the formalism which has been subsequently popularised as "Hopfield networks". And probably the Nobel Committee should have chosen someone like Shun-ichi Amari for the third spot on the prize.

It also seems much fairer to the wider field and long history of neuroscientists, computer scientists, physicists, and so on to call them "associative memory networks" and not "Hopfield networks", i.e. Hopfield was definitely not the first/only to propose the network some call "Hopfield networks". For instance, after the proposal of Marr (1971), many similar models of associative memory were proposed, e.g., those of Nakano (1972), Amari (1972), Little (1974), and Stanley (1976), which all have a very similar (or exactly the same) formalism as Hopfield's 1982 paper.

Today, notable researchers in this field correct their students' papers to replace instances of "Hopfield networks" with "associative memory networks (sometimes referred to as Hopfield networks)" or something similar. I would encourage you to do the same in your current/future videos.

I deeply regret making a similar mistake regarding this topic in one of my earlier papers. However, I am glad to correct the record now and in the future.

Refs:
D Marr. Simple memory: a theory for archicortex. Philos Trans R Soc Lond B Biol Sci, 262(841):23–81, July 1971.
Kaoru Nakano. Associatron-a model of associative memory. IEEE Transactions on Systems, Man, and Cybernetics, SMC-2(3):380–388, 1972. doi: 10.1109/TSMC.1972.4309133.
S.-I. Amari. Learning patterns and pattern sequences by self-organizing nets of threshold elements. IEEE Transactions on Computers, C-21(11):1197–1206, 1972. doi: 10.1109/T-C.1972.223477.
J. C. Stanley. Simulation studies of a temporal sequence memory model. Biological Cybernetics, 24(3):121–137, Sep 1976. ISSN 1432-0770. doi: 10.1007/BF00364115.

[SunnyKimDev]

1:19 I like his way of thinking. ML didn't "steal the Nobel Prize" from physicists, but owes a lot to physics and the Novel prize is asserting such.

[TommiLipponen]

Turing award winner Leslie Lamport always refers to physics when he talks about his papers and algorithms.

[shadowmil]

Even Nobel Winning Physicist jobs aren't safe from AI.

[zombieinjeans]

The irony is, science started without any disciplinary borders. It was just “natural sciences.” Those borders peaked and we’ve been slowly moving back to Natural Sciences ever since.

[feedthechunk]

I love Phil's passion! I wish I had Physics Professors like that.

[AI-Life-]

I just discovered this channel and I’m already hooked! The way you break down topics is brilliant. Subscribed immediately!

[EdeYOlorDSZs]

As an MSc AI student at UvA I took a course in 'theory of complex systems' where statistical physics was at the forefront. I was fimiliar with the concept of energy based models from my deep learning/ML courses and I could see the connection to neural networks where the particles in the system find an orientation that minimises system energy. This is very analogous to stochastic gradient descent in multi layer perceptrons (term coined by Hinton himself). It's really cool to study AI right now :D

[bryceblankinship]

I love how this channel keeps the early 2010s style of video journalism, only thing they've changed is adding visual edits versus literally drawing on a pad.

[Dndo]

So, why bring up the Ising model and then not explain anything?

[tomholroyd]

8:54 Coffee, a Reversi board, and a pair of dice --- tools of the Physicist

[tradetor]

I am just glad us statistical & computational physicist get more recognition

[clbgrmn]

This was a great video on a very interesting topic. And even moreso, very well edited to show the crossover of the fields and expertise. Thanks for creating this and sharing it with us!

[EarendilMitsos]

What i kept from this is that there are a couple of lucky/happy souls that have both Prof. Moriarty and Prof. Pound as their supervisors! 🎉🎉🎉

[KingTine]

"Yes, I believe the fundamental research in machine learning, such as deep neural networks, is inherently physics. Reality is filled with intelligent systems, where intelligence is a sort of emergent property. Therefore, studying these systems is fundamental to physics, as it is the goal to understand our reality

[dkickelbick]

I still don't understand why an achievement in computer science gets an Nobel Prize in Physics. Just because, there is Boltzmann in it? Or you can describe an neuronal network as particle interaction? It doesn't make it Physics. Economy uses physics models as well, it doesn't make it an achievement in Physics neither. "Physical Systems"? Humans are physical systems as well, it does not make it an achievement in Physics. The "problem" is not that a computer scientist gets the price, you don't have to be an Physicist to get it. Or that Computer Science isn't important or no hard science. But it should be an achievement in Physics to get the Nobel Prize in Physics. If computer science led to new insights in Physics, it would totally fine. But here all the explanations are the other way around. Physics is used to make progress in Computer Science, which is great. But it is not Physics. Like chemistry is not Physics or Biology is not Physics.

But what the heck, who cares. Now, Physics and Chemistry are AI slop as well.

[aidanthompson]

Backpropagation -> self-correction mechanism for ANN

[VincentGroenewold]

Personally I really don't care much for the Nobel prize, I worked in science for a long time and the one thing I hated about it is the politics (internally and between groups and institutes). I worked there for the science, in pure form preferably. I was that guy that came in in the weekends, while it wasn't required at all for my level, but I just liked doing it. Later on I started to see more of the inner workings and the management of things... and I bailed out eventually. Still love science, not so much the current implementation. And a prize kinda makes that point in a symbolic way, science isn't about prizes or competition... well, it shouldn't be.