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

AI winter is well on its way

blog.piekniewski.info

Post reply on HN
Print view View original

441–450 of 518 posts

Re: AI winter is well on its way

#441

Post by sh33mp »

timr wrote: post #422
bitL wrote: post #417

Earlier quoted context omitted.

Yeah, but the point here was that radiologists on average fared even worse. 83% is not impressive, but better than what we have right now in real-world with real people, as sad as it is. Obviously, best radiologists would outperform it right now, but average ones, likely stressed under heavy workload might not be able to beat it. And of course, this classifier probably works on certain visual structures better than h…

"Yeah, but the point here was that radiologists on average fared even worse." Except they don't. See the table in the original post. Also, comparing the "average" radiologist by F1 scores from a single experiment (as you've done in other comments here) is meaningless. Unless my doctor is exactly average (and isn't incorporating additional information, or smart enough to be optimizing for false positive/negative rates…

This thread is a microcosm of this whole issue of overhyping.

On one hand, we have one commenter saying he can train a model to do a specific thing with a specific quantitative metric, to demonstrate how deep learning can incredibly powerful/useful.

On the other hand, we have another commenter saying "But this won't replace my doctor!" and therefore deep learning is overhyped.

The two sides aren't even talking about the same thing.

Re: AI winter is well on its way

#442

Post by pizza »

bitL wrote: post #440
Florin_Andrei wrote: post #436

Earlier quoted context omitted.

> AI winters are a result of a massive disparity between the expectations of the general public and the reality of where the technology currently sits. I think they also happen when the best ideas in the field run into the brick wall of insufficiently developed computer technology. I remember writing code for a perceptron in the '90s on an 8 bit system, 64 k RAM - it's laughable. But right now compute power and data…

I wish GPUs were 1000x faster... Then I could do some crazy magic with Deep Learning instead of waiting weeks for training to be finished...

somebody is working on photonics-based ML http://www.lighton.io/our-technology

Re: AI winter is well on its way

#443

Post by sgt101 »

Erlich_Bachman wrote: post #267
sgt101 wrote: post #212

Earlier quoted context omitted.

It shows something about the game, but it's clear that humans don't learn in the way that alpha zero does, do i don't think that alpha zero illuminated any aspect of human intelligence.

I think that fundamentally the goal of research is not necessarily human-like intelligence, just any high-level general intelligence. It's just that the human brain (and the rest of the body) has been a great example of an intelligent entity which we could source of a lot inspiration from. Whether the final result will share a the technical and structural similarity (and how much) to a human, the future will tell.

In principle you are right. In practice we will see. My bet is that attempts that focused on the human model will bear more fruit in the medium term because we have huge capability for observation at scale now which is v. exciting. Obviously ethics permitting!

Re: AI winter is well on its way

#444

Post by kbp »

visarga wrote: post #306

AI winter is not "on its way". There is AI hype and anti-AI hype, and then there is actual practice. This article is anti-AI hype, just as bad as its opposite. In practice there are tons of useful applications. We haven't even begun to apply ML and DL to all the problems laying around us, some of which are quite accessible and impactful. The hype cycle will pass with time, when we learn to align our expectations with…

> AI winter is not "on its way". There is AI hype and anti-AI hype, and then there is actual practice. This article is anti-AI hype, just as bad as its opposite. In practice there are tons of useful applications.

I don't think the article was saying that AI isn't useful, but just that deep learning specifically is not an AI panacea, and that the current hype around AI is on its way out. The hype dying down and the associated buzzwords starting to repel money instead of attract it is all that's meant by AI winter, I believe, not that we'll run out of places where the techniques would be useful.

Re: AI winter is well on its way

#445

Post by erikpukinskis »

fourfaces wrote: post #158

The inconvenient but amazing truth about deep learning is that, unlike neural networks, the brain does not learn complex patterns. It can see new complex patterns and objects instantly without learning them. Besides, there are not enough neurons in the brain to learn every pattern we encounter in life. Not even close. The brain does not model the world. It learns to see it.

You’re close.

“learns to act in it”

we don’t even see the world, we see a hyperdimensional action space. Anything we can’t relate analogically back to some embodied action will literally be invisible to us.

Re: AI winter is well on its way

#446

Post by erikpukinskis »

edejong wrote: post #169
fourfaces wrote: post #158

The inconvenient but amazing truth about deep learning is that, unlike neural networks, the brain does not learn complex patterns. It can see new complex patterns and objects instantly without learning them. Besides, there are not enough neurons in the brain to learn every pattern we encounter in life. Not even close. The brain does not model the world. It learns to see it.

This post is very uninformed. "It can see new complex patterns and objects instantly without learning them." Except, it doesn't. It is clearly false. When animals grow up in an environment without certain patterns, they will be unable to see these patterns (or complex combinations of these) at a later stage. We see complex patterns as combinations of patterns we have seen before and semantically encode them as such.…

We learn to see patterns, but we see through physical and cultural action patterns that are simply present, not learned.

It’s like a river flowing... yes, the water molecules each “discover” the their path, but the path of the river is a property of the landscape. It is not learned.

Re: AI winter is well on its way

#447

Post by timr »

sh33mp wrote: post #441
timr wrote: post #422

Earlier quoted context omitted.

"Yeah, but the point here was that radiologists on average fared even worse." Except they don't. See the table in the original post. Also, comparing the "average" radiologist by F1 scores from a single experiment (as you've done in other comments here) is meaningless. Unless my doctor is exactly average (and isn't incorporating additional information, or smart enough to be optimizing for false positive/negative rates…

This thread is a microcosm of this whole issue of overhyping. On one hand, we have one commenter saying he can train a model to do a specific thing with a specific quantitative metric, to demonstrate how deep learning can incredibly powerful/useful. On the other hand, we have another commenter saying "But this won't replace my doctor!" and therefore deep learning is overhyped. The two sides aren't even talking about…

Agree that the thread is a microcosm of the debate, but ironically, I'm not trying to say anything like "this won't replace my doctor".

That kind of hyperventilating stuff is easy to brush off. The problem with deep-learning hype is that comments like "my classifier gets a ROC/AUC score of 0.8 with barely any work!" are presented as meaningful. The difference between a 0.8 AUC and a usable medical technology means that most of the work is ahead of you.

Re: AI winter is well on its way

#448

Post by ujal »

placebo wrote: post #433
ujal wrote: post #389

Earlier quoted context omitted.

I don't get the sentiment of the article either. I can't speak for researchers but software engineers are living through very exciting times. State of the art in numbers: Image Classification - ~$55, 9hrs (ImageNet) Object Detection - ~$40, 6hrs (COCO) Machine Translation - ~$40, 6hrs (WMT '14 EN-DE) Question Answering - ~$5, 0.8hrs (SQuAD) Speech recognition - ~$90, 13hrs (LibriSpeech) Language Modeling - ~$490, 74h…

My thoughts on AGI (at least in the sense of being indistinguishable from interaction with a human) are the same as my thoughts on extraterrestrial life: I'll believe it only when I see it (or at least when provided with proof that the mechanism is understood). This extrapolation on a sample size of one is something I don't understand. How is the fact that machine learning can do specific stuff better than humans dif…

This is what we know from Google about Duplex:

"To train the system in a new domain, we use real-time supervised training. This is comparable to the training practices of many disciplines, where an instructor supervises a student as they are doing their job, providing guidance as needed, and making sure that the task is performed at the instructor’s level of quality. In the Duplex system, experienced operators act as the instructors. By monitoring the system as it makes phone calls in a new domain, they can affect the behavior of the system in real time as needed. This continues until the system performs at the desired quality level, at which point the supervision stops and the system can make calls autonomously." --

Re: AI winter is well on its way

#449

Post by timr »

bitL wrote: post #429
timr wrote: post #422

Earlier quoted context omitted.

"Yeah, but the point here was that radiologists on average fared even worse." Except they don't. See the table in the original post. Also, comparing the "average" radiologist by F1 scores from a single experiment (as you've done in other comments here) is meaningless. Unless my doctor is exactly average (and isn't incorporating additional information, or smart enough to be optimizing for false positive/negative rates…

> Unless my doctor is exactly average Just to get back to this point: what if the vision system of your doctor is below average and you augment her by giving her a statistically better vision system, while allowing her to use the additional sources as she sees fit. Wouldn't be that an improvement? We are talking about vision subsystem here, not the whole "reasoning package" human doctors posses.

Again, check that table. It says a lot:

https://stanfordmlgroup.github.io/competitions/mura/

On just about every test set, the model is beaten by radiologists. Even the mean performance is underwhelming.

Re: AI winter is well on its way

#450

Post by colordrops »

paulie_a wrote: post #107
colordrops wrote: post #103

Earlier quoted context omitted.

Improvements in search, translation, image recognition and categorization, voice recognition, and text to speech off the top of my head. I'm sure there are a lot more.

Yeah but those are all pretty terrible to the actual end consumer. They might be cool technologies but at the end of the day, I am a user that hates dealing with them. IVRs are terrible a experience. Image recognition is iffy at best. Text to speak is terrible. In 10 years maybe they will have it hashed out... just like 10 years ago, or 20 years ago

I'd have to disagree. While IVR sucks (most current implementations don't use ML by the way), image recognition and categorization is at or better than human levels in most cases now. Cutting edge TTS is now nearly indistinguishable from a human. Just check out some samples [1]. And while translation still sucks, ML based translation is still far better than previous approaches.

[1] https://www.theverge.com/2018/3/27/17167200/google-ai-speech...

Post reply on HN