IBM Deep Blue defeated world chess champion to prove that it is competent enough to rule the cognitive computation world
The California research laboratory of International Business Machines Corporation, located near San Jose, after spending many years as a hardware- centric organization selling mainframes, servers as well as personal computers, the tech corporation has progressed dramatically in the field of cognitive computation.
Deep within the application we use, the medication we take, the food we consume as well as the medical diagnosis done by us, one can find traces of the corporation. Like its Calif research lab, there aren’t obvious signs it is there. Look cautiously though, and the work of the organization- especially from its laboratory- is everywhere.
The cognitive computation business of the Corporation –including analytics, algorithms, machine learning and artificial intelligence – contributed to 35% of IBM’s 81 billion is sales revenue in 2015. It is the rapidlygrowing niche at the New York-based organization, where total sales are decreasing.
The work at Calif played a significant part in this. Hundreds of researchers work at the facility establishing algorithms, artificial intelligence and chips to support more powerful and faster self-learning technologies. It doesn’t feel less like the rest of Silicon Valley technology companies, where ping-pong tables and hover boards are popular sights, and more similar to college- campuses, which don’t have students.
The hallways are clean and quiet. The tenured professors, in the present case, the company’s engineers and scientists do coding in their unnoticeable offices.
The Corporation is already a main player in mainframes and servers - its technology is powering the airline reservation and banking systems. It also has people making efforts in sectors like cloud and cyber safety.
However, to comprehend why it views cognitive computation as its future, it takes one to 2011, when artificial intelligence of the organization, Watson, defeated humans in the game show “Jeopardy."
Winning "Jeopardy" is not anything like winning at chess. An initial IBM artificial intelligence, Deep Blue, defeated the world chess champ 21 years ago by computation of every potential measure and exercising the best option. But for the AI system to even play "Jeopardy", it needed to understand riddles, learn natural language as well as answer questions in clear sentences. Things human beings do.
The organization had continued to research AI since the 1970s, but the victory of Watson was a huge breakthrough. If the organization could teach Watson anything it required to know to win over a difficult game show, what else could it learn?
Shortly after the "Jeopardy" matches, researchers of IBM, including those working in its Calif lab, taught AI system to peruse medical journal abstracts and patent databases.
In drug research, every given molecule can have at most 100 synonyms-- generic names, various chemical strings and brand names.
No comments:
Post a Comment