miket

That's a good question.

One answer is that in each of those cases there are alternative mapping/news/music services. An open-source, privacy-first device could be an intermediary that wraps other services, and if it got significant adoption, serve as a incentive for those services to compete on quality/price/privacy-awareness.

There are even startups, like ours (Diffbot) that are building general purpose knowledge graphs for developers to use.

Was recently reported that Google receives 100K DMCA requests per hour: http://www.theverge.com/2016/3/7/11172516/google-takedown-re...

Exactly, you got it.

Founder, here.

It is an imperfect analogy, but let me try to elucidate what our model actually is.

We currently power over 250 companies, from some of the most popular applications like Instapaper and Shopspring, to search engines like Bing, DuckDuckGo, Yandex, and eBay to business intelligence tools like Crunchbase.

Our thesis is that access to large volumes of structured data will be just as critical a resource as the learning algorithms or the computational hardware for developers building the next generation of smarter application, and we offer our knowledge-as-a-service as a platform/component to build end-user applications around.

Part of the new funding is going to our effort to build a comprehensive database of knowledge synthesized and fact-checked from the sum total of all of the information on the web. There is some parallel to Google's knowledge graph project, except that our method is fully autonomous (no human curation) and we are building a self-sustaining business around it with a fully-featured API that everyone (outside of Google) can use.

Here's a page with various benchmarks: https://github.com/soumith/convnet-benchmarks

An issue has been created to add TensorFlow to this shortly.

Working on it. The reliance on visual features means it works decently well on extracting from international pages. Try putting in foreign-language product pages or article pages into our homepage testdrive or with a developer token to see an example.

Unfortunately, I can't speak about this customer integration without permission. You can try to ask questions about Diffbot, though!

Founder here. I wouldn't describe what we are doing as cloning Google's Knowledge Graph. Rather, we're trying to build a machine that autonomously learns from reading the web, and makes that structured data accessible and queryable to any app, smart device, or business that uses data. You can also tell Diffbot what part of the web to crawl.

Our APIs power consumer services including Instapaper and DuckDuckGo to enterprise applications like content management and market intelligence.

Would love to have this for email.

Is there any way to generate this algorithmically, or otherwise view the most upvoted comments on HN?

Here's a list of services Google has shut down: http://en.wikipedia.org/wiki/List_of_Google_products#Discont...

Diffbot - Palo Alto, CA

We make the web understandable to machines.

Checkout our recent Techcrunch coverage: http://techcrunch.com/2015/03/31/diffbot-discussions-api/

And see the technology applied to this very comment thread:

http://www.diffbot.com/testdrive/?url=https%3A%2F%2Fnews.yco...

If interested, please send an email to introduce yourself to the team at jobs@diffbot.com

Gigablast is alive and well at Diffbot! (founder here)

https://gigaom.com/2013/09/10/diffbot-brings-big-time-search...

Hey Sam, this is Mike, the founder at Diffbot, the tech that powers Instapaper. Stitch looks great, and we'd love to help out. Send us a PM.

http://diffbot.com

Not beats experiential education and building an early network if intellectual peers.

I'd recommend to a 14-year old an AI camp being offered by the Stanford AI Lab: http://ai.stanford.edu/sailors/

Diffbot (Palo Alto, CA)

We're a 14-person AI startup right by the Stanford campus. We apply deep learning, computer vision, and natural language processing techniques to the problem of understanding webpages. Our APIs convert billions of webpages automatically into structured data for the likes of Bing, Digg, Instapaper, eBay, Adobe and others.

We use very few 3rd party frameworks and strive to develop our own performant machine learning techniques.

Looking for published ML/CV/NLP specialists, data fusion / knowledge graph, and web-scale crawling experts with a track record of building intelligent systems that perform at human-level accuracy rates. If interested, send us a note at jobs@diffbot.com (due to limited HR bandwidth, naked resumes will be discarded).

I'm a big fan of the startup FancyHands - https://www.fancyhands.com/. The value proposition offered by the two services is similar: text a task, have an on-demand assistant find a solution for you. I do wish sometimes FH would return something faster: response for me range from 30m to a few hours. Though it sound like the with the current scaling pains at Magic, they aren't fully living up to their promised responsiveness, either.

UPDATED: Just saw that FH has released some real-time communication capability with their assistants. Very cool: https://play.google.com/store/apps/details?id=com.fancyhands...

We wholeheartedly support this cause at Diffbot! http://www.diffbot.com/robots.txt

+1 for Casetext, it's designed for this.

Diffbot (Palo Alto, CA) - INTERN/REMOTE/VISA

AI/Computer Vision startup spinout of Stanford, looking to revolutionize web search. We apply techniques in NLP and CV to analyzing webpages, converting them into structured data. Our extraction APIs currently power Microsoft, eBay, Facebook, and many startups.

We're looking for experts in: Machine Learning, Computer Vision, Data Fusion / Record Linkage, and crawling and scalability.

Read more about our mission here: http://goo.gl/bcx7cW and let us know who you are: jobs@diffbot.com

Check out http://casetext.com. Their "ReCites" extraction is a pretty useful way of summarizing a legal case text.

Looking forward to seeing further development!

For a more comprehensive word API, check out the excellent Wordnik API: http://developer.wordnik.com/docs.html#!/word

Diffbot - Palo Alto, CA Hiring Computer vision, NLP, deep learning, web search engineers.

http://diffbot.com

Scratch is a great project. For those of you looking for a gentle intro to programming for your kids, I highly recommend checking out http://tickleapp.com, a recently Kickstarter-backed Scratch implementation for the iPad.

Thanks a lot for writing this tutorial, Andrej. Super clear, the next Feynman in the making here!

Interesting choice of language, maybe you could add little run buttons to the code snippets to try all the examples in the browser :-)

Don't forget that a large chunk of Mozilla's revenue comes from Google: http://thenextweb.com/insider/2013/11/21/mozillas-reliance-g...

Diffbot - Palo Alto, CA - http://diffbot.com

At Diffbot, we're applying sophisticated deep learning, computer vision, and natural language processing techniques to convert the web into structured data. Join our growing team of world-class experts as we build the world's largest collection of structured information to power apps and more intelligent search. We are currently seeking roles in machine learning, product engineering, and business development. Shoot us a message at jobs@diffbot.com.

At Diffbot, we have an automated discussion thread parser, currently in beta testing, that might be exactly what you need. Send me a note at mike@diffbot.com and I'd be glad to hook you up.

(disclosure: I work there)

Would love to hear more about your use case Denny--sending you a PM.

As for web-scale crawling of particular verticals such as products and news, you might want to try: http://www.diffbot.com/products/automatic/

We're planning on releasing support for jobs, companies, and people later.

(disclosure: I work there)

Check out http://diffbot.com/products/automatic/product/ to do this fully automated.