The machines made by humans have been engaged in defeating humans for some time. Let's know more about it...
Recently, researchers from the Massachusetts Institute of Technology (MIT), developed a gaming bot 'DeepRole' based on Artificial Intelligence (AI) Technology. This bot can win multiplayer games.
Researchers used DeepRole against human players in an online game. DeepRole took participate in about 4000 rounds. And during each and every round bot was winning over human players. These kinds of advancement in science can't be compared by a win of the chess game but programmers tried to make those machines which can beat even humans or stand against humans.
Let's know about more such competitions and challenges in which robots had beat human beings:-
1. In 2011, Watson by IBM (International Business Machine Corporation) defeated two former champions of Jeopardy Challenge.
Watson is a question-answering computer system capable of answering natural language questions, developed by a research team led by lead Researcher David Ferrucci in IBM's DeepQA project. Watson was named after Thomas J. Watson, the founder and first CEO of IBM. To know more about Watson click on this hyperlink.
Jeopardy Challenge! is a classic game show and kind of questions and answer quiz which consists of three rounds. There are three contestants including the previous show's champion which fought against each other. In 2011, of those three contestants, one was Watson, the second one was 74 times jeopardy champion, Ken Jennings and the last one was 20 times jeopardy champion, Brad Utter. This competition lasts for 3 days and the winner was Watson. This showed the power of a supercomputer to the world that IBM had.
2. In 2014, Deep face Facial Recognition by Facebook had mastered in recognizing with 97% accuracy.
DeepFace is a Facebook research group's deep learning facial recognition program. In digital images, it identifies human faces. Facebook Research reported that the DeepFace system is 97.35 percent accurate compared to 85 percent for the Next Generation Identification System of the FBI. Still, some researchers questioned, "Is AI facial recognition is really very comparable to humanized facial recognition?"
3. In 2016, Microsoft's Speech Recognition can transcribe audio with fewer mistakes than human beings.
Speech Recognition is an AI technology that can convert audio input into a machine-readable format. It can identify words, phrases and convert them into text. According to Microsoft, their speech recognition software gives us results with a maximum error of 0.4% whereas the error rate of transcribed audio by human beings was 5.9 percent.
4. In 2017, an AI system performs better than 75% of Americans people in the Visual Intelligence test.
Northwestern University developed an Artificial System, which can solve the logic puzzles just by taking a glimpse of it. In a competition of Visual Intelligence, this program performs better than 75% of American adults. Where most of the focusing on Vision-based AI recognition this program was one step beyond which focuses on visual reasoning.
5. In 2017, Libratus defeats top four world-class poker players in a match.
Libratus is an AI computer program developed by Carnegie Mellon University, Pittsburgh. This was designed to play poker. Libratus defeated top-4 champions in poker's world (Jason Les, Dong Kim, Daniel McAulay and Jimmy Chou). Poker is a game of cards in which the second player is completely unaware of the first player's move. In a game like Poker, the win of an AI bot gets more effective.
3. In 2016, Microsoft's Speech Recognition can transcribe audio with fewer mistakes than human beings.
Speech Recognition is an AI technology that can convert audio input into a machine-readable format. It can identify words, phrases and convert them into text. According to Microsoft, their speech recognition software gives us results with a maximum error of 0.4% whereas the error rate of transcribed audio by human beings was 5.9 percent.
4. In 2017, an AI system performs better than 75% of Americans people in the Visual Intelligence test.
Northwestern University developed an Artificial System, which can solve the logic puzzles just by taking a glimpse of it. In a competition of Visual Intelligence, this program performs better than 75% of American adults. Where most of the focusing on Vision-based AI recognition this program was one step beyond which focuses on visual reasoning.
5. In 2017, Libratus defeats top four world-class poker players in a match.
Libratus is an AI computer program developed by Carnegie Mellon University, Pittsburgh. This was designed to play poker. Libratus defeated top-4 champions in poker's world (Jason Les, Dong Kim, Daniel McAulay and Jimmy Chou). Poker is a game of cards in which the second player is completely unaware of the first player's move. In a game like Poker, the win of an AI bot gets more effective.
6. In 2018, AI by Alibaba defeats human beings in a Comprehension Test held by Stanford University.
There was a comprehension test held by Stanford University between AI by Alibaba and some students of University. After answering approximately ten thousand questions, the score makes the win in favor of AI. AI scored 82.44 while human beings scored 82.304.
LET'S WIND UP
We have just seen seven very major examples of wins of machines over human beings society. Now the question arises that Is AI is really good for human beings? So the answer is everything looks goods and suits till a limit. After that limits, the thing may get useless or very dangerous. AI falls in the dangerous category if it started using limitless. If they can beat their creators then that day is not too far when machines running on AI are ruling on human beings. So AI should be also limited as this is also making the population of a country unemployed which is bad for the country economics. But at the other side Speech Recognition, Image Recognition, etc. are one of the fabulous inventions by programmers of this decade.
0 Comments