Facebook Hate Speech AI Blocking The Wrong People
Published in News

Hate is a powerful item and when it spills out it can be violent, rude and many other things. Because of the power of hate found in images, media, mems, etc., many have wondered why there are not more efforts to prevent the posting or sharing of such information. After all why would a media (social or other) want to allow hate speech or images posted on their pages. Facebook took this thought process and turned it into a policy that is designed to help stop hate speech from showing up. Now the system has inadvertently started censoring the wrong people.

Tuesday, 28 January 2014 13:33

Google buys up AI company DeepMind...

British start-up DeepMind became a part of Google, for which the search giant paid about 400 million US dollars. Thus, this company became financially Google's largest acquisition so far carried out in Europe.

Monday, 28 October 2013 21:08

CAPTCHA Allegedly Broken By New AI

In the war against spammers and other malicious individuals there have been many protections invented. One of these is the CAPTCHA system. This is a system that presents an image that a user must type in to prove they are human. This image is generally not readable by bots as anything other than an image file so it can stop many attempts at spamming. The system is not fool proof and some of the simpler systems have already been hacked through prediction of the codes used to generate the images. Now, it seems like that someone has found a way to actually read the image and input the information.

Tuesday, 15 January 2013 09:18

When supercomputers learn slang

watson

There has been quite a lot written about IBM's Watson. The supercomputer that once won on the famous quiz show Jeopardy!. One of the most covered topics is about the algorithms which provide an excellent understanding of conversational language. This nice bit of programming not only understands what’s spoken, but it also sounds human, with sentences that use informal relaxed syntax and vocabulary of everyday life.

Thursday, 20 October 2011 22:09

Should Google Worry About Siri?

Siri-2Has anyone gotten tired hearing about Siri? I mean people are just going on and on about it. People are even interviewing the phone for crying out loud. The funniest thing about it is that some are looking at these interviews with Siri and claiming this is proof that Apple has developed some form of new AI. Ugh… the things some people will believe. Once and for all; Siri is a VERY cool product which can be fun to tinker with, dangerous if you are not careful and an extremely simple concept.

We have said this before, but it bears repeating. Siri is a speech to text engine that takes the input and then searches through a series of commands and responses. These are preprogramed and Siri cannot make decisions of judgment calls. If it cannot find it in its internal database as a command or a pre-programed response it will attempt to find it using an online search engine (Google, Bing, Wolfram Alfa…) Once it has a response it then converts that from text back to speech. The algorithm for this is not that complicated and is not much more than an extension of traditional voice commands. In fact I have a remote control sitting in my lab from Amulet Devices that is capable of understanding my commands and looking for the proper responses. If I want to watch an action movie it will search through my collection or in an online catalog for that genre. It can do the same thing with music. I have had this for over a year now, so really Siri is cool but not ground breaking.

This leads me into a rather ridiculous comment made by one of the principal investors in Siri before Apple bought it. Gary Morgenthaler, has said to Cnet that Siri represents a real danger to Google and Microsoft.  He also goes on to make a claim that Siri is capable of “understanding” that is simply not true. Siri takes the natural speech and converts that to text that it can respond to. It does not understand anything more than the pre-programed responses. If Siri truly understood it would be able interpret what you are saying and then respond with questions for clarity or with real information. When I asked Siri if it liked Doctor Who, it responded to me with a listing of Doctors within 10 miles all pulled from Google.

Even funnier is that Cnet comments that Siri must be popular because it has spawned a popular blog called Sh** that Siri Says. If this is the case then I guess that the auto correct function on Android and iOS is very popular as well. After all there are multiple sites that are dedicated to just how badly these products screw up what we are trying to say and do… oddly enough, just like Siri does sometimes.

In reality, we have someone riding the popularity of a novel API that is truly in its infancy as far as technology and application. Microsoft, Google and others are all working on similar APIs. Apple just did what they do best, they got to market first with a workable product, but if you look closely you can just make out the duct tape holding parts of it together…

Source Cnet

Discuss in our Forum