DailyDirt: Does It Take A Village To Teach Artificial Intelligence?
from the urls-we-dig-up dept
The accomplishments of artificial intelligence are making it a popular topic in the news again, both for its wins and its (apparent) failures. General artificial intelligence hasn’t quite lived up to its full potential yet, but more open source AI projects could help speed up development. Here are just a few reminders that open source AI projects are making progress — hopefully towards a more ‘John Henry‘ type of AI and less of a scary Skynet program.
- Google is sharing some of its artificial intelligence capabilities — like image recognition, speech translation, etc. Maybe people will find cool new uses for AI, and hopefully, some of the machine learning bugs will be caught faster. [url]
- Yahoo hasn’t been getting too much positive press lately (for the health of its business), but it also released its own open source AI tools (aka CaffeOnSpark AI engine) recently. Yahoo still has a lot of Flickr image data to help train its deep learning software, but selling off Yahoo’s “core business” could have an effect on Yahoo’s AI projects. [url]
- OpenAI is a non-profit artificial intelligence research company that has just set up new offices in San Francisco. It has $1 billion in funding from several sources, aiming for a long-term benefit to humanity from artificial intelligence projects. [url]
- China’s main search company, Baidu, has released its own open source AI software — used to build systems like Deep Speech 2 which recognizes human speech better than some humans (at least for short phrases or messages). Baidu’s code is called Warp-CTC, and you can play around with it all you like — but you’ll probably need access to some heavy-duty hardware to make your own HAL9000. [url]
After you’ve finished checking out those links, take a look at our Daily Deals for cool gadgets and other awesome stuff.
Filed Under: ai, artificial intelligence, cloud, deep learning, deep speech 2, image recognition, john henry, machine learning, open source, speech recognition
Companies: baidu, flickr, google, microsoft, openai, yahoo
Comments on “DailyDirt: Does It Take A Village To Teach Artificial Intelligence?”
Does It Take A Village To Teach Artificial Intelligence?
Depends upon:
A) The intelligence of the village
and/or
B) The artificiality of the intelligence desired.
On the other hand, I see it like GMO foods. When it has been tested in controlled situations, with humans, for 50 years or so, I will begin to think about taking a more favorable view of the subject.
Beating a human at chess or go or Jeopardy is one thing. Understanding humor, or satire, or innuendo, or tongue-in-cheek, or micro-expressions, or body language, or eye contact longevity is another hemisphere entirely. When the machines have proven to understand human communication, in all its forms, it is time to start thinking about believing they understand you.
Then qualifying their ‘intelligence’ beyond even complex math is another matter. Let them explain poetry, the differences between poetic styles, why a movie is good or bad with the only input being the movie itself, describe a sunset, tell someone about the colors of fall in New England, interpret the look in a scorned lovers eyes, etc.. In addition, open source and heavily reviewed by non-partisan and community trusted coders over a long period of time.
In the mean time, it is going to be a lot of fun.
you missed the internet teaching Microsoft’s AI to be a bigoted sexpot.
http://www.complex.com/life/2016/03/microsoft-tay-tweets-about-sex-hitler
Re:
OK, maybe you didn’t, but I would have thought it would be worth a bullet point.
It Certainly Takes A Village ...
… to corrupt one.
Sounds more like a setup for Toner Wars, ah, i mean AI Wars.