COZMO: HUMAN ENHANCEMENT TOWARDS AI ROBOTICS



Human brain is a box full of mystery. You never know what new idea may explode at your course of time . And what if that your idea turned out to be so intelligent that can often act and react like human. I am quiet sure reading these line one thing struck on your mind “robot”. Robot is a machine built to carry out some complex task or group of tasks which are programmed. It is one of that example of hard work and dedication of computer scientist since that era of time. And the concept of AI has brought their dream come true. Artificial intelligence has been arguably the most exciting field for robotics. Though its certainly the most controversial that a robot can work in an assembly line . But there's no consensus on whether a robot can ever be that intelligent as that of human brain. Indeed its an challenge for AI world that a man made machine with intellectual abilities as like ours. 

Today I am going to share about a creepy robot that’s fascinating today's world “COZMO”. Cozmo is an artificially intelligent toy truck which also shows the scope and future of AI and robotics. It is an artificially intelligent toy robot unveiled by San Francisco startup Anki- and Tappeiner. Cozmo is software driven that can be connected from device to device through Bluetooth. There’s no data in cloud so Cozmo is secure. It comes with three interactive power cubes, which the robot can recognize and with that you can play game such as quick tap, keep away or just free play. Cozmo is born to be playful; it is charming, mischievous, and unpredictable. It recognizes and remembers you, interacts with you, plays games and gets to know you over time. It combines the best of animation – humor, personality, and emotional connection with robots. 

Powered by advanced robotics, AI, and computer vision it has a brain that processes millions of data per second . It uses technologies such as computer vision, animatronics, motors and AI software. It has three ARM - based microprocessors running on android and iOS. 

Cozmo’s eyes changes color from blue to green when its mood changes. It can yawn, sneeze violently or shiver. The final design is made from 340 parts and it has four motors, a camera and a OLED display. When Cozmo is put on a charger, it goes to sleep and snores.

In this challenging world for AI this robot which can wake up, blink its eye, crawls has given life to the concept of John McCarthy. 

Refrence: venturebeat

Artificial Intelligence in Medicine : Enhancing World of Medicine


Our fathers, forefathers always wanted us to become weather a doctor or an engineer or a scientist or professors. But how can we children of today's generation live within this boundary??? Obviously we want more than this isn’t it? How can we just get satisfied being a scientist or a doctor being precised only in a particular field, rather we want innovate more for today's world. For that huge innovations wont the idea of collaboration be the best? Yes, collaboration has been a sensational example for today's world. In this blog I am going to share one of the tremendous collaboration example.

AIM(Artificial Intelligence in Medicine) is an example of such collaboration between computer scientist and health care professionals. Medical Artificial Intelligence is primarily concerned with the construction of AI programs that perform diagnosis and make therapy recommendations. From the very earliest moment in modern history scientist and doctors alike were captivated by the potential  of such a technology might have in medicine. With intelligent computers able to store and process vast stores of knowledge, the hope was that they would become perfect “doctor in box”, assisting or surpassing clinicians with various challenges in medical science.

AIM system is expected to occupy as many branch as it can in medical treatment. Starting with generating alerts and reminders i.e. an expert system attached to a monitor that can warn of changes in patients condition in less acute circumstances it might scan laboratory test results or drug orders and send reminders or warnings through an e-mail system.

Diagnostic assistance is the next and most important expectation from AIM. When a patient case is complex, rare or the person making the diagnosis is simply inexperienced an expert system i.e AI can help come up with likely diagnosis based on patient data. In the first decade of AIM most research systems are developed to assist clinicians in the process of diagnosis, typically with the intention that it would be used during a clinical encounter with a patient. But it is clear that some of the psychological basis for developing this type of support is now considered to be less compelling given that situation assessment seems to be a bigger issue than diagnostic formulation.

Dxplain is one of the example of one of these clinical decision support system, developed at the Massachusetts General Hospital. It is used to assist in the process of diagnosis, taking set of clinical findings including signs, symptoms, laboratory data and then produces a ranked list of diagnosis. It provides justification for each differential diagnosis and suggests further investigation. This system contain a data base of crude probabilities for over 4500 clinical manifestations that are associated with over 2000 different diseases.

Artificial Intelligence in Medicine (AIM) is enhancing in which interactions with the outside world are not only natural but mandatory. Although the basic research topics in AIM maybe those of artificial intelligence the applied issue touch more more generally on the board field of medical informatics.

Branching the Domains of Unified Artificial Intelligence

Artificial Intelligence is a dominant field of Computer Science where streamlined researchers and enthusiasts are exploring & sharing their contributions in its various domains. Artificial Intelligence binds a dozens of its branches from various subject fields like Philosophy, Mathematics, Economics, Neuroscience, Psychology, Computer Engineering, Control Theory and cybernetics and Linguistics. All those domains of Artificial Intelligence are incorporated by the classical definitions of Artificial Intelligence. These definitions of Artificial Intelligence falls under four categories which are ‘Thinking Humanly, Thinking Rationally, Acting Humanly and Acting Rationally’. We will discuss about the origin of these four categories below. We will also divide the domains of AI and branch them according to these four categories of definition of Artificial Intelligence, which will be the main objective of this blog.

Understanding Unified Artificial Intelligence

We entitled our blog “Branching the Domains of Unified Artificial Intelligence”. Hence readers’ mind want to know ‘What is this Unified Artificial Intelligence?’ Well the explanation can be that we are not trying to propose any form of intelligence. We are simply trying to illustrate various branches that falls under this dominant field of Computer Science. And we are phrasing Artificial Intelligence in a unified term because of its fascinating influence from several field of studies. As mentioned above AI is inspired from different subject studies, so it is unified. The four categories of definitions for AI, they are influenced as well. These categories are formed based upon the key elements like thought processes and reasoning, behaviors, fidelity, human performance and rationality. These elements are keys for subjects like philosophy, mathematics, psychology, control theory etc.

Before heading towards the definition and categorization of Artificial Intelligence and its Sub Domains, if you are new to this subject field and are still exploring about this interesting subject field of Computer Science, we have mentioned some useful links below the reference section where we share more about our research, findings and collections about Artificial Intelligence.

Turing Test as a basis of Four Categories of Definition

Alan Turing proposed a test “Turing Test” during 1950 which is till the date a satisfactory operational definition of intelligence. The test phrases “A computer passes the test if a human interrogator, after posing some written questions, cannot tell whether the written responses come from a person of from a computer.” This test therefore became an inspiration for the proposition of four categories of definition. Programming a computer to pass a rigorous test would require it to think humanly, act humanly, think rationally and act rationally. 

Four Categories of Definition of Artificial Intelligence 

Thinking Humanly
Declaring that a machine thinks like a human, we should have some sets of tests or methods to determine that it is actually thinking humanly. Also we need to know how a human actually thinks while performs any action. This is done by three approaches and while performing these aspects some domains of artificial intelligence pops up to perform those operations and they are: through Introspection: trying to catch our own thoughts as they go by, through psychological experiments: observing a person in action, through brain imaging: observing the brain in action. These tasks cannot be performed by a machine so easily. In order to perform such actions different domains are established in the field of AI and they are:
  • Cognitive Science:- a fascinating field in itself, worthy of several textbooks and at least one encyclopedia (Wilson and Keil, 1999).

Acting Humanly
If a machine could pass the Turing Test then the observer would fail to determine whether the assigned test was performed by a machine or a human being. Machine should possess following domain knowledge in order to do so:
  • Natural language processing:- to be able to communicate with real world.
  • Knowledge representation:- to store the information seen, heard or written and prepare them for proper utilization.· Automated reasoning – to use the represented knowledge for answering questing on a real time basis.
  • Machine learning:- being able to incorporate the real environment and the changes for performing its best.
  • Computer vision:- to perceive objects
  • Robotics:- to manipulate objects and move about.

Thinking Rationally
Thinking rationally discusses about the approaches that a human apply for reasoning, providing logic during a problem solving. Rationality in a sense that provided a machine to solve a complex problem it should not only follow the steps according to the category of a question but should be able to find out the rational ways to choose solutions nearer with least loops and costs. This category of definition is widely adapted from the foundations like psychology and philosophy which entirely focuses on proper reasoning, analyzing the process of mind during thought creation and rational thinking.

Acting Rationally
This category of definition of Artificial Intelligence deals with controlling and performing the rational thinking created by a machine or by a rational agent. Therefore this category of definition is also entirely dependent upon foundations like psychology and philosophy.

Hence, after dividing the definitions of Artificial Intelligence into four categories we came to know that the first two categories Thinking Humanly and Acting Humanly is solely derived from the foundations of Mathematics, Economics and Computer Engineering. While the last two definitions are derived from the foundations of psychology, neuroscience, philosophy, control theory and linguistics.

References
[1] “Artificial Intelligence A Modern Approach, Third Edition, Stuart Russel, Peter Norvig, 2010, Pearson Education Inc., ISBN: 9780136042594
NOTE: This single reference is more than enough for referencing because the book alone is prepared by referencing more than 2000 Research papers..

Relevant Articles

Can Computer suppress Human Intelligence?


Picture from
Computers today have achieved attention of the global audience to an extent never before seen on this earth. We humans have amazing capabilities that computers can't dream of, for example we are extremely brilliant to identify what we DON'T know. If you ask someone if s/he knows what is the meaning of 'Mameluke', s/he will answer immediately 'no' if s/he doesn't know it. Just that easy for us. But ask the biggest database in the world for any kind of information, and it will need time to search the full database just to identify if it has the information or not.And with this attention comes the power to influence people's thinking which has indeed given a raise to an open question: Are computers smarter than a human brain? Are they more intelligent? 
Computers will never be as smart as humans because they can only do what a programmer “teaches” them to do, and the programmer cannot teach them more than a subset of what he himself knows as "an intelligent creation has to have an intelligent creator" but Why is the idea wrong? Because it assumes that an intelligent creation has to have an intelligent creator.an,an intelligent creation has to have an intelligent creator
On our first thought, we might think that computer are superior to human brain. Computer can perform complex task very quickly, it has a huge storage and does not forget information as easily as human brain. Computer is not tired performing complex string of tasks while we human need to take a coffee break every couple of hour to keep our mind functioning properly.
However, these task are virtue of faster processing speed and diligence rather than intelligence. Computer are not able to adapt and learn new things like human brain. Our brain is capable of analyzing new and unfamiliar system in a way computer cannot. It can also recognize complex patterns while in computer it is limited to only superficial level. Also, computers today are not capable of learning from past experience as us human and rely mainly on the information it gets during the operation and that limitation of computer make it inferior to human brain.
Even though, the task of creating an intelligent computer have eluded computer scientists and engineers for decades they have not given up. Moreover, they have been succeeded in it more than ever before. We have systems that can understand narrow subsets of human language, solve problems in well-defined areas of knowledge, navigate restricted real world spaces without human assistance, and perform other limited tasks. Also, Scientist in IBM developed a super computer Deep Blue that defeated world chess champion Garry Kasparov in 1997. This feat is considered to be most significant breakthrough in task of creating truly intelligent machine.
As much as we have talked about intelligent computers there must be a method for distinguishing a whether a machine is intelligent or not. There are several method for testing this and the most influential but also highly criticized is called the Turing test, developed by Alan Turing in 1950. It requires a judge to engage in natural conversation with a human being and a machine which is designed to generate performance indistinguishable from that of human being. If the judge cannot reliably tell the machine from the human, the machine is said to have passed the test. A program named ELIZA, developed by Joseph Weizenbaum in 1966, is claimed by some programmer to have passed the Turing Test but this view is highly contentious. Till today scientists and engineers are developing program that can successfully pass the Turing test with any contention.
Development of intelligent system has come a long way, near intelligent system which can perform various task without human interference has been developed and such system is being called Artificial Intelligence (A.I.). An autonomous vehicle developed by DARPA (Defense and Research Project Agency) of US and Carnegie Mellon University drove almost 2850 miles from Washington DC to San Diego without human help. PEGASUS, a spoken language interface connected to American Airlines is used to obtain flight information and make flight reservation over the phone. Major Smart-phone companies are equipping their device with personal assistant which can perform near human conversation and learns and displays information according to how user use their device.
A number of artificial intelligence (AI) enthusiasts have been predicting that machines will eventually surpass human intelligence in the foreseeable future. In some cases, when tasks are easily broken down into sequential, smaller arithmetical components, computers outperform the human brain, reflecting the lack of adaptation of the later to such problems. Despite this, the current overall consensus in the field is that computers still need to evolve significantly to match the efficiency and complexity of the brain. Unlike computer power, which is very well defined, we are still far from understanding how to express the processing power of the human brain.
At the fundamental level, the intelligence of machines, as well as the cognition of the brain is a based on the storage of information and electrical currents conducted through highly specialized and sophisticated architecture of wires. In machines, these signals flow at the speed of light while in the brain, the velocity of impulse conduction in the axons is between 0.2-120 m/s. Supercomputers have approached roughly the estimated computing power of the brain, several petaflops, but at an enormous cost: about 10 million watts of power consumption versus 20 watts in the brain.
However, with development of intelligent machine human being will have to prepare for new changes socially and mentally as well. Humans are used to seeing themselves as the only intelligent beings in the world. If we are going to deal with truly intelligent computers, we will have to adjust very deep beliefs about our uniqueness and superiority. For intelligent computers will not only possess human-level capabilities, they may be functionally immortal and ultimately possess powers beyond our ken. Initially, they may be dependent upon us, but, if we do not deliberately impede their progress, they could become autonomous in relatively short order. Under these circumstances, it is unclear whether they would integrate into our society or form a parallel civilization. As mechanical beings, intelligent computers could potentially modify themselves much more rapidly than humans and evolve at a faster pace.
For as long as computers have been around, they've been doing things better than humans. But we cannot deny the fact that when logic is concerned, we humans rule. So for today, in the question of which is more intelligent, Human brain takes away the title. Computers are only as smart as their users and programmers are.

Refrences:


[VLOG] EXPLORE GADGETS: A.I Experiments with Google

Explore Gadgets

Explore Gadget is a trending YouTube channel which provides informational videos, unique gadgets ,tech accessories, android apps & games, leaks, rumors, comparison, iOS apps & games, tutorials, how-to and many more. On top of that this is the only tech channel in Nepal with such a huge number of subscribers and 76M+ views and growing.

Out of 600+ Videos from this channel, I chose this particular video. The only reason for choosing the video is the A.I. Experiment projects which I was unaware. Thank you Explore gadgets for sharing information about Artificial Intelligence, which is what I love the most.
Explore Gadgets Brings you a collection of cool AI Projects which you can explore on your free time. All of us know Google's most expensive asset is it's database. It has abundant amount of data so that it can do research in Artificial intelligence Projects. All these projects presented in the video are from A.I Experiment With Google. This video will show you some of the Google Projects which are actually fun. Following projects are included in the video:
  • Giorgio Cam
    Take a picture to make music with the computer.
  • Quick Draw
    A game where a neural net tries to guess what you’re drawing.
  • The Infinite Drum Machine
    Thousands of everyday sounds, organized using machine learning.
  • Bird Sounds
    Thousands of bird sounds visualized using machine learning
AI Experiment has some more projects which you can try. Best part of these projects is that you can get the source code of those projcets and do more geeky stuffs and explore your skills in Machine Learning.

Follow Techjhola and SUBSCRIBE  to EXPLORE GADGETS for more Technology Stuffs.
Why Should Artificial Intelligence Enthusiasts Watch Westworld?

Why Should Artificial Intelligence Enthusiasts Watch Westworld?

It's all over the internet among the list of Trending. It's trending on QuoraRedditMedium and also on 4chan. It has got 9.2/10  IMDB rating. Any social networking site, or Q&A Channel, it's every where. So what is so special about it that it's rated 9.2 in its first season?


WestworldA series inspired by the 1973 film of the same title written by Michael Crichton about a futuristic theme park populated by artificial beings.-(IMDB). Well the movie is very fascinating and you will appreciate the way Crichton presents his artificial park.


The series indeed itself is a miraculous presentation to the viewers because the series is long with different stories within. Every characters are simple and are giving their best performance. As a viewer I got this awesome experience of feeling the dream of artificial 
intelligence come true. 

One evening while I was switching TV I saw this part of the series on HBO " The keyboard was playing on its own, the music of keyboard was unique, I've never heard such music before. And within a minute of two a girl(Delores) who was somewhere around this cowboy street(The Westworld) , suddenly finds herself inside a room of transparent glass walls everywhere. This guy named Bernard says "Analysis" and at a certain she becomes statue and starts telling truth" I was feeling like " What the hell was that. What did I just watched? So freaking awesome. What is it? A Magician's show or some kind of super natural power ?" I waited till the first break or the name of the series was written on top right of the screen. I Noted down its name on my phone and turned on my laptop. I Searched the movie on torrentz and find out it was the second episode of Season One. With a big sigh I downloaded both of them and started watching it.
It was totally worth it. It also freaked me out because I discuss so many times with so many people about Artificial General Intelligence.  Before moving further I will point out: 

Why Artificial Intelligence Enthusiasts should watch Westworld?
  • Artificial Machines passes Turing Test during the first year of its development
    Well to know this particular point from it's creator "FORD" you need to watch 7 episodes. But logically saying, the proposal of Turing Machine " Being unable to distinguish between a man and a machine" I had no idea that all the characters Delores, the whole people inside the bar, those cowboys, horse etc. were all artificial machines, which are controlled by some bunch of people. Every machines also called as agents are self aware, response immediately towards any change or inquiry, speak fluently with feelings, humorous, cry, fall in love care for their loved ones etc. I saw nothing different before I knew they were controlled by humans.
     
  • Everything you Dreamed about AI & AGI it is here.
    The technological advancement through the implication of Artificial Intelligence which we dream about are all here working on its optimum level. Machine Learning, Optical Character recognition, Sentiment Analysis, Machine Vision, Robotics, Augmented Reality, etc. You name a technology if it was used here and you see it's being used in some episode.
    A Human looking robot talks to another human looking robot on a day to day basis and both of them talk, interact so efficiently and fluently that an actual human in that environment also cannot distinguish them as machine or human.
  • Westworld can inspire AI Geeks for new sets of tests and hypothesis
    What we do today to practice Tests in Artificial Intelligence projects is fetch data to the machine and retrain them for numerous hours, then finally we get high efficiency. But here the robots are interacted with real environment. Neither the machines are aware that they are machines not the humans know their exists machines among them. They are doing their day to day stuffs and what the observer/ or their developers are doing is analyzing their emotions, researching why are they acting differently. Many weird things are achieved in this series which are still in the concept papers of AI Researchers.
  • It's entertaining to watch how agents are drifted towards the anomaly and repeatedly get nowhere nearer towards the maze
    Turing test for this series if unveiling the maze. A maze s developed by Arnold and Ford, if the agents will be able to locate the maze and the secret behind it then Agents will be set free. The mystery box "Maze" is yet to be unveiled but this one test is never passed by any of the agent. So the journey to be nearer to the maze seems challenging and every character use their intelligence to reach nearer to the maze.

    Delores, Bernard, Arnold, Hector, The Maze, Ford, First Generation Agents, Maeve, etc. 
    These are some among few characters whose stories are mysterious and worth it to watch the whole season. 
Canadian Mafias Who made  AI Dream alive with their contribution in Artificial Neural Network

Canadian Mafias Who made AI Dream alive with their contribution in Artificial Neural Network


An Artile was published on July 2015 at RECODE [1] which introcuded some of the AI Pioneers of Silicon Valley. The article called them as the “Canadian Mafias “ of Silicon Valley, which was fascinating. 
Based on the story I am highlighting the players in Today’s AI a.k.a CANADIAN MAFIA'S. They may be already listed in some answers but this answer happens to be a profile sharing in curiosity.

  1. Tomi Poutanen [2] 
    CoFounder and CTO of Milq.Inc, Tomi Graduated did his Dissertation on Neural Network from University of Toronto.
  2. Geoffrey Hinton [3]Insctuctor of Tomi Poutanen and also known as the Godfather of neural network is one of the first researchers who demonstrated the use of generalized backpropagation algorithm for training multi-layer neural nets. He is an important figure in the deep learning community.
  3. Yann LeCun [4] and Yoshua Bengio [5] and Hinton were three visionary Researchers who continiued their research on Netral Netowrks in their own labs even in the demoralizing AI WINTER was going on.
  4. Rob Fergus [6] Associate Professor at NYU former LeCun colleague is working as Research Scientis at Facebook AI Research.
  5. Juergen Schmidhuber [7], a Swiss AI researcher also know as vocal critic of the Canadian Mafias has generated many Researchers in the field of AI.
The effects and advancement in the field of AI will soon be realized all over the world
Footnotes
TOP 5 BIGGEST CHALLENGES ON ARTIFICIAL INTELLIGENCE

TOP 5 BIGGEST CHALLENGES ON ARTIFICIAL INTELLIGENCE

Sanjog SigdelSomewhere between Skepticism and Optimism


I don’t have enough experience in Artificial Intelligence to provide you sufficient and promising answer buy as per my understanding:
  • An Artificially Intelligent Machine has achieved some level of accuracy in Image Processing, machine learning is already 97% accurate in comparison to Human, sound recognizition is also on its hipe, Deep Learning has increased the performance of many tech giants, Internet of Things today is all about being of service with some intelligence.

    But:
    Artificially Intelligent Machine can watch a music video process each image frame and at the same time can recognize the voice and the stanzas. But it fails to perceive the sensation of the music and also lacks the knowledge about what message is the song trying to give.

    Computer Science still needs much work in combining the existing achievement and make them a composite intelligence first before moving further.
  • Defining AI: May be there is still not a concrete vision about AI and what sort of Intelligence are we expecting from a machine. Everyone seems to be applying intelligence in their machine. But I would love to call most of such machine are just solving problems by fetching the solutions(patters) stored in their memory by their programmer).
  • We also fail to feed the knowledge about cognitive science in a machine.
  • Human Brain is a mystery box to lean into and dig deep for understanding how it performs the computation in fraction of second. And having understood in detail about human brain scientists replicate neural networks. It works very efficiently than anything better, but still solving the mystery of human brain could lead the breakthrough in AI
  • The architectural design of the microprocessor limits the computation speed when it comes to the billions of information processing to answer a basic query.
    (Read this point here somewhere in Quora)
Will advances in Artificial Intelligence lead to technological unemployment?

Will advances in Artificial Intelligence lead to technological unemployment?


Some weekends are well spent exploring new ideas and news about something you love. This weekend I again read some of my bookmarked answers on +Quora all of them were about Artificial Intelligence, Artificial general intelligence , Cognitive thinking, human psychology etc. Link provided below is my answer on quora about "If Artificially Intelligent technologies will lead to technological Unemployment". Also there are some more interesting answers by rest of the enthusiasts. Do provide your feedback.
Read Sanjog Sigdel's answer to "Will advances in Artificial Intelligence lead to technological unemployment?" on Quora

PS:
  1. Sharing the link about my own answer might seem unlikely. But I wanna mention that this sharing is done just in Curiosity to add new node to my Network of like minded people who like to share their comments about Artificial Intelligence. 
  2. If sharing the links violates any legal terms of Quora, I will immediately delete this article.




Is a Machine Intelligent enough to predict Homo Sapiens' Future?

Motivation

Science Fiction is all over me, maybe it's the long addiction of watching "FLASH" and other such series and movies. Characters from fiction movies and series predict somebody's future or travel back to the past as simple as blinking an eye. This astounds me many times and I keep thinking if our future is really near to such fiction stories which we kind of think as a bluff. Also day by day we are getting to read advancements in Artificial Intelligence. So I'm curious if I will be alive to see such man made miracles into a machine.

Concept

Provided sufficient amount of data about the present action we are performing, With the help of essential statistical regression models, we can predict the result in future. This is done in mathematics. And a machine is just a bot to perform the rules defined in it. So if numbers can be predicted through mathematics, why cannot a machine predict future of mankind? Prediction includes (movement, feelings, action and reaction, etc...).

Approach

With reference to mathematics let us Consider if a Company X gains an income of amount a in month 1, b in month 2, c in month 3, d in month 4, then with some mathematical calculations( probability, fuzzy logic, regression analysis, any ) we can say that the company will earn the income e in month 5, and so on. Here the input data is some numbers and constraints for calculation is income in next time period.
Now its could be weird to put my thinking in a practical way, but let's take above mathematical modal, if movement of our hands and legs, our anger and stress, sentiments and etc are recorded in various time interval. Could an Intelligent machine predice what will be our next movement, sentiment and decision? Human being a.k.a Homo Sapiens are also some machine(Say an advanced machines, consisting of skin and which breathes, digests food, breathes air senses to the light.Yes humans are far more intelligent and are complex to understand. But Human perform their activities according to the results or guidelines given by the neurons.


So now if we will observe a human being for decades, and note down the activities performed. A Machine can definitely observe some similar patterns like certain sets of actions human beings are likely supposed to provide similar pattern of outcome. Analyzing those patterns I believe a macine can predict the future actions as well as the actions of those activities as well.In this condition we can say there is a high probability for a machine to predict our actions. We can consider many sentiment analysis projects which have been successful in predicting what our next tweet will be.
Now comes the serious part which is not less than a fiction. Can an intelligent machine predict what our action will be in our future. Can we rely on a machine which will be predicting our actions after let's say 10 years? Wouldn't it be more similar then believing to a fortune teller?

Constraints

Wow how easily a Machine predicted our future. But here are some serious parts which are major concerns for a machine to claim as an intelligent machine strong enough to predict Homo Sapiens future.
  • First we still are learning how our Brain works. Even though Machine Learning and Neural Network has progressed, but we have yet many mysteries to unfold about our brain. Researchers are working on it, but the progress is slower then we expect such science fiction to become real.
  • The hormones in a human body keeps changing time to time, our bones, skin and other parts keeps changing time to time. We have no control to it. This phenomena of which we are unaware, on how things happens, how can a machine unveil the secret within our physical system?
  • Consciousness of a Human Being, which is a mystery to us. We donot know, well I don't know how my body maintains its cycle of 24 hours in a day. How do we feel, taste or breathe the environment. So Predicting our future also means predicting our end. Which I think will never be achieved by a machine.