Hacker News new | comments | show | ask | jobs | submit login
YC AI (ycombinator.com)
663 points by craigcannon 189 days ago | hide | past | web | 219 comments | favorite

If you're the kind of person that's interested in taking up this challenge, but you currently have the coding skills without the deep learning skills, we built something that can equip you with most of the current best practices in deep learning in <2 months: http://course.fast.ai/ . It doesn't assume anything beyond high school math, but it doesn't dumb anything down (key mathematical tools are introduced when required, using a "code first" approach).

We don't charge anything for the course and there are no ads - it's a key part of our mission so we give it to everyone for no charge: http://www.fast.ai/about/

And yes, it does work. We have graduates who are now in the last round of applications for the Google Brain Residency, who are moving into deep learning PhDs, who have got jobs as deep learning practitioners in the bay area, etc: http://course.fast.ai/testimonials.html . Any time you get stuck, there's an extremely active community forum with lots of folks who will do their best to help you out: http://forums.fast.ai/ .

(Sorry for the blatantly self-promotional post, but if you're reading this thread you're probably exactly the kind of person we're trying to help.)

This course was discussed here on HN 3 months ago: https://news.ycombinator.com/item?id=13224588

thrawy45678: I see a lot of criticism about tmux and other non-core items being included in the overall curriculum. I think the author is trying to portray the workflow he is currently on and exposing the full tool kit he uses. I don't think he is saying - this is "THE" approach one has to follow.

derekmcloughlin: If you've no experience with ML stuff, you might want to start with Andrew Ng's course [...] Paperspace (https://www.paperspace.com/ml) have ML instances starting at $0.27 per hour.

webmaven: Any recommendations for the cheapest (but not "penny wise, pound foolish") HW setup that meets these requirements for course completion? - (answers: https://news.ycombinator.com/item?id=13227437 )

The course is excellent, and thank you for making it and offering it for free - but a word of caution for those considering following it: Along the way you will incur not-insignificant costs for Amazon EC2 GPU instances, and, even if your instance is shut down, SSD EBS storage costs.

Edit: To be clear, I'm not suggesting it's not worth it, just highlighting that theres's more than a time commitment to budget for.

Yep, the course is free, but you'll need to pay for the computing power one way or another.

If you have a workstation with a fairly recent Nvidia GPU (I used a GTX 980 TI) and bunch of spare disk space, you don't need AWS at all. You'll still pay for the electricity, of course, but it's not what I would call a significant extra cost. That is, if you already have the hardware.

What's the minimum nVidia GPU spec (VRAM) and SSD/HDD disc space (TB) required, from your perspective?

Get a GTX 1070 at a minimum. GTX 1080 Ti if you have a bit more cash to spend. Disk space is entirely dependent on your data set size. We're not talking huge numbers for most things you would be doing

I have GTX 970 and how much spare space do we need?

btw, I have a doubt, the course has specific video to setup aws, for local machines, what to do? should I install the required packages python, keras or whatever they use?

The install-gpu.sh script found in fastai's github can be used to set up the required environment on an Ubuntu box.

A GTX 970 should do. The exercises are designed for GPUs with more memory, so you'll need to use smaller batch sizes here and there.

To give a more exact estimate - if you don't use spot instances ($0.20/hour) it'll cost $0.90/hour, and if you do the suggested 70 hours of work for the course, that's $63. And then there's around $6/month cost for the EBS volume.

Not exactly.

The time you spend on the course, and the time spent by your instance running workloads are two completely separate things.

Your cost per-hour is accurate (plus a few bucks for an IP address), but the number of hours is off by an order of magnitude.

Also, I'm paying double that per instance per month for SSD using the provided scripts to build the instances. It's the smallest part of the cost, but I mention it because it can take newcomers to EC2 by surprise if they have an instance shut-down but consuming disk space.

I'm going through this course, but using Google cloud instead of AWS. I can confirm that at least the first Jupyter notebook works well for me.

I had to adapt the aws-install.sh script, but it was easy enough. I ended up using a snapshot instead of a persistent volume, as the monthly cost when you're not running is much cheaper. So I have a script to create a new instance and then restore that snapshot. It's faster than installing the dependencies each time).

Google Cloud has a $300 free trial for new users: https://cloud.google.com/free/

(Disclosure: I work for Google, but have no involvement in Google Cloud other than being a happy user)

Any chance you can share the adapted scripts? I assume they are easy to make, but the less barriers to entry the better.


1) Create the instance the first time: https://gist.github.com/rahimnathwani/40ed3b6f496d377e3d17de...

2) SSH into the instance and run this script (replace anaconda mirror if your instance isn't in Asia, and definitely set the password hash): https://gist.github.com/rahimnathwani/b63ebc5900b832743d9e22...

3) After you've tested the server, snapshot the disk (using the web console) and destroy the instance and the persistent disk.

4) Run this script to recreate the instance and disk from the snapshot (replace with your snapshot name): https://gist.github.com/rahimnathwani/9019df33d4ad7ec4607413...

Buy a laptop with a GPU. If you are serious about learning AI, it's worth spending few hundred dollars to get a proper workstation.

I heard that nowadays it's possible to buy a laptop and a external GPU?

I have a mac, is it possible to really buy an external GPU??

Yes, but GPUs should only be consider for acceleration when there is insufficient local CPU power to accomplish something valuable. Otherwise, it's like buying a Ferrari to get groceries.




AKiTio Node | Thunderbolt3 External PCIe Box for GPUs https://www.amazon.com/dp/B06WD8KS52

Gigabyte GeForce GTX 1080 XTREME Gaming Premium Pack Video Card (GV-N1080XTREME-8GD Premium Pack) https://www.amazon.com/dp/B01HHC9Q3U

There's been some work on using AWS spot instances for the course, which can save quite a bit:



Like how much? If you are going to get into deep learning for real seems like it might be worth building a multi-GPU workstation?

Don't buy equipment before you have demonstrated a real need for it.

This applies across basically all of life, and it's so frustrating to see people ignoring it, because what ends up happening is they use a string of 'gonnas' to justify buying stuff they don't need. Gonna get fit - buy $1500 worth of gym gear. Gonna learn electronics - buy oscilloscope, power supplies, tons of components. Gonna get your motorcycle license - buy brand new bike and stick it in the garage.

If you have a desktop computer, you're good to start. When you've done enough that your available CPU/GPU is limiting you on your own projects (not on something you pulled off github) then you can look at upgrading.


>Gonna learn electronics - buy oscilloscope

A fairly accomplished electronic engineer told me that they'd never once solved a problem using an oscilloscope, but that it helped to keep them occupied while they were mulling over what might have gone wrong. (That's presumably why the better ones have so many knobs and dials to play with, like one of those children's toys.)

I've certainly solved problems with a storage scope before, but not for a long time, and they were mostly software problems rather than hardware problems (ie. using it as a poor man's logic analyzer to infer what's going on with the code via a couple of spare IO pins). I really kinda want one though.

There is actually an accepted term for what you describe in many circles, called GAS -- Gear Acquisition Syndrome.

One forum user put together a <$700 complete PC with an adequate GPU. You definitely don't need >1 GPU.

Or you can use spot instances, as a sibling comment mentions - about $0.20/hour generally.

Ah, cool. I am familiar with AWS and spot instances, but after that comment was assuming it required multiple instances for training or something.

Build a fast deep learning machine for under $1K https://news.ycombinator.com/item?id=13605222 a month ago with many tweaks recommended in the comments.

gcp: Contrary to the claims there, the CPU does matter

dsacco: yes, this will work, it will quickly become suboptimal [...] as you scale your hobby into something resembling more professional work

brudgers: I'd start with a used Dell Precision T7xxx series off of Ebay for a <$300 including RAM and a Xeon or two.

croon: If the limit is a firm $1000, I would get something like this: https://pcpartpicker.com/list/XHV9Fd

I'd consider $60 bucks to pick up deep learning skills an "insignificant cost".

Well, that depends on how you define your cost function. :-)

"Along the way you will incur not-insignificant costs for Amazon EC2 GPU instances"

Is there some particular constraint involved that I cannot run the computations at home?

Heat death of the universe.

Buy a pc/laptop with an NVIDIA GPU and setup Linux on it. It is a better option than using EC2 instances.

How much VRAM is required?

While I have found it quite easy to take a code-first approach to deep learning/machine learning, I have encountered a lot of scepticism from existing ML/data science practitioners about such an approach, and I feel like investors will be even more risk-resistant.

Funnily enough big tech companies have seemed the most willing to accept people making a switch. I'm guessing because their appetite for ML/DL people is currently unquenchable.

Hard to say. Universities aren't very good at instilling good applied engineering. If you are 18, very high GPA and you've passed this course and have strong programming skills, I think some investors would be interested. Similarly, if you have say a physics degree from a good university and you've done this I think investors would be interested.

If you're 22+ with a high school degree and not much of an engineering resume... Yeah, your pitch deck better be good.

Besides graduating from your fast.ai course, what were the other qualifications of those Google Brain applicants? I'm imagining they would have, or be in the process of getting, an MA/PHD in non-AI-related area.

No, the person I know who is in the last round is an economics major with no Masters/PhD. She has a very impressive background in industry.

Although a lot of folks in the course do indeed have graduate degrees in other fields (including English Lit, Neuroscience, Radiology, etc...)

I am sorry if I sound rude or naive or both.

But I really do not understand why we hate ads. I have seen many tutorials which are given out for free. Of course, I am grateful that you decided to offer the course for free of charge, but I really would not mind a little adverts just to get you as the creator/maintainer some $$.

I am a self published author of a decent intro to web development book in Go language, it is an example driven tutorial/ebook and while the main book is open source on Github, there is a leanpub.com version which I offer for variable pricing, 0$ to 20$, and it has been working great for me, rather than getting nothing for the tutorial, I am getting something.

Without ruining the current tutorial, there are ways of getting something from it, of course.

My $.02 -- if I were to provide a resource with the goal of being of great value to people, and it's within my wherewithal to maintain it with what I make elsewhere, the $$ return from ads are far too low to justify how much they detract from the experience I set out to provide. Ads aren't the absolute worst, but I think we can agree they are on average negative to the experience you visit a non-shopping page for.

For a fairly niche resource such as this (it'll never reach a "how do I get a boyfriend" level of audience), it's unlikely to ever draw as much ad revenue to pay for itself. To do so, a high quality, specialized ad system would need to be deployed, which honestly becomes a high touch deployment and maintenance project, that isn't directly tied to the core goal of just providing an awesome resource publicly for the greater good, which is just distracting (or costly) for whoever is behind it.

I appreciate the mature choice to not try to gain small change and instead eat the cost for hosting and development to feel good that you are providing something not just great, but unadulterated as well.

True, what made me write it was a YouTube channel I came across, they don't accept donations or show ads on their channel and they have crazy views, they say that we don't want to earn money on this, granted that they give it away for free, but it isn't "evil" to get some money out of it, I am not saying rip off students by charging 10000$ per session, but providing something like a PDF version of the online guide for a small amount say $5 would, in the long term, give you some return.

That's passive income which you don't have to bother about, like you'd have to bother about ad deployment. We as an industry are funny, we expect everything to be free of cost _and_ the author should not monetize it in any way possible, it isn't evil to monetize that's what I am saying.

Ads are basically a tool for psychological manipulation. It's unfortunate that this is the only practical method of monetization for some creators. Micropayments in future may help with this. To me, ads do feel disrespectful of my audience.

While ads might be that, but all I wanted to say is that it could be monetized somehow like a pdf on leanpub for 0-20$ pay what you want, I would have gladly paid.

I'm pretty new to AI -- would it be better to do this course or the machine learning course by Andrew Ng on Coursera?

The Coursera course is "machine learning", which is a more general terms, while this course (based on the description) is "deep learning", which is more specific - the ML course ends with Neural Networks, while this one starts with them.

I'd start with the Coursera one, if only to learn when _not_ to use a neural network and use something simpler. But if you already know how to cluster data with K-means, what a linear classifier is, and/or what an SVM is, then you can probably skip the ML one.

Could you add a simple description of the practical usecase for the lessons? I know what "image recognition" is useful for. I have no idea whether I need to or want to learn "CNN", "overfitting" "embedding" "NLP" or "RNN". I am interested mostly in image recognition and text classification.

Thanks for putting that up. I'm gonna try to find the time, since I'm hoping to do a PhD someday.

Thanks, I look forward to going through this course.

I wanted to point out though, that on the main page http://course.fast.ai/ the courses appear out of order; 1, 2, 0, 3, ... instead of 0, 1, 2, 3, ...

Some learning resources:

Learning AI Advice from Open AI, Facebook AI leaders


Thanks! Thanks, thanks, thanks a ton for what you are doing. As someone without a strong math background, it is kinda hard to enter the field, so I'm going to take 7 weeks now and hope I can understand deep learning better than I do now.

Checkout https://www.floydhub.com as a replacement for AWS!

This might be exactly what I have been looking for... thanks!

I wonder if the AI/machine learning revolution will open new jobs for people coming from IT admin roles.

Came here expecting to see this mentioned somewhere, was not disappointed :)

Thankyou, this looks to be exactly what I need! Starting the course now.

Thanks so much for highlighting this!

thank you .. your course was really helpful.

Thank you

I love your project but I totally detest your landing page with the huge animated panels that sit there eating up cycles that could be used for something better. I'm sure it's great for conversion and all that but that doesn't stop me from being irritated.

If the course is free and you're doing this to improve the world what's the point of using such tactics? At least shut down the rotation after one or two iterations.

"that sit there eating up cycles"

Sure, about as much as rendering a single glyph of the entire text on the page.

Kidding - I have no idea but if it's just panning it that should be the ball park for the necessary compute load.

As someone who's capable of implementing and understanding many of the most fashionable tools in AI, I don't know what to do with the current economy. I think there is far too much attention being paid to the pie-in-the-sky research, and even though wealthy investors think the things that they don't understand and I do are capable of accomplishing things that they think can be accomplished and I don't, the problem is that they want to pay people like me to chase their dream. And while I do love money, I also love the idea of living a meaningful and fulfilling existence by pursing technologies that will advance mankind.

Can other people who've actually seen real promise in their AI research chime in and help convince me that we are actually on the precipice of something meaningful? Not just more classification problems and leveraging the use of more advanced hardware to do more complicated tasks.

I don't think we are on the precipice of general AI, perhaps we are not even close.

But we don't need fully general AI to have a huge transformative impact on society and the economy. Think about the full spectrum of jobs that exist in the modern world: doctor, teacher, construction worker, truck driver, chef, waiter, hair dresser, and so on. How many of those jobs actually require the full power of human intelligence? In my view, almost none. Maybe the intelligence required is still more general than what a computer can do, but probably a determined research effort could be enough to make up the gap for any specific task (say, to cut someone's hair, you need to have good visual understanding of the shape of the head, and very good scissor-manipulation skills, and a good UI so that the customer can select a style. It hardly seems like an insurmountable technical challenge).

Maybe a pair of scissors near a head are not the best example of imperfect AI being applied in a great way.

I think you vastly underestimate the creative problem solving skills that even "unskilled" humans are capable of, which is required to cut hair.

> I don't think we are on the precipice of general AI, perhaps we are not even close.

In all probability, I want to agree with you.

However, deep learning unnerves me. Seriously, how many people who evolved the deep-learning-trained models actually understand how those models work? But the models produce good-to-great results.

So there could be a small chance that general AI could randomly and spontaneously come into existence. Either we are too close to general AI or infinitely far from it. We can't tell it because we are not sure if intelligence is a evolved goal oriented function or an engineered substance. Being an atheist, I believe the first part - and thats what deep-learning based AI looks like.

A major difference here is that humans don't have user defined stimuli.

Human intelligence evolved because of the need to survive which evolved because of the need for genetic replication. Our need to survive lead to a nervous system which created the need for a central management platform (the brain).

People do understand the deep-learning models they create. They're based on a user defined limit for error which is the mathematical distance between what is and what is not.

I think that until we have an algorithm that can rewrite itself optimizing for existence (bug-less-ness and a continuous power supply?), we won't even scratch the surface of general AI.

Hi kolbe, have you seen this research paper: http://www.nature.com/nature/journal/v533/n7602/full/nature1...

They train a model that translates signals from the motor cortex to directly stimulate the patient's forearm muscles. It is the first case of a quadriplegic man being able to control his arm directly with his mind.

That is really cool. Just like it was really cool when we learned how to translate neural signals farther away from the motor cortex to do similar things a decade ago. But I'm failing to see the evidence of revolutionary AI. I see a really well-researched piece of hardware paired with a solid understanding of anatomy and probably a few great data scientists to assist in it all.

And the consensus in the Neuroscience community, from what I understand, is that this is two things (1) a great feat of engineering and (2) a great demonstration of the brain's ability to learn and adapt.

Given the OpenAI initiative and the explicit intention of "democratizing" AI, it seems they're interested in funding application, not more research.

I think the results seen from deep learning along with recent developments in commodity hardware(GPUs have made major leaps in the past two years) to run them in a reasonable amount of time has kicked off a frenzy. IMO it does seem a little overhyped in the same way the internet was -- its a tool that enables ideas beyond our imagination, so VCs throw money at companies on a dart board hoping one takes off. In the end, the internet was a revolution but it wasn't as quick as everyone thought.

TL;DR - AI is becoming a tool that can be used by anyone.

>> AI is becoming a tool that can be used by anyone.

The fact that I can do it should be evidence enough for that statement. ;)

As far as the bulk of your comment goes, maybe. That's a prediction, but I can't say that I feel especially confident in the prospects of it. But who knows. Cold fusion looked like a lock for a revolution 60 years ago, and social media looked like a fad a decade ago.

Robotic technology has improved greatly over the last few decades. Just over the last decade, smartphones have created a bunch of innovation in sensors, mobile processors, batteries, etc. And reduced the cost of those things greatly.

But robots haven't really proliferated outside of narrow applications. The main thing holding them back is they are blind and dumb. They can only do very simple rote tasks over and over again. They can't identify objects in an image. They can't figure out how to pick something up. They can't learn from trial and error.

ML has improved massively over the last few years. It's now possible to do machine vision at (arguably) super-human levels on certain tasks. When you see videos of reinforcement learners learning to play video games; that's not a huge stretch from controlling a robot. There have been similar jumps in ability in many other areas. From machine translation to speech recognition.

Sure, there may be lots of over-inflated expectations. But I think a vast increase in automation is well within the limits of what is possible in the near future. You may not get Jarvis. But we will get robots that can understand simple orders and learn to do simple tasks.

That's huge. That's replacing most human jobs huge. It will completely change our economy and our way of life.

If you're worried AI is having no impact at all, some of the places I can think of where it's already had an impact are Google Search, Palantir, and driver assistance (as in Tesla's Autopilot).

So the tools we have now aren't total frauds, and there's probably more to be discovered.

What all three of those examples have in common is that they involve an AI assisting human decision-making. I think that this will be the most lucrative area in the future as well. If you have a set of data a human can't even look at, like a million web pages, the AI doesn't have to be that good at processing it to be useful, since it has no competition from humans. On the other hand, it's rare that you ever need to come up with a million new concepts - you probably just need one. So humans easily outcompete AI at coming up with new ideas. Maybe it's a bit vague what "new ideas" means, but certainly if you're capable of generating new ideas, then you're at least generating turing-complete outputs. Nobody is even trying to do that right now. In the case of image recognition, there are some complex ideas needed to make it work, but at the end you're just outputting a label, getting your model to use the most basic DSL possible.

Palantir doesn't do any machine learning (or AI).

Care to elaborate?

Source? Are they lying?

They are actually very public about the fact they are not a machine learning company and are more or less opposed to machine learning. Their goal has always been to wrangle data and make it useful to human decision makers. not automate decision making.

You don't think that progress in both qualification, vision, and robot control can lead to robots that can replace people in tasks like packaging, assembling, painting or cooking?

What would you say is the missing piece for a robot that knows how to wrap a football for amazon, (and knows what do if the ball falls down, and if there are people around?)

> We want to level the playing field for startups to ensure that innovation doesn’t get locked up in large companies like Google or Facebook.

AI and ML are exciting because they promise to help us evolve systems and machines quickly to perform more accurately.

This requires access to a constant flow of large, proprietary datasets.

Providing cheap access to datacenters and compute power is a great first step for leveling the playing field for startups.

I'll be interested to see how YC tackles the (IMO) more important problem of providing access to the data needed to train models.

I think IBM has taken an extremely wise first step by acquiring the data assets of The Weather Company. This will give its Watson IoT portfolio a leg up on other companies that need to rely on a patchwork of public and private data sources when factoring something as integral as the weather into algorithms and logic engines.

Perhaps YC can consider something similar, pooling investors and VCs together to acquire/partner with providers of essential data.

> We want to level the playing field for startups to ensure that innovation doesn’t get locked up in large companies like Google or Facebook.

It's not just the data. It's also acquisition: Facebook and Google are figuring very large in the exit path of a lot of these companies.

Which means that after one or more years the stated goal here could be nixed. You'd almost have to close that route before this statement is meaningful.

The two companies are like two giant vacuum cleaners sucking competition out of the technology industry.

Taken in isolation, the two trends (data acquisition and talent acquisition) are already disturbing. Combine the two, and you have to wonder how dumb we actually are to allow things like Facebook's acquisition of WhatsApp.

Isn't most weather data gathered by the government?

IBM also bought Blekko, a general purpose web search emgine like Google and Bing. They use it to harvest data to build something like Freebase internally, now that Googe closed down Freebase (to ise it as well internally only). https://en.wikipedia.org/wiki/Blekko

So it would definitely help smaller AI companies if an open minded consortium would launch a new FreeBase alike service and open web crawler with free data dump access. (WikiData is at the moment (and for the next months) several magnitudes smaller, and Freebase frozen stale data gets more useless every day)

When I first moved to SF I basically was building a FreeBase competitor for entity resolution and ontology building. I think opening up another service like this is required for the future. :)

Completely agree. It's the (lack of) data that's the issue, not exposure to technology.

reinforcement learning, the most promising approach for robotics today, doesn't require a lot of data to train. Or rather the data required to train it can be synthesized from the problem domain itself. Access to compute and hardware, possibly, but that is it.

>> RFS: Robot Factories.

From all the places where AI could help, why focus on that field, the place where the most vulnerable employees are found, the hundreds of millions from china, Bangladesh, etc - who have little chance of having a meaningful Social safety net ?

>> job re-training, which will be a big part of the shift.

I don't believe that this is realistic, even in the west. Why ? because the internet, who is obsessed about this subject, and doesn't lack imagination, can't even supply a decent list of what jobs will the future offer that could employ the masses who's jobs will be automated.

>> job re-training, which will be a big part of the shift.

To me there's no other way to look at this other than having faith in our ability to find work to do. Wherever there is deficiency in technology, humans will find a way to fill in. Now the question is whether this will be at the same scale, and will there always be gaps to fill in. No one knows. We never have known. We've just blindly innovated and had faith. It's worked so far. Until it doesn't we will probably just continue doing it.

In other words, this is and has always been unplannable. An example. When DARPA invented the internet, no one explicitly said "We're going to put a bunch of mail staff, brick and mortar commerce employees, etc, out of work so we better ensure we train programmers to help fill the void."

I'm with you though; it concerns me.

>> When DARPA invented the internet..b&m employees.. programmers

As for being less brick-and-mortar employees ? i think that was an historical trend, so it makes sense it will continue, supported by tech.

Also It's somewhere between 1960 and 1970(darpa and the internet), i think. But Moore made his law in 1965. So we guess that in 2017 computers are extremely abundant and fast, connected at high speeds(moore + internet), and we know they are versatile.

So we can guess they'll have many uses(and a lot of room for imagining), and maybe we'll need many programmers for that.

furthermore , see my comment on predicting jobs, i don't think it's that hard(in general) :https://news.ycombinator.com/reply?id=13910181&goto=threads%...

It concerns me as well.

But jobless people are a resource that entrepreneurs will find a way to utilize for profit. It's not that the technology will require new kinds of work, but that idle people will enable new kinds of companies.

I think that Uber/Lyft/Prime Now/etc were enable because of high unemployment / low participation rates.

Factory floor robots will be of use for economies where their populations are in decline, Japan, and soon China, much of Europe. The pace of adoption; as you imply, will negatively impact the workforce --China being China, though, I can see them doing this in a more methodical way, they do not wish to upset their population lest they unwittingly foment unrest.

That said, despite whatever you think, VCs want return on investment and they do not really care about who and what they replace --for example, environmental concerns, they do their PR thing but they will still build/invest in projects which will develop in places with less stringent regulation effectively undermining their domestic posture on environmental issues as well as their posture vis a vis respect for the working person. They just want money and influence --but to feel conscientiously at ease they may engage in political issues that locally will appear to be progressive.

The number of industrial robots in Japan is actually dropping, whereas China is hockey-stick. (Source: UN report[0] released October 2016)[1]

I visited two Chinese factories this week, one for injection-molded plastics and one for metal. Both were essentially collections of very expensive automated machinery such as 5-axis steel boring CNC controllers and high powered laser cutters, with a few gantries, finishing/packaging and storage areas. There were as many people on computers working with 3D models as there were standing on the factory floor. In the latter case, the primary interface was usually a keyboard and mouse or touchscreen.

[0] http://unctad.org/en/PublicationsLibrary/presspb2016d6_en.pd... (page 3, citing UNCTAD secretariat calculations, based on International Federation of Robotics, 2015, World Robotics 2015: Industrial Robots, available at http://www.ifr.org/industrial-robots/statistics/ (accessed 19 October 2016))

[1] https://news.ycombinator.com/item?id=12991293

The bit about Japanese robot numbers going down is interesting, are factories producing less or are they replacing prev gen robots with newer ones capable of more output (more efficient)?

Japanese GDP is declining and hasn't grown much at all in real terms since the 1990's[1], so... maybe producing less and not replacing with more capable robots?

[1] http://www.tradingeconomics.com/japan/gdp

While it's true that there have been, are, and will continue to be a loss of manufacturing jobs as a result of automation, robotics, and AI, we also need to consider the other side of the economic spectrum, and look historically to see what has happened when these kinds of shifts come.

For example, I'm sitting next to someone who is working on robotic apple-pickers. The amount of cost from buying apples is about 75% represented by labor. So yes, there would be job loss, but what if apples start to cost 25% of what they currently do? What if we extrapolate that across all fields?

What if we could do this with everything we need - food, consumer goods, housing, etc? It's a shift as big as the industrial revolution and moving away from an agricultural society from a consumer perspective. In short: Life would get much better and cheaper for all of us and fast.

But then there's unemployment. What will the masses do? Well, what did the masses that used to be farmers do when we moved from 50% of the United States being farmers to the current levels of ~2%?

The answer historically has been that entirely new types of employment are created, and the pain is remarkably short-lived considering the size of the shift and the decrease in costs, and the Luddites look silly in retrospect. It's usually hard to see looking forward, but somehow it's always worked out.

I'm not sure what the entire picture looks like; I do think that we will need massive re-education as a society, and we're completely unprepared for that. In the past we've replaced unskilled labor with new unskilled roles, and I don't see those roles hanging around anymore. (I moved to San Francisco from a very rural town, and I get stressed whenever I think of the economic future of that small town. Until there's risk-free education available at will that area is economically doomed.

But I don't think the proper response is to slow innovation because of pending job loss. If the AI shift will be as big of a deal as the Knternet, imagine how much better it makes the lives of everyone - is that really something we should actively fight against? I'd argue history suggests we shouldn't.

> In the past we've replaced unskilled labor with new unskilled roles, and I don't see those roles hanging around anymore.

The sad reality is that there's a nontrivial chunk of the populace that isn't able to pick up highly skilled roles. It also ignores the role of unskilled jobs in providing space for people whose job class has been destroyed and need to retrain (or mark time until retirement).

I'm not advocating slowing innovation to prevent job loss. I am advocating avoiding magic thinking ('there's always new jobs to go to'): we need to start a serious conversation about what we do with our society when we have the levels of unemployment we can expect in an AI-shifted world. Right now we're trending much more towards dystopia than utopia.

ahem basic income

Rich economies could (barely?) afford that, how could mismanaged economies afford a basic income given they don't have much in the way of economic might to afford a basic income. Think Indonesia, Bangladesh.

I keep pointing out that Bucky Fuller calculated that we could pretty much take care of everyone on Earth while working ten hours a week if we just applied the technology we have in an efficient manner. (By the mid-1970's he claimed.)

>> but what if apples start to cost 25% of what they currently do? What if we extrapolate that across all fields? ... do this with everything we need...

That's great, as long as there are new jobs, or a decent safety net(which i doubt).

>> The answer historically has been that entirely new types of employment are created, ... It's usually hard to see looking forward, but somehow it's always worked out.

It's not always worked out - remember horses ? and what about the terrible transition period ? but let's leave that.Is it really that hard to see looking forward ? That's a more interesting question:

So before the industrial revolution, rich people bought a lot and a variety of clothes/foods/things/services(doctor/lawyer/cook/traveling/transportation), large houses, etc . So assuming things would get cheaper, they'll be more available , you could imagine that regular people will want those things,a lot of them, but maybe with even more variety(more people). So you could imagine a consumer culture.

And the same for the business side: businesses wanted advertising, transportation, financial services, manufacturing, knowledge, etc.. So if all those things will become cheaper - and better, you could assume they would want more and more of them.

And furthermore - before the industrial revolution there we're many problems - the world wasn't ideal(you got sick, life we're boring). In general , we can assume many of those problems will get solved, so they'll create new demand.

So so many new jobs! And we know all that - we've seen it happen!

So why the hell we cannot think of new jobs ?

> Why can't we think of new jobs?

My job didn't exist 50 years ago. Did yours?

Which job, existing or new, cannot be automated?

Cross-cultural educationalist philosopher.

Well, what did the masses that used to be farmers do when we moved from 50% of the United States being farmers to the current levels of ~2%?

Didn't that work in the opposite way, though? New opportunities appeared that pulled (sometimes forced) people away from farming; it wasn't like there was a bunch of unemployed ex-farmers who suddenly started finding new jobs.

Which time period in history? In the 1830s, you'd get younger sons or daughters of a farming family who would move away to the city to earn extra spending money for their family and find a measure of independence; oftentimes the eldest son would inherit the farm, so there was no place there for the other children.

In the 1930s, the farms were themselves collapsing. Mechanized agriculture created both a huge oversupply of produce (which drove down prices) and also ruined the ecology of the plains, which eventually led to the dust bowl. Farmers absolutely were forced off their land: that's where we got Okies, Hobos, the Great Migration, Grapes of Wrath, and all those other subcultures of migrant workers from.

I thought that's exactly what it was. Dustbowl -> Great Depression -> New Deal -> WWII -> Industrial Revolution takes off

What do you suggest they focus on then? I think it's a good thing. That change is coming one way or another, so why not own it?

Considering that UC is also funding Basic Income research, I think they'd be one of the better people to be automating stuff.

The legal and healthcare industries being disrupted by AI would be preferable. That's where a lot of inefficiency exists. Both of those are making it very hard for "normal people" to afford legal help or good medical care, due to how expensive they can get. If a single lawyer could use an AI to do 95% of his or her work, I imagine that would drastically cut his or her total fees.

Doctors are overworked as is and won't have their jobs replaced, they will only benefit. And doctors don't deserve to lose their jobs anyways. Lawyers on the other hand...

I've also become cynical in this regard - "better in the long run" often means that entire generations of people will suffer, which is a huge societal impact in the end. It's even scarier when you consider this new industrial revolution will need even fewer people, which means wealth will be redistributed even less, and will probably bias those who already have access to capital to fund such ventures. More and more I'm convinced that we owe it to ourselves to provide a safety net for every person that allows them to live a decent life without worry for basic needs - UBI or otherwise.

Here's an article about that idea from the POV of startups, though I think it's more generally just a moral thing we should do: https://www.theatlantic.com/business/archive/2012/02/the-ent...

Reducing the cost of manufactured goods helps the poor as much as the rich.

For instance: Reducing the cost of farm equipment could enable poor rural farmers to rise above subsistence farming. I could think of countless examples.

In fact, reducing the cost of manufactured goods helps the poor more than the rich in some sense, since a marginal reduction in cost of living for them is much more impactful than for a rich person.

Think about it from a revenue standpoint. It doesn't matter if your cost of living becomes 50% cheaper if you lose 100% of your income.

If 10 people's cost of living drop 50%, then one drop of 100% of income would be acceptable.

I guess YC is looking at this from the perspective that it would prefer if its startups wouldn't have to deal with manufacturing issues in China, when the entrepreneurs could build everything they need in robot-powered factories in the United States.

I'm not agreeing with them, just saying that I think this is where they are coming from.

Surely YC isn't to blame for the 'vulnerable employees' you speak of?

Why attack YC instead of the ones exploiting those employees, or even the systems that allow exploitation to occur in the first place?

But no, by all means, lock those 'vulnerable employees' into a job that they will be treated as sub-human for the rest of their lives, all in the name of 'job security' (or whatever you think you're preserving).

Because YC is both the one advancing AI work and proposing this "solution".

If the problem is truly exploitation, then let's attack exploitation.

Because aside from exploitation, there is another very real problem of replacement. Who said there's only one problem to be attacked at a time?

The point wasn't that this employees have terrible jobs. Some sure do, some doesn't.

The point is that if factory work is automated - they'll have no other possible job, and no safety net, unlike the west. Thus they are vulnerable.

>if factory work is automated - they'll have no other possible job, and no safety net

I know, and it's morally disgusting and I wish systems like this didn't exist in the first place.

What I'm saying is that there is no future in which automation does not replace (most) human labor, so if the goal is ROI, then why would YC ever invest in something other than that? It would be delaying the inevitable, would it not?

Let's say YC provides a UBI to all of those displaced workers; now what? A bunch of unemployed are (still) surviving in the same fucked up system as before and the problem still exists.

To somehow pin the problems of the system on YC's investments is ignoring the larger, more pressing problems of socioeconomic inequality and exploitation as a whole.

I've wanted to break into robotics from machine learning for a long time, but I haven't found a good entry point for the problem. It seems like a rather large vertical cliff I have no way of scaling.

One of the handholds I'd need is a physics engine which which models a robotic arm down to the finest levels of motor control and feedback. I am not a mechanical engineer, and I do not know where to begin with a problem like that. I don't imagine this is hard to build using existing physics engines, but it wouldn't work out the box. It requires some deep domain knowledge of things like friction, tensile strength, and so on. A system like this would spur progress in robotics immensely.

Have you looked into ROS and Gazebo?

I am an undergraduate leading an open source autonomous vehicle project [0] and we're currently almost done setting up simulation testing. Gazebo and ROS are very poorly documented but they serve their purpose really well when it comes to creating pub sub and physics simulation for robotics systems.

[0] https://github.com/gtagency/buzzmobile

It doesn't have to be an arm; for example Kiva, the robotics company Amazon bought for $775 million in 2012, just found a better way to move parts around in a warehouse; pretty basic idea and system too.

If you or for that matter anyone want a cofounder to apply to YC for AI, happy to help.

Would be helpful to have your email here or in your profile

Thanks, I only do outbound contact; if anyone wants me to contacts them, simply let me know and I'll be in touch with 24-hours.

My email is in my profile, pls reach out

contact me, I am interested

Why do you think you need a physics engine (outside of simulation/testing)? Don't humans get by pretty well with only an intuitive one?

Human motor planning involves, at least, forward and inverse models for performing stochastic optimal control (at least, that's the standard formulation of the problem). An intuitive physics engine may be built into our heads, but whether or not it is, just running it forward doesn't seem to suffice for motor control.

it is for simulation, as you say - but this is a critical for employing reinforcement learning. The internal (inverse) model will of course, be based on a simplified system, like a DNN

Maybe or maybe you could just model what the workers already do, predict what they should do, and dynamically provide feedback, simulation, etc.

im not quite sure what you're suggesting. there needs to be a source of ground truth to inform the robot if what is being done is right. And actually having a physical robot arm for this would be way too slow

Right, the ground truth is the existing human labor - and yes, run sims at scale makes sense, but if you don't build and model using the real world you'll rapidly get a solution that only shows the sims errors in modeling the world. All the best robots I've seen run sims, but very narrowly and with a well defined problem to refine via a well designed model.

Offer stands to help you, but given the deadline for the next batch is days away, you would need to let me know as soon as possible via a comment, then I'd contact you via email.

shoot me an email with your background etc. I'm not aiming for this batch actively but i wouldn't mind a chat. Google my name to get my website an email. it should be the first result

How exactly does this "democratize" AI? Doesn't this only potentially prop up another AI company with the hope that YC will be backing it (and thus profit from its success)?

It just seems like "democratize" is one of those new startup buzzwords. But yes, it seems like really what they want to do is bring it to market, and by democratize, they mean have everyone buy it.

Yep, No one was taking "disrupt" serious anymore, so they needed a new word to strip from its meaning.

It's official. HN has disturbed democratization.

Did you mean 'disrupt'? Hadn't heard of "disturb" before used in startup lingo.

I like it.

"How This Startup Is Disturbing The Equipment Financing Market"

"A startup is disturbing the consulting industry"

"16 Startups Poised to Disturb the Education Market"

Woops, meant to say disrupt of course.

I'm not a buzzword historian but I'm fairly sure "democratize" has been around for a while: https://en.wikipedia.org/wiki/Democratization_of_technology

Wouldn't a "third AI company" be a net positive for society? Especially if it wasn't tied to a large network of other interests/products like the existing ones are? The biggest risk right now is that a select few companies are buying up every single company that "does AI" which massively limits the market.

What would prevent this company from being bought up? And is "select few" and "select few +1" all that different?

Slightly off topic but had 2 questions.

The last part reminded me of google training a bunch of robot arms [1]. Haven't seen much being done with it, does anyone know if anything is being done with the data?

We don't really know how much of each resources it takes to grow something. How much oxygen does 1 tomato plant take? I think growing in space would require one to investigate such questions in more detail. Ever since I came across ML and DL, I wondered if there was a way to train a deepFarmer. Something that understands relations between minerals nutrients and the fruiting body. Sometimes you want to grow stuff that might have less of certain stuff but you want more off it. Eg: low calorie high volume foods. Or you might want high protein food. If an "AI" could figure out how to maximize vitamin C for example in tomato for a population that needs more of that vs it learns to make the tomato more water rich for some other reason. AI is interesting.

[1] https://research.googleblog.com/2016/03/deep-learning-for-ro... Edit: Forgot link

I've spoken to a few people who are interested in applying machine learning to hydroponic farming with the aim of creating vertical farms.

If you could isolate the different rooms sufficiently, you could run machine learning to optimise yield (or Vitamin C).

I have doubts as to the commercial viability of this, as farming is pretty low margin, so the overhead requirements for the capital expenditures would require pretty substantial gains in efficiency, which I think just don't exist.

I have at least one serious potential investor ready for anyone attempting this (vertical farming automation).

Also, I am interested to cooperate on creating reliable contracted urban demand without the need for retail packaging - http://infinite-food.com/ - currently relocating to Shenzhen.

If automated urban vertical farming happens it will probably happen in China because dense/cheap/economies of scale/rising food prices/rapid urbanization/transition toward smaller households/cultural preference for fresh and less processed foods/less regulatory issues.

Why vertical farming specifically? No doubt vertical farming will form part of the solution to feeding urban populations in the coming decades, but the area covered by open-air "horizontal" farming is immense. We just can't do construction on that scale, not to mention the incredible environmental impacts. Growing things with LEDs seems pretty wasteful when we have a great big star shining down on us all day. Vertical farming will bring fresh food to high-density populations for sure, it just isn't going to change the face of farming across the planet.

Although it's a much more difficult problem to solve being an uncontrolled environment, my company is working on this automation but for traditional farms (http://touch.farm). The first step is gathering rich data from different crops/climates, which is why we're starting by designing the cheapest and most usable ag sensors we can. If we can make sensor tech accessible to a wide range of farmers, we'll gain a rich enough dataset to start applying ML (and farmers and the environment win in the process).

tl;dr: vertical farming is great, but what we really need is innovation/investment in traditional farming.

Why vertical farming specifically?

Ethnic Chinese investor, China focused. China has removed a lot of arable land recently for urban development, people are urbanizing at the fastest pace in human history, and food prices are rising quite rapidly.

Your comments are logical for the US and your project sounds interesting. From what I understand EU agriculture is a sort of middle-ground, with smaller-scale automation.

Also depends on crop types. Makes sense to do large fields for corn, for example. But being able to grow herbs etc locally where it's consumed, inside cities makes way more sense. This in turn frees up resources tied to growing and transporting such stuff to cities.

There are pros and cons to both but I think you can get richer data for vertically initially to train systems than move on to larger scale horizontal farms. Tough this doesn't mean there are low hanging fruits their either.

That's a good point - the choice of crop will have a large bearing. I don't know if low margin crops like grains could provide an adequate ROI given electricity costs etc, let alone the outlay required to build structures over huge areas.

In my understanding vertical farming is usually a hydrocultural practice (doesn't use soil). That works well for a lot of horticulture, but not so much for broadacre crops.

Interesting to get that perspective. I'm based in Australia, which has a similar farming landscape to the US/Canada; I'll admit my view is probably a bit skewed.

Is rice agreeable to vertical farming? i.e.: could the bulk of China's crops be moved to VF? In my understanding it's typically a hydroponic practice, so I imagine so.

Yeah I figured out you were in Victoria. I'm from Sydney. Expect a possible phone call from the interested party. No idea about rice, which I've only ever seen planted in fields, but a close friend of the investor is a qualified biologist and a cursory web search suggests that rice hulls are themselves an appropriate hydroponic medium for some crops, so perhaps you could potentially save further on transport by going multi-crop and full circle by re-using waste product as growing medium for other crops. Try Rice endosperm iron biofortification by targeted and synergistic action of nicotianamine synthase and ferritin, Plant Biotechnology Journal, 2009 (7), p. 631-644, Wirth et al

More than happy to start a conversation with them; also feel free to pass on my email: simon at touch dot farm.

Interesting thoughts on the reusing the waste. I toured the Carlton United breweries the other day, and found out it's their waste that has formed the basis of Vegemite since its inception. i.e.: there are massive wins to be made when you can "close the loop" on food production. When you've got production, processing, and reuse all under one roof like you could with vertical farming - gold.

> The last part reminded me of google training a bunch of robot arms [1]. Haven't seen much being done with it, does anyone know if anything is being done with the data?

The data gathered using aforementioned robot arms is available here: https://sites.google.com/site/brainrobotdata/home

Google (specifically Google Brain Robotics) is continuing research on this. Exciting time to be alive.

On the agriculture topic NASA has been working in this area for a while:


As I understand it growing crops in a small, sealed environment requires much more precise understanding of how plants behave than you do for traditional Earth-bound farming.

Ah, interesting. I know NASA were doing some research in aeroponics [1] but I haven't seen anything recent on it. Lot of commercial systems I see don't have a lot of research behind but they are still getting better yields I think. I also saw an old but interesting system from MIT [2]. It would be very useful to mine data from these to train ML systems.

[1] https://en.wikipedia.org/wiki/Aeroponics [2] http://openag.media.mit.edu/

> Some think the excitement around Artificial Intelligence is overhyped. They might be right. But if they’re wrong, we’re on the precipice of something really big. We can’t afford to ignore what might be the biggest technological leap since the Internet.

1. We need to work on big things whether or not they're overhyped and whether or not we're on some precipice.

2. Those who think what marketers call AI is overhyped (myself among them) don't think that it isn't something really big. Even though we haven made very little progress since the algorithms that power most modern machine learning were invented fifty years ago, there is no doubt that machine learning has become quite effective in practice in recent years due to advances in hardware, and heuristics accumulated over the past decades. It is certainly big; we just don't think it has anything to do with intelligence.

3. Are there any machine learning experts who think we are on the precipice of artificial intelligence? If so, do they think we can overcome our lack of theory or do they think that a workable theory is imminent?

I think this is one of the few RFS that I would drop everything for.

The factory environment has to be one of the most frustrating as a software developer. So much low hanging fruit that you really have to develop too much if you are not following the typical industrial idioms. (Throw ladder logic at it, keep at it till it barely works, then run it till the wheels fall off.)

Even just something as simple as reporting on a 30K$ PLC driven machine is painful. You can't buy a 10K HMI with Wonderware (a still painful piece of software) so you just ignore it. In my particular case I developed a simple reporting system on a RPI.

These industrial systems are usually not capable of even SQL or MQTT. You have to strap on an extra piece of hardware and a ton more money for licensing.

Even deployment is painful. You can't just push new code because you have no way to mock your industrial systems. Even if you could you will need to live edit your code into the PLC or you will wipe its current state of recipes and other user data. Because your PLC wasn't able to use standard SWE tools to get that data. God help you if you need to roll back.

So I am applying to this. Everything is broken. Where do you even start? Fix the platform, fix the robots, fix the job setup, fix the people.

Industrial IoT modules are solving these problems. I have been developing/deploying these modules for the past two years that interface to the controllers and send MQTT messages OTA to the cloud for analytics. Ofcourse, it is a bit of a challenge to interface to these legacy industrial systems, but not impossible.


Exactly. These systems are pretty foreign to most people. It has completely different vernacular, idioms and tradition.

It is very different than say backend development.

I asked this last time but got no response.

1. Is there a firewall between the information companies applying give you and the rest of the OpenAI effort?

2. What's to prevent a partner from seeing a good thing and passing the info along to a potential competitor already funded inside the program?

Overall it seems that this may be used to give OpenAI a strategic competitive advantage by using ingress application data for market analysis/research/positioning/signaling/etc.

I really hope the bit about "free machine-powered psychologists" is satire, but given YC's unironic techno utopianism fear that it is not.

Reading "Computer Power and Human Reason" by Joseph Weizenbaum (published over 40 years ago!) should remind people that attempting to build a "machine-powered psychologist", even if it's something that can be done, is not something that should be done.

Even if you say it isn't people will still interpret it as they wish:


Was against the creators insistence taken as the real deal.

Why? Can you explain for those that haven't read the book

Perhaps due to the article starting with discussing how AI might be overhyped, but I'm very much not blown away by this post.

Reinforcement learning for self improving robots is one of their called out areas? I've never found companies focused on tech or research problems first to be all that successful. In terms of a research projects it's not very interesting or socially beneficial compared to self driving cars or robotics applications in medicine.

It all leaves me wondering what YC's strategy is here. Maybe it's easier to establish a fund in AI, get smart people to apply, or that their expected future returns are higher?

This seems neat.

Are there any other future verticals which you would consider domain specific perks for?

Also, what exactly constitutes an AI startup? If you utilize a ML library to handle a small feature of your product, are you an AI startup?

> Some think the excitement around Artificial Intelligence is overhyped. They might be right. But if they’re wrong, we’re on the precipice of something really big.

I mean, even if it is overhyped, I think there's a lot to be excited about. Weak AI is still an amazing breakthrough for automation. The trick is to not try to do too much at a time. We do ourselves a disservice by not considering how amazingly efficient humans augmented by ML can be. The research for AI doing everything just isn't there yet, and that's OK.

Ahhh this is the moment I've been waiting for. Hype has officially hit stratospheric proportions.

Time to add the words "deep" or "learn" to your startup name and reap in the dough!

It's only hype if it doesn't work. Beating humans at games which were previously thought only humans could play.

Speech recognition, image recognition, translation have all made leaps in last 5 years.

Sure it's silicon Valley and a lot of companies will scam themselves in but I absolutely believe there will be another Google coming out soon or has already been born who will capitalize on AI.

Another Google as in a startup specifically using AI to improve search?

Is Google really going to miss improving search with AI with all their billions and brains?

Parent probably meant another company that goes from $0 to megacorp before your newborn child graduates college. Not a search company.

There may well be that kind of "hype" in the near-term, in startup land, but there is 0 chance that AI itself is hype.

At some point, making food/housing/healthcare cheaper seems like a more achievable goal than finding work for people (who are in theory competing with AI).

I'm not surprised that the general public is worried about AI but I would expect many others to worry about data. Exaggerating a bit: mathematics and AI skills is something any talented person can get individually, but gathering useful personal data on a large scale requires a huge infrastructure. So if we want to "democratize" then I'm wondering why not democratize access to data.

"We think the increased efficiency from AI will net out positive for the world, but we’re mindful of fears of job loss. As such we’re also looking to fund companies focused on job re-training, which will be a big part of the shift."

I find a lot of wishful thinking, denial and cognitive dissonance in this sentiment, which is found everywhere.

"Computers will eliminate these jobs, but humans will always have more to do."


If computers can learn at an accelerated pace, what makes you think that by the time you learn that next thing, it won't already be eliminated (or shortly afterwards) by a fleet of computers that use global knowledge? Do you really think that Uber driver - turned - newbie architect is going to be in demand vs the existing architects with their new AI helpers?

It's not black and white, but the AVERAGE demand for human labor is going down, not because EVERY human job will be eliminated but because automation allows LESS people to do the job.

So wages drop. On average.

The only real solutions are either unconditional basic income, or single payer free healthcare / food / recreation.

My goal is also to democratize AI, in particular AI research. I believe that every developer should have at their disposal the same kind of tooling and, even more important, the same ability to intersect their data with the world's data. Engineers at Facebook, Google, Microsoft and so on can test their models or even enrich them by using the Facebook, Google or Bing dataset. Independent entrepreneurs cannot do the same thing with the same ease. If we want to reach general AI any time soon, indie entrepreneurs must be let in to play.

My strategy is to build a service, free for non-profits to use, that would solve the problem of "if I only had the same data Google engineers had, this product would be perfect". Here is how it would work.

1. Go to my webpage and register a site you want me to index for you. The site URL you enter may already have been registered by another user, but to be sure the data is in my index, register it again. I will now continue to index this site every 24 hours for as long as I live. You need higher frequency indexing? Sure, no problem. You will owe me for the additional cost.

2. Download a client of choice from the website, we have them in c#, java, python, R ect. The client will let you query your own private data as well as the data in the cloud (the data I'm now generating and refreshing every 24 hours). The query language will also let you join or intersect between the two datasets. In fact, due to the nature of RPC you can use your local data and all of the data I'm generating and refreshing, as if it was your data.

3. In the end, I will be indexing such a large part of the internet that there will not be much use for Google anymore, or ads. That's the vision.

I'm not American and can't see how I'm a good fit for the YC program this summer. However I will be needing funds for cloud machines pretty soon and so far I've found noone at OpenAI to contact. Is there anyone from OpenAI reading this? This should be right up your alley. Care to speak?

I have a question about your site - "Google engineers" have a lot more data than just "Google's index of the web", they have street view data, data from people doing captchas (including the new street sign ones), click data for how people use their products (gmail, maps), maybe android autocorrect data, speech recognition etc - how would you provide access to those?

I wouldn't provide access to most of those data because I don't have the means to and I wouldn't want to either. My businss strategy is to build "strong NLP" without having to treat users as bags-of-valuable-data that I can sniff. But to integrate with a open map service would absolutely fall within the scope of my offering.

The research we would do in my team would be cutting-edge. But we would never even attempt to achieve what Google is achieving when they sniff their Android users. Why would we be cutting-edge? I don't know, but that would be our aim. Here's an example of what we would be doing in the NLP domain:

"Give me the latest sales numbers."


Give=exec latest=DateTime.Now-x sales=sp_daily_sales


exec sp_daily_sales '2017-03-23'

If you find it hard to sell AI based solutions, just call it automation rather than AI. It is a term that people react to differently.

It's like the term "technology". Spoons, chairs, bricks are technology. The modern usage of the word technology is what people used to call "high technology".

I'm interested in this per my earlier discussion on machine learning in radiology (see: https://news.ycombinator.com/item?id=13571847). It's disappointing that you have to be in the Bay Area to participate. I'm just starting residency and don't have the time to drop everything and enroll in an incubator. I think I'm one of the few people with a master's degree in computer science and (soon) to have a medical degree. I can handle the technical and medical sides of a radiology informatics / machine learning business, but I'd need someone to manage the business, marketing and sales sides.

> If the experiment works out, we’ll expand what we offer to include things like access to proprietary datasets and computing infrastructure.

Datasets, that's the most important point in Machine Learning and exactly what Google has been collecting for the past decades.

I wanted to start a project about dermatological images but how would I get that information? Then decided to start an agricultural project but then again, how to get a million images to identify a thousand species? Birds? Legal documents? Human faces? Fashion? Speech? Translation? Everything needs a huge collection of datasets.

The tools are there, that's the easiest part.

Perhaps it's an idea for YC to start a "job agency for AIs".

On the "demand" side, client-companies can offer problems to be solved by AI.

On the "offer" side, startups can provide algorithms solving specific problems.

YC can be a mediator, running the algorithms, and keeping the data of client-companies safe from anybody else (including the AI startups).

Here's an example of such an agency: http://www.aigency.co/about/

The post is clear that A.I. startups in this vertical will be given special resources, but not clear if there will be more startups selected specificially for this?

I see ML and AI raising on the market quite fast. How does this work? How many ML engineers are out there? If, broadly speaking, we have shortage of software engineers, what is the demand for ML engineers? It is not something you learn overnight as a dev or mathematician, so where are there coming from?


So how to correctly "mention this post in your application." just insert the URL ? "https://blog.ycombinator.com/yc-ai/"

Seeing as how sam altman has invested in an AI startup (vicarious.com), which is pursuing robotics applications, why would he/YC want to fund competitors for RFS ?

Democratize all the things.

Democracy is dead. Long live AI Democracy.

Would YC be interested in funding a deep learning chip startup?

Is YC still interested in Solo Founders?

I'm excited to see what new ways companies will find to call a bunch of if-statements "AI".

Curious, what would qualify as AI under your definitions?

The joke I was trying to make is that a lot of startups are doing the same things as always but calling it "AI." A chatbot using NLP technology from the early 2000s is suddenly "AI."

There are plenty of actually innovative products in the AI space.

Curious again, what are the innovative products in AI right now?

Major verticals that come to mind in terms of progress are healthcare imaging and research and self-driving cars.

At Risk of being downvoted to oblivion and out of a very well-meant interest: what if you'd replace AI with Chinese traditional​ medicine in the post above? Why would AI be more viable than that?

(Note: I am very skeptical towards CTM and quackery. But also towards AI and ML in general. Any pointers would be great. Why is this the new industry to look out for, for example?)

Recent advances in AI/ML (i.e. deep learning) have shown ground breaking progress in a number of fields especially in computer vision problems, natural language processing, and audio processing.

AI is a tool that we are now seeing solve many problems which computers couldn't even touch 10 years ago with better than human accuracy today. This field is very new and there is significant room for improvement and new approaches. IMO, a lot of the excitement comes from the really impressive progress in a short time and a very wide open field of problems still to be explored.

It isn't just the advances in the field... it is the accessibility of it all. The resources to run it are becoming more available, as GPU-focused cloud instances are now generally available, as well as just buying your own GPUs. Education is available to coders on how to get up to speed on this field, so it isn't just those with a academic background in it who are able to contribute to projects anymore. I'm not saying that us weekend dabblers in this field will surpass those who have dedicated years of study to it, but we are gaining enough understanding to come up with innovative ways to use it, and to contribute to projects. Even if we are not doing the ground breaking work, more people working on more projects will bring more innovation.

Thank you for the serious reply.

I would like to ask you and anyone reading this: what exactly gives AI/ML the "edge"? I mean, faster computers are a given these days compared to any other time in humanity. They are giving us simulations so accurate that people in the 80's would be extremely jealous. However, my question is why these new techniques are superior to the established techniques. Why would an AI solution be better? In what sense would it be different from any other method?

Again, I am not anti-AI or anti-ML but I'm interested in hearing opinions. If you could give me concrete examples where AI or ML is better in a way that is different from "faster computing" I would be sort-of satisfied.

For the ones pursuing the CTM line, forget that. I know it's BS and I hope nobody​ falls for that stuff. However, s/AI/TCM/g in the post would not make clear what the "edge" of AI is over conventional computers.

Thanks for the civil discussion.

The imagenet competition is a great concrete example. This is a massive dataset of images and the competition is to identify correctly the object in the image (truck, person, etc). The results the past few years look something like this: https://lh4.googleusercontent.com/0FLG_HLy-dHk5kPFIocPMHn5jc...

Prior to 2012, this was done using traditional computer vision algorithms with hand picked features. The accuracy was often around 30-40%, until the first deep learning model beat the 2nd place competitor by near half the error. Today, almost every single competitor is deep learning based and accuracy is equal to human classifiers.

Since deep learning has so many applications, it's hard to generalize what the "edge" is against all other approaches. In my opinion, other approaches often require handpicking features (where a ANN learns them) and struggle to recognize patterns in many areas (where a ANN does).

Thanks! This is exactly the kind of comparison I'm looking for (other fields are still welcome).

If I look at the link you provide, I see an improvement which is great. However, the improvement seems to be somewhat linear year-on-year. In the original post, the first sentence is: "Some think the excitement around Artificial Intelligence is overhyped. They might be right. But if they’re wrong, we’re on the precipice of something really big. We can’t afford to ignore what might be the biggest technological leap since the Internet." Am I right in assuming that there is the expectation of AI/ML to have "hockey stick" growth in the (not-so distant) future? If so, what kind of methods could possibly make this real? If anyone could give me some papers, an essay or some words to search for this would help me a lot. Essentially, I am intrigued why there is this expectation of hockey stick growth. If my premise is wrong, please enlighten me too.

It's a family of new techniques that has had great success doing things that are usually very hard or tedious to do with computers. These techniques seem more similar to how a human would do things than most other ways of programming things. This is the reason for a lot of the interest and hype.

Just one random example: http://lmb.informatik.uni-freiburg.de/Publications/2016/TDB1...

> what exactly gives AI/ML the "edge"?

Backpropagation. [1]

Are you familiar with the old Polaroid instant film? [2] You took a picture and pulled it out of the camera and it developed over the next few minutes. The picture went from dark grey nothingness to the image you took, like "magic".

Anyhow, so-called "neural" networks (they are no more neural than Kermit the frog is amphibious) are basically just giant polynomials with changeable coefficients. [3] The input variables are e.g. pixels from an image, and the polynomial is solved to generate a single output value indicating that it is e.g. a cat. (But before backprop this will be uncorrelated, random, useless.)

Now this image-to-yes/no-cat function is very difficult for a human programmer to write out by hand. I mean it's so difficult that no one can do it very well at all.

But using backprogagation, and a big pile of images (some of which are cats and some of which are not) and an index of which are cats for the computer to use (humans will have had to already sort and tag the images for this to work, TANSTAAFL), you can run a simple loop to develop a "function" (in the form of coefficients for the polynomial) that can "see" cats. The "training" phase adjusts the variable coefficients until the result correlates with the tagged "catness" of the images, and then you're done. After that, you can input new images to the function and it will be able to indicate whether they depict cats or not.

The function forms out of the data+backprop like an image developing on the film, projected there by the data set.

So yeah, this AI/ML stuff lets us generate useful functions that we have no other way of generating (literally no one knows how.) It's a big deal.

[1] https://en.wikipedia.org/wiki/Backpropagation

[2] https://en.wikipedia.org/wiki/Instant_film

[3] https://en.wikipedia.org/wiki/Polynomial

I worked at a startup where we used machine learning to review legal documents.

When you have 10M+ documents, it takes a long time for a team of lawyers to review them. So a company with AI/ML is at a huge advantage in identifying the documents a lawyer needs to win a case.

To get accurate results, you need a lot of data, but to crunch that data you need a lot of power, not just speed. A lot of the processing is done now with GPUs (thousands of cores)...traditional computers are CPU intensive and have limited number of cores. Think for example a multi-threaded application running on GPUs vs CPUs...the GPU can divide the problem in more tasks and the processing can be done more accurately not just faster.


PS: I think people overuse "Artificial intelligence" we're no where close...but it excites people compared to "machine learning" or "deep learning" or "convolutional network" or just statistics...It's just trying to find an equation that fits the data like linear regression.

I agree with your sentiment: to analyze these amounts of documents you need computers. And I understand the importance of parallelism, GPU computing can be way faster than CPUs for certain tasks.

In a sense, and correct me if I'm wrong, I get the feeling that the whole ML is just starting now because we have faster computers. To analyze such amounts of text and make some sense of it, you'd need a massive "database" or "fitted model" to decipher what the legalese means. Only recently has this been possible to create (thanks to Moore's law). It's not like there is a breakthrough in mathematics (statistics) right? The theory probably has been around for ages but is only now becoming practical.

> It's not like there is a breakthrough in mathematics (statistics) right? The theory probably has been around for ages but is only now becoming practical.

That's absolutely correct. But to pooh pooh it on those grounds is like pooh-poohing the fourier transform around the 70s (which theory had been around for 200 years) and only had then "become practical" with the advent of transistor-based computer. Now the FT and relatives are used in everything from audio processing and compression, image filters, statistical analysis, pretty much every form of scientific spectroscopy, MRIs, etc. If you had generally made investments in "FT-related technologies" you'd be doing well.

Only if you consider techniques like Word2Vec[1] as a non-breakthough.

No one knew you could do it, but yes, it build upon previous work. I was working in the field before and after and if anyone had asked me if one could represent all human languages in only 300 dimensions, and have vector composition be meaningful I'd have laughed at them.

Take using back-propagation to train deep neural networks. People had shown it worked in 1 or 2 layer networks, but despite years of work no one had been able to train anything deep enough to be useful. Then Krizhevsky, Sutskever and Hinton won ImageNet[2], proved it was possible and kicked off this whole ML craziness.

Neither of them is exactly because of more powerful computers, nor magical math breakthroughs. It was more lots of hard work by researchers trying many combinations of techniques until something worked.

These techniques, combined with huge volumes of data and - yes - more powerful computers are what have made the difference.

[1] https://papers.nips.cc/paper/5021-distributed-representation...

[2] https://papers.nips.cc/paper/4824-imagenet-classification-wi...

Has Chinese medicine solved any long-standing problems recently? Are there new developments which suggest that factors contributing to its previous disappointments are about to be overcome? Is there a clear path for its development? The answer to all these questions, for AI, is yes.

Hi AlexCoventry,

Please don't get me wrong: there is no future for CTM. You are 100% correct about this. However, quackery is still not a thing of the past. Therefore, I would ask of you not to debunk CTM or to point to its weaknesses (which has been done infinite times). I would like to know the AI-specific answers to your three questions (in your words or in a linked essay / paper). In a way, I don't want AI to be the "quackery of computer science": people use and believe it but it has no advantages. I would love it to be real.

I am all for a future in which we can have AI/machine learning/deep learning if it actually makes a difference. But I would love to know your opinion on what difference that could be and a little on how to achieve that. Obviously it seems I am missing something since YC is eager to invest in this field (or fields).

Your CTM analogy is not the best choice: I'd hazard to say probably around 90-95% of CTM is utter nonsense (and 99.9% of the internal justifications are), but, without CTM they'd have never discovered artemisinin, which is being deployed as a first-line treatment for malaria in areas where there is resistance to chloroquine.

Also at risk of being downvoted:

If I had general_ai.exe or worlds_best_nlp.py, what should I do with it? It isn't even clear to me how they would be useful.

Why do you expect people will be trying to build a general_ai.exe? Vertical applications of machine learning and light-AI are what most private comapnies that are getting traction are doing today, while big-cos and universities (and things like OpenAI!) are funding the research that could lead to more generalizable AI.

People certainly talk about general_ai.exe.

Often I imagine I had a technology such as general artificial intelligence or world-class NLP. Perhaps I lack imagination, but I have not yet thought of an exciting application. These are tools, much like Perl scripts, and it is not clear to me which will displace more jobs during my life.

Not serious people in any near-term way. Even bullish AI folks think general AI is MINIMUM of 10, more likely 50-100 years away.

You would never give away general_ai.exe at most you would rent it as a service. If its capabilities were linear with computing power you would just have it exponentially fund its own capabilities. Start by having it operate as a poker bot, run mechanical turk, do data analysis, imitate a software development contracting business. Then as capital accumulates add more capacity. Even if general_ai.exe doesn't get smarter than a human but can just perform more operations its probably the biggest opportunity ever.

Turn it on and ask it your question?

But seriously, narrower forms of AI are useful for things like self-driving cars.

In this thread, I have posed the question to general (biological) intelligence, since I do not have access to general artificial intelligence.

Because there is dumped model out there, and you can download and run it and see the accuracy yourself. If you cannot even do this, then your question or suspicion is worthless

Applications are open for YC Winter 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact