Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

nobelprize.org

Post reply on HN
Print view View original

711–720 of 876 posts

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#711

Post by ngcc_hk »

slashdave wrote: post #631
api wrote: post #73

Earlier quoted context omitted.

My sense is that we might have reached the limits of what we can do in high-energy or fundamental physics without accessing energy levels or other extreme states that we currently can't access as they are beyond our capacity to generate. From what I've read (not a professional physicist) string theory is not testable unless we can either examine a black hole or create particle accelerators the size of the Moon's orbi…

Although rare, there are cosmic rays that do span very high energies. You can access these from, for example, atmospheric showers.

Looking at history it is a bit odd that cosmic rays as high energy collider were not a field by itself.

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#712

Post by ngcc_hk »

jhbadger wrote: post #626
passwordoops wrote: post #62

Earlier quoted context omitted.

You know what, I can support this for its predecessor, Lotus123. If ANNs are worthy of the prize in physics, then so is this

VisiCalc, surely. Lotus 123, much like Excel, was just following in the footsteps of the original spreadsheet. https://en.wikipedia.org/wiki/VisiCalc

This practical idea of spreadsheet has much more impact to society than many innovation. Nobel prize worthy.

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#713

Post by Onavo »

KoolKat23 wrote: post #692
Onavo wrote: post #656

Earlier quoted context omitted.

The deep learning variety of neural networks are heavily simplified, mostly linear versions of biological neurons. They don't resemble anything between your ears. Real life neurons are generally modeled by differential equations (in layman terms, have many levels of feedback loops tied to time), not the simplified ones used in dense layer activation functions. Here are some examples https://snntorch.readthedocs.io/en…

Would that be equivalent to Weight matrice parameters?

Ish, take a look at the curves of the spiking neural network function, they are very different from the deep learning nets. When we "model" biological neural nets in code, we are essentially coming up with a mathematical transfer function that can replicate the chemical gradient changes and electrical impulses of a real neuron. Imagine playing a 3D computer game like Minecraft, the physics is not perfect but they are "close enough" to the real world.

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#714

Post by cha42 »

hoseja wrote: post #152
openrisk wrote: post #115

Earlier quoted context omitted.

> it’s hard to do experiments at the boundaries of what we know this applies primarily to fundamental physics. There are many areas of applied physics (materials, fusion, biophysics, atmospheric physics, etc. etc.) where the main constrain is understanding complex systems. These areas are quite crucial for society.

As evidenced by the search for room temperature low pressure superconductors.

Or toothpath dynamic.

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#715

Post by MichaelDickens »

erinnh wrote: post #600
ajkjk wrote: post #581

Earlier quoted context omitted.

Everyone's first thought when they read something is whatever the social norms say you're supposed to think (peer review = good, publishing without peer review = not science somehow?), but shouldn't you stop and wonder why the esteemed scientist wrote that line instead of just dismissing it? Otherwise you are only chiming in to enforce a norm that everyone already knows about, which is pointless. One of the really re…

Im not a scientist and do not know how these things work out, but wouldn't it be possible for scientist to simply publish their papers online without peer review if that is what they want? What's stopping them from doing so?

Nothing stops them, some people do do that. Two examples that come to mind are Aella's research on fetishes[1] and Scott Alexander's research on birth order effects[2]. But you don't get academic credibility by publishing online without peer review, and it's much harder to get university funding.

[1] https://aella.substack.com/p/fetish-tabooness-vs-popularity

[2] https://slatestarcodex.com/2018/01/08/fight-me-psychologists...

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#716

Post by ngcc_hk »

sph wrote: post #159
acoupleofts wrote: post #148

Earlier quoted context omitted.

> The least amount of wars in any century since WW2 :/

We still have a decade or so to get back to average Also, WW2 being so utterly destructive, back to back after an arguably even worse global war, skews the stats a little.

Ww1 is not a global war. The Asian and pacific theatre is not in play. It is mostly an European war.

But it triggers ww2 because the treaty is too hard on Germany. And crazy people has its soil prepared for their madness.

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#717

Post by matusp »

etiam wrote: post #478
matusp wrote: post #353

Earlier quoted context omitted.

The techniques highlighted in this prize are not really that useful for deep learning.

You mean besides bringing it into existence at all?

Please explain how Hopfield network influenced modern deep learning models based on supervised differentiable training. All the "impactful" architectures such as MLP, CNN, Attention, come from a completely different paradigm, a paradigm that could be more straightforwardly connected to optimization theory.

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#718

Post by blululu »

FrustratedMonky wrote: post #582
blululu wrote: post #567

Earlier quoted context omitted.

Physics is pretty old and it has always been about understanding the fundamental structure of reality. If it doesn’t tell you how it all goes round then it is not physics: plain and simple.

Do physicists know the fundamental structure? They have some mathematical approximations that work for some measurements. They can make some predictions in some areas, but the same approximations break down in other areas. So the fundamentals aren't 'known'. Some think measurements is engineering. So are the physicist that focus on building an apparatus to measure a theory, they are engineers? So only the theoretical…

Physicists have learned quite a bit about the fundamental structure of the cosmos in the last 500 years. We can get into philosophical quibbles over what is knowledge and the relationship between approximations to reality, but we have clearly developed a very rich understanding of how the world works. A lot of the fundamentals are very clearly known. Entropy and statistical mechanics have been part of physics for 150 years and have clearly enhanced our understanding of the universe. Claude Shannon’s work definitively helped us understand the world more deeply. I think deep learning is interesting but it would be a stretch to claim that this has enriched our understanding of the universe by a large margin. Definitely not as much as Shannon’s work.

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#719

Post by zeofig »

aDyslecticCrow wrote: post #587
DangitBobby wrote: post #550

Earlier quoted context omitted.

"Massive" seems overstated. Also doubt that it's net positive.

You're probably only thinking about the modern chatbots when you say that.

Perhaps you have less idea what he's thinking than you think you do.

Re: Nobel Prize in Physics awarded to John Hopfield and Geoffrey Hinton [pdf]

#720

Post by zeofig »

Looks like the Nobel wants to fast track its march towards irrelevancy. What a joke.

I see a lot of people saying "physics has stalled" etc., which is not the case. It may be the case for high energy physics (I would not even make that statement myself with any confidence), but there is a lot of other physics being done.

Post reply on HN