An AI that helps you in making Awesome Sketches - Google AutoDraw !

Sketching is your love and passion and you always adore making doodles, but there might be time when you find it difficult to make any design or some sort of sketch using your mouse and keyboard, and even after hours of effort and practice the output you get is not satisfactory or not what you expected it to be. Google has just launched an artificial Intelligence based bot called "AutoDraw" that turns out or convert not so efficiently drawn doodles or designs into meaningful sketches. If you want to use it you can visit the following link and have a go with the AutoDraw https://aiexperiments.withgoogle.com/autodraw. 

You can use AutoDraw right from your laptop or desktop or any smaller devices such as your smartphone or your tablet.Let us get into the technical side of this awesome AI product from Google and let us try to find out, how it works and how it converts or helps you in making awesome sketches from not so proficient doodles.

It is with the help of Neural Networks and Machine Learning, Google AutoDraw analyses your drawing and suggests different objects and sketches based on it from its database, in its database, there are a large number of designs are already available that might have been added by Google(for initial roll out designs have been taken from designers and from design studios) itself or the designs might have been submitted by some designers, if you want to submit your own design then you can submit it to AutoDraw as well. To begin with, AutoDraw provides a white canvas where you can scribble and just above the canvas is a bar showing possible matches to your drawing. You can also choose to draw whatever you want without any suggestions from Google, along with text box, shape, and fill colour options. Doodles, as well as individual shapes and lines, can be resized as well.

The Blank Canvas where you can draw

I have simply drawn a line and see AutoDraw has suggested me with number of designs just after the Title in a bar

Once you get the desired result or once you get the design of your choice, you can download it from AutoDraw as well and later on can share it on the social network as well. For the starting purpose of AutoDraw Google has used Designs from artists and design studios and have thanked them on its Blog Post. If you want to see the designs submitted by various artists you can see it from the menu option as well.

Well, this is an awesome tool for the people who wants to make sketches but somehow don't have the creativity to make one, as here a non-designer person can also make an awesome and professional looking sketch and can then download it and use it wherever he/she wants.

Thanks to Google for providing this awesome AI tool.

Source of Information: http://gadgets.ndtv.com/internet/news/google-autodraw-turns-your-doodles-into-meaningful-sketches-with-machine-learning-1680419

Robot Solves Rubik's Cube in less than a Minute !

Robotics and Artificial Intelligence has started growing heavily from the past few years, as we all know that Artificial Intelligence is the field that deals with making the Machine as Intelligent as humans are in order to do the work faster in comparison to the work done by humans. Although most of the scientist and researcher believes that the growth of Artificial Intelligence can result in human being loses their jobs but on the other hand, some Scientists makes it very clear that Artificial Intelligence does need the involvement of Human beings for their handling and proper functioning.

SUB1 RELOADED

In February there was a news regarding a Robot that passes through the Google's Captcha Test that basically been used for separating humans and robots, but this particular robot surpasses it and made itself count as human. Now another big news is coming through the Internet today that there is a Robot called Sub1 Reloaded who has solved the Rubik's cube in less than a minute and made a new Guinness World Record. The Robot Sub Reloaded solved the Rubik's cube in just 0.637 seconds, the previous record of Sub1 was 0.887 seconds. After making a thorough check about this amazing job of the Robot, The Guinness World Record has confirmed this amazing achievement made by the Sub1 Reloaded Robot.It only took 21 moves to solve the Rubik's cube by using its 6 motors that were connected to each side of the cube.

The Spokesman said that "We want to show that problems can be solved much more efficiently using MicroElectronics".

Check out the Video here, that have been Uploaded by Guinness World Records:

Artificial Intelligence is surely growing and growing big time, and It is sure that we are going to see lots and lots of great invention and usage of AI in our day to day life in near future.

Source of Information: http://www.guinnessworldrecords.com/news/2017/3/video-robot-breaks-world-record-solving-rubiks-cube-in-0-637-seconds-464392

Deep Coder- An AI that Writes Code !

Artificial Intelligence has come a long way, Microsoft along with the Computer Scientists of the University of Cambridge has discovered an Artificial Intelligence called "Deep Coder", that writes code without even the prior knowledge of programming or coding.Deep Coder works by analyzing the code of previously written applications and then writes its own code. 

• This new System is based on a technique called Synthesis Programming which is a form of automatic programming that writes codes by itself, thanks to the development of this new System that can perform a search in the database of previously written codes and after analysing the suited code for the given requirements start writing code to solve the given problem.

• At present this System can only solve small programming problems and can write only 5-6 lines of codes but this invention has a great potential to become an automatic tool for coding. A person who wants to develop a software but doesn't know how to code has just to describe its requirement to Deep Coder and the Deep Coder will then use Machine Learning technique to analyze the previously written code and then will start writing code for the given requirement.

• According to Microsoft Research Team and Cambridge Computer Scientist, Deep coder at present is only able to solve simplest of problems as they have tested it for solving online coding challenges, but they are aware of its potential and are optimistic about its bright future.


• However, According to Armando Solar-Lezama, a Massachusetts Institute of Technology(MIT) researcher, Deep Coder is not a threat to a manual developer, as they believe that this system can prove helpful to handle basic tasks or simpler tasks so that the developer team can concentrate on the complex issues and doesn't waste their time on solving routine tasks.

This is a great news for the followers of Artificial Intelligence, as we have already seen Artificial Intelligence based products playing games, writing encryption and other things, this isn't a surprise that an Artificial Intelligence is writing code as well. Hope to get more out of it in near future as it can really prove helpful for developers to develop software at a rapid pace with a more optimized version of coding.

Source of Information: http://www.wired.co.uk/article/ai-write-code-microsoft

What is Machine Learning?

All of us must have heard about Machine Learning for at least once in our lives. It is one of the hot topics or we can say it is one of the areas that is containing high possibilities in near future in the field of Artificial Intelligence.

Machine is Learning by itself :P

Let I start with a simple definition that can be understood by even a non-technical person:

When we talk about machines in computer science, directly or indirectly we are talking about Computers. Computers can perform all the tasks pretty efficiently that has been given to them by users, for example, I have given a complex mathematical problem to my computer, it will hardly take milliseconds to come up with the correct answer that is the power of the computer. But if I provide it a picture and want it to recognize that who is in the picture whether it's a human or an animal, it won't be able to tell or answer. That shows, our machine is not intelligent by nature.

So what is Machine Learning then, Machine Learning can be defined as a type of Artificial Intelligence that provides the ability to a computer to learn by themselves without being programmed explicitly.

The Computers are by default programmed to perform a particular task for example if it is programmed to perform a calculation it will only do that, it won't do some other task. Machine Learning provides the capability to a System or a Computer to start learning by itself without any intervention of human being. 

• Let us take the real time examples of Machine Learning Implementations, We all do shopping online, we all purchase clothes, gadgets from Amazon and all e-commerce giants. Whenever we visit the websites there is a section of Products Recommendation, where we see all the products listed according to our previous searches, how it is possible it is possible because of Machine Learning.
  
  
  
  
  

• Another example is of Facebook News Feed, there we often see that news and activities that are according to our liked posts or shared posts, for example, if News Feed is showing activity of a particular friend and we are providing a response to it , then Facebook will start showing more activities of that particular friend in News Feed it is all because of algorithm that has been developed by Facebook using Machine Learning.


• Self Driven cars have created such a hype in the market and it is believed that they become reality in near future, it is all because of Machine Learning. Let us assume that a self-driven car is running on the streets and there is a board of Stop written over there but due to some problems it is not clearly visible, as a human being we can recognize it by seeing its red color and we will stop over there but how the machine the car recognize it is the concept of Machine Learning that will tell the machine that you have to stop it here. A company named Tesla is working on this concept of Self-driving Cars and the Google self-driving car project is now "Waymo". Waymo stands for a new way forward in mobility.
  


• Email Spam Detection is another application of Machine Learning, the Mails that seems spam to mail clients like Gmail or yahoo they directly send them to Spam Folder.


• Behavioral Advertisements that we see when we make a search on Search Engine is also the implementation of Machine Learning, the pages we visit often is being taken into consideration while showing ads to us.


• Recently Google Launched and app called Allo, in it the app will automatically provide us the possible replies to messages there also Machine learning is implemented. It also recognizes the images and gives us possible replies that can be made for them.
  

It is believed that the area of Machine Learning has not been explored heavily till now, so there are lots and lots of possibilities there. If it is implemented well in near future which it seems like will happen for sure then we surely will see machine becoming more intelligent and can understand most of the things by themselves only.

What is Artificial Intelligence and What are its Types?

Artificial Intelligence has always been a topic of excitement for me, AI has given new heights to science and it is making wonderful products for years now. Whenever we hear the term Artificial Intelligence the first thing that comes to our mind is Robots, Although Robotics has been one of the major areas of AI, in real AI is more than that. 

Let us first start with the definition of Artificial Intelligence.

Artificial Intelligence is defined as the Science that makes computer to do certain things that if done by a normal user requires intelligence. According to Techopedia Artifical Intelligence is an area of Computer Science that emphasizes on the creation of intelligent machines that works and react like human beings.

There are many definitions out there on the web, but the main conclusion that we can draw from all of them is that AI can be defined that branch of computer science that is working on the process of making the machine intelligent and make them think like a human brain.

Now as we have the idea of what AI actually is, let us shift our focus to types of AI now:

• Artificial Emotional Intelligence (AEI): Programming emotions into a machine is the field of Artificial Emotional Intelligence so that after reaching the state of singularity (Singularity is explained in below)  machine can identify us and won't be taking unwanted actions against us and also will be able to follow our instructions.


• Artificial Narrow Intelligence (ANI): ANI is one of the types of Artificial Intelligence it is also called as Weak AI. ANI is that type of AI where the machine is only expert in a particular area for example if our machine is expert in playing chess, It can only do that and can't do anything else.


• Artificial Wide Intelligence (AWI): AWI is one of the types of Artificial Intelligence it is also called as Strong AI. AWI is that type of AI where the machine will be as powerful as a human being and the brain of a machine will be started to work like a normal human being does and Common sense of a human will also be there in the machine.


• Artificial Super Intelligence (ASI): Artificial Super Intelligence is also called as Singularity, Once the Machine becomes as powerful as the human being after that there will be time to time improvement in it and then it will reach a level that will be called as the Singularity level. If Singularity level reached and the machine will become more powerful that human beings then there might be the case that machine won't be taking the order of humans , so all these things have to be taken into consideration that make machine as powerful as you can make them but always keep in mind one thing that program them in such a way that they should never judge them above their creators and follow their instructions always.

       Application areas of Artificial Intelligence:

        

        Here are some.

          1 ) Game Playing

You can buy machines that can play master level chess for a few hundred dollars. There is some AI in them, but they play well against people mainly through brute force computation--looking at hundreds of thousands of positions. To beat a world champion by brute force and known reliable heuristics requires being able to look at 200 million positions per second.

          2) Speech Recognition

In the 1990s, computer speech recognition reached a practical level for limited purposes. Thus United Airlines has replaced its keyboard tree for flight information by a system using speech recognition of flight numbers and city names. It is quite convenient. On the other hand, while it is possible to instruct some computers using speech, most users have gone back to the keyboard and the mouse as still more convenient.

          3) Understanding Natural Language

Just getting a sequence of words into a computer is not enough. Parsing sentences is not enough either. The computer has to be provided with an understanding of the domain the text is about, and this is presently possible only for very limited domains.

          4) Computer Vision

The world is composed of three-dimensional objects, but the inputs to the human eye and computers' TV cameras are two-dimensional. Some useful programs can work solely in two dimensions, but full computer vision requires partial three-dimensional information that is not just a set of two-dimensional views. At present, there are only limited ways of representing three-dimensional information directly, and they are not as good as what humans evidently use.

          5) Expert Systems

A "knowledge engineer'' interviews experts in a certain domain and tries to embody their knowledge in a computer program for carrying out some task. How well this works depends on whether the intellectual mechanisms required for the task are within the present state of AI. When this turned out not to be so, there were many disappointing results. One of the first expert systems was MYCIN in 1974, which diagnosed bacterial infections of the blood and suggested treatments. It did better than medical students or practicing doctors, provided its limitations were observed. Namely, its ontology included bacteria, symptoms, and treatments and did not include patients, doctors, hospitals, death, recovery, and events occurring in time. Its interactions depended on a single patient being considered. Since the experts consulted by the knowledge engineers knew about patients, doctors, death, recovery, etc., it is clear that the knowledge engineers forced what the experts told them into a predetermined framework. In the present state of AI, this has to be true. The usefulness of current expert systems depends on their users having common sense.

          6) Heuristic Classification

One of the most feasible kinds of the expert system given the present knowledge of AI is to put some information in one of a fixed set of categories using several sources of information. An example is advising whether to accept a proposed credit card purchase. Information is available about the owner of the credit card, his record of payment and also about the item he is buying and about the establishment from which he is buying it (e.g., about whether there have been previous credit card frauds at this establishment).

Note: This above application areas  information is taken from

http://www-formal.stanford.edu/jmc/whatisai/node3.html

That's all friends in this article of mine, just go through the article and get to know about the     awesome science of Artificial Intelligence.