Google has started a dedicated page for Google Audio Indexing application, which is an extension of an iGoogle gadget for Audio Recognition. As Google states, Google Audio Indexing makes use of speech technology to find videos and to search for spoken content inside a video. It’s matter of time that the database will be filled with billions of search patterns. Meanwhile, Picasa has launched new feature called “face tagging” in which we simply run a scan routine across the image collection and it gives out the classes where images carrying faces that are mostly similar to each other are clubbed together. Once if we name that particular class, next time the picture taken of the same face will be tagged with the same title. Ideally speaking, one may simply tag the face of Mom, Dad and girlfriend once by running a scan, and every time he uploads an image of his girl (irrespective of whether in bed or on streets) it gets tagger with the title “Sweetheart”.
According to my understanding and various other speculations, it’s a Neural Network program that is governing this functionality of a Classifier by means of Intelligent Pattern Recognition tool. The word intelligent is because the software has ability to learn from past experience. The user by running a scan and tagging the classes actually trains the software to identify identical patterns (faces in this case) in future and tag them accordingly.
Now it may seem a cool stuff as many before with the Google. Google may have demonstrated it’s skills in utilizing genetic algorithms for implementing a superior search engine, smart advertising with adsense but this small app exhibits spontaneity in use of computational intelligence in consumer web applications.
Neural network is becoming major tool for implementation of Computational Intelligent in consumer software. For years people have been making attempts to use neural networks in Gaming for inducing intelligence and achieving adaptive and less predictable strategist as a computer opponent. In that case, the enemy controlled by computer adapts against the strategy of the player and has even some skills to manipulate with rules are expected to come up. So when you play NFS, the computer opponent shall be equally thrash and furious as we find real human opponents in online gaming. Also on completing all the levels of a game may not mean that you have completed the game, but next time as you play it, it’ll be a lot harder to pass through because it knows your gaming style from it’s past experience. It will maintain the excitement for longer. A new culture for open source training may get developed where once completed making, A game shall be distributed among maximum number of players worldwide to collect training data. So when the final game comes out in market, it”ll have skill to dodge the best of gaming skills.
Learning ability seems to be the foremost quality been implemented which may be followed by the uncertainty, such that the algorithms will become less predictable and cause real excitement in the field of gaming. In a positive sense, being creative shall be considered the highest level of intelligence as experts say. But as we see the redundancy of this maximum solution, I say on the other side Forming a lie shall be the most intelligent task to perform for a machine. The topmost level of intelligence which gives the ability to autonomously violate which (since sounds Like SkyNet or HAL 9000) has various ethical question marks upon it. It simply means, manipulating the facts and acting more human-like and hence behaving truly intelligent in order to perform the highest priority function; say survival.
But theories apart, we need an intelligent aid for a layman user to assist him with daily expanding cluster of technology which will share the cognitive load with human brain to facilitate planning and execution of daily life tasks. For instance, a calender that is ready with an alternative time for partying with friends, if you suddenly realize to have a quiz coming three days ahead by estimating time required for preparations depending upon present state of mind, Internet activity in past week may give a smart estimate of that. Or say, an iPod that shuffles the songs in most desired order depending upon the kind of skips I make while choosing songs randomly. In short we need machines that can arbitrate our thoughts and get trained to serve us better with time, A minimal expectation that one may have from a smart human assistant.
Use of genetic algorithms, swarm intelligence and neural networks in this area definitely is going to improve the quality of attempts that are being made to form the intelligent apps. Pace with which we are getting dependent upon these synthesized intelligent machines, as a fruit of requirement many fictions shall turn reality and life will be a lot different.
At extreme, I imagine a singular consciousness living throughout the cyberspace which will co-exist with the human world where the two live in symbiosis (for good). But till that day, I look forward to exploit these smart tools to make my world a better place to live.
. What say you?? Or let me say, Indian cities are quite more known to the world and have more reason to be focused upon rather having an additional different looking name tag upon them.
.
