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

191–200 of 518 posts

Re: AI winter is well on its way

#191

Post by raducu »

empath75 wrote: post #170
joe_the_user wrote: post #81

This is a deep, significant post (pardon pun etc). The author is clearly informed and takes a strong, historical view of the situation. Looking at what the really smart people who brought us this innovation have said and done lately is a good start imo (just one datum of course, but there are others in this interesting survey). Deepmind hasn't shown anything breathtaking since their Alpha Go zero. Another thing to co…

What, no progress six months after achieving a goal thought impossible even just few years ago? Pack it up boys it’s all over but the crying.

I was thinking just that when reading the paragraphs about the uber accident. There's absolutely nothing indicating that future progress is not possible, precisely because of how absurd it seems right now.

Re: AI winter is well on its way

#192

Post by annabellish »

flohofwoe wrote: post #183
jarym wrote: post #177

I’m not sure how the hype wagon started but I for one am glad it’s about to pop. I am working (founded) a startup and while we have AI on the roadmap for about a years time, it isn’t something that’s central to our product. (We already use some ML techniques but I’d not confidently boast its the same thing as AI). Cue an informal lunch with a VC guy who takes a look, says we’re cool and tells us just to plaster the w…

> ...just to plaster the word AI in more places... I bet that's how the new "AI-assisted" Intellisense in Visual Studio got greenlit: https://blogs.msdn.microsoft.com/visualstudio/2018/05/07/int... If an AI-infested text editor isn't a sure sign that the bubble is going to pop soon then I don't know ;)

I think part of this ridiculousness may just be that the common lexicon doesn't have enough words for AI. "AI-assisted Intellisense" at the moment seems to boil down to a marginally novel way of ordering autocomplete suggestions, and like...

It's not wrong? It's neat, it's potentially useful, and it's powered by something under the umbrella of "AI". The problem is that that umbrella is gigantic, and covers everything from the AI system providing routes for the AI system in a self-driving truck on AI-provided schedules for a hypothetical mostly-automated shipping business, to a script I whipped up in ten minutes to teach something 2+2 and literally nothing else.

So we get to the nonsense position where there isn't a better way to describe a minor improvement to what is essentially the ordering of a drop-down list except by comparison to the former example.

Re: AI winter is well on its way

#193

Post by tmalsburg2 »

Stopped reading after the first half. The evidence for the idea that deep learning is failing is that Deep Mind haven't produced anything revolutionary since Alpha Go Zero which was published not even a year ago? And that preformance doesn't scale linearly with the number of parameters? And speculation about why Lecun made a certain career decision? Not very convincing.

Re: AI winter is well on its way

#194

Post by ollin »

halflings wrote: post #151
ozy wrote: post #150

We call it deep learning, but it is deep pattern matching. Extremely useful, but don't expect it to result in thinking machines.

Are our brains magic? If they aren't then surely they must be doing something that we can reproduce. We've built so many things that we considered "thinking machines" in the recent past (realistic speech synthesis, image recognition and captioning, human-level translation, elaborate recommender systems, robust question answering) on "deep pattern recognition".

Brains are not magic, and will be reproduced eventually, but DNNs are a fundamentally weaker architecture and won't be enough. Neural nets can solve some problems that brains can solve easily and lots of other ML methods couldn't solve, which is great. But the space of problems that brains can solve and neural nets can't is still rich, and will remain so until better methods are developed.

Re: AI winter is well on its way

#195

Post by jimothywales »

biswaroop wrote: post #47
bra-ket wrote: post #26

Earlier quoted context omitted.

it's not a stepping stone, if you look closely it's a dead end

I don't see how systematically accurate image classifiers and facial recognition systems built on deep learning is a 'dead end'. Products are products. If deep learning has led to actual profits in actual companies, it's not a dead end. As to whether this leads to AGI is a completely different question.

The point is that the profits may not be as grand as the current level of hype may indicate

Edit: additionally it could be a dead end because the hype tends to narrow the directions we explore with ML. If everyone is obsessing about DL, we could be infuriatingly ignoring other research directions right under our noses.

Re: AI winter is well on its way

#196

Post by kgwgk »

danielbarla wrote: post #138
mastrsushi wrote: post #113

Warning 23 year old CS grad angst ridden post: I'm very sick of the AI hype train. I took a PR class for my last year of college, and they couldn't help but mention it. LG Smart TV ads mention it, Microsoft commercials, my 60 year old tech illiterate Dad. Do any end users really know what it's about? Probably not, nor should that matter, but it's very triggering to see something that was once a big part of CS turned…

Your post has two different points, and I think they should be separated. Yes, there's an AI hype train, and it's pretty tiring. I'm starting to wonder when the first AI-enhanced shoes are going to come out. Or an AI enhanced blockchain, right? That said, the rest of the rant about AI is less solid. Sure, AI today is fairly boring and run of the mill data processing / optimisation stuff (I sort of know, though I only…

> I'm starting to wonder when the first AI-enhanced shoes are going to come out.

Wait no more:

The first connected cycling insole with artificial intelligence: https://www.digitsole.com/store/connected-insoles-cycling-ru...

The World's First Intelligent Sneaker: https://www.kickstarter.com/projects/141658446/digitsole-sma...

Re: AI winter is well on its way

#197

Post by sanxiyn »

nopinsight wrote: post #77
felippee wrote: post #61

Earlier quoted context omitted.

> The difference between the current AI renaissance and the past pre-winter AI ecosystems is the level of economic gain realized by the technology I would argue this is well discounted by level of investment made against the future. I don't think the winter depends on the amount that somebody makes today on AI, rather on how much people are expecting to make in the future. If these don't match, there will be a winter…

Because there is a dearth of experts and a lack of deep technical knowledge among many business people, there are still a great many companies that have not yet started investing in deep learning or AI despite potential profits based on current technology. Non-tech sectors of the economy are probably underinvesting at the moment. This is analogous to the way electricity took decades to realize productivity gains in t…

While I agree there is underinvestment in non-tech sectors, I don't see why that would change and they will use deep learning. There are lots of profitable things in non-tech sectors that can be done with linear regression but not done.

Re: AI winter is well on its way

#198

Post by EliRivers »

batmansmk wrote: post #164
felippee wrote: post #137

Earlier quoted context omitted.

It's not like humanity really needs another chess playing program 20 years after IBM solved that problem (but now utilizing 1000x more compute power). I just find all these game playing contraptions really uninteresting. There are plenty real world problems to be solved of much higher practicality. Moravec's paradox in full glow.

I guess there are reasons why researchers build chess programs: it is easy to compare performance between algorithms. When you can solve chess, you can solve a whole class of decision-making problems. Consider it as the perfect lab.

What is that class of decision-making problems? It's nice to have a machine really good at playing chess, but it's not something I'd pay for. What decision-making problems are there, in the same class, that I'd pay for?

Consider it as the perfect lab.

Seems like a lab so simplified that I'm unconvinced of its general applicability. Perfect knowledge of the situation and a very limited set of valid moves at any one time.

Re: AI winter is well on its way

#199

Post by danielbarla »

johnmoberg wrote: post #173
danielbarla wrote: post #152

Earlier quoted context omitted.

So, Hawking's preposition is that AI can surpass human intelligence. Is this the part you're disagreeing with? It's one thing to disagree, and it's another to call it ignorant. I for one agree that it can / will surpass human intelligence, it's just a question of when.

Unless natural disaster and/or war resets humanity, it seems inevitable. One only needs to concede that there is (probably) nothing special about the biological nature of the brain.

That's my reasoning as well, regarding the biological part. On the one hand, it seems to be an incredibly energy-efficient solution, but has several detractors as well (in a lighthearted fashion: lack of scalability and high-bandwidth networking, several billion years of legacy code, etc).

On the other hand, it's unclear whether computers will continue to get faster in the fashion that they are today for long enough. If not, then huge efficiency improvements would be necessary on the software side, though those do seem to be happening as well.

Re: AI winter is well on its way

#200

Post by SomethingStars »

Grue3 wrote: post #153
ehsankia wrote: post #88

Earlier quoted context omitted.

How can something be biased when it's listing facts? Those are actual features that are available today to anyone, that were made possible by AI. Do you think it would be possible to type "pictures of me at the beach with my dog" without AI in such as short time frame? Or to have cars that drive themselves without a driver? These are concrete benefits of machine learning, I don't understand how that's biased.

Can't speak for the other items, but these >translate over 100 languages in Google Translate; >caption over a billion videos in 10 languages on YouTube; barely even work. Yeah, it's a difficult problem but it's not even close to being solved.

YouTube captioning in English works surprisingly well, the improvement over the last few years is huge. It still chokes on proper nouns but in general it mostly works.
Post reply on HN