JOIN US×
ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Over the last few years, advancements in the field of Artificial Intelligence (AI) has constantly been in the headlines. These advancements have, in fact, revolutionised many industries. AI has also become increasingly ubiquitous — and its applications are visible even in many of our daily online activities.
Right from tagging friends in Facebook photos to managing your Amazon shopping cart, AI algorithms play a crucial role in making the magic look mundane. This is just the tip of the iceberg — the rate at which the AI sphere is evolving looks to continue, if not increase!
Latest developments
What are the latest AI developments? As AI is an extremely vast field, here are some of the latest research ideas in a few AI sub-fields:
Computer vision: Computer vision is the extraction and processing of content from images and videos. Computer vision, in conjunction with robotics, is being used for a variety of applications, most notably in self-driving vehicles, through a combination of convolutional neural networks for object detection, object collision algorithms and predictive modelling. Some computer vision-related focus areas include computational photography (capturing photographs which cannot be taken with regular cameras).
Reinforcement learning: Reinforcement Learning (RL) refers to learning through trial-and-error, by rewards and punishments, similar to the traditional human learning methods. RL is used in the design of software-based game-playing agents. RL agents essentially observe the environment and choose actions that maximise payoff. Google DeepMind’s AlphaGo, based on RL, in 2016, was the first computer program to beat a professional Go player. There is also a growing interest in inverse reinforcement learning in Markov decision processes, that is, the problem of extracting a reward function given observed, optimal behaviour.
Natural language processing: Natural Language Processing (NLP) refers to the ability of computers to analyse and understand human inputs, both text and speech. One of the biggest challenges in NLP is training computers to understand languages while accounting for domain knowledge and ambiguity.
Long Short Term Memories are being used to create character-level models with reasonably high degree of accuracy.
The future of AI
The possibilities in using AI to make the world a better place are endless. AI could even be instrumental in solving climate change and finding a cure for many diseases. While AI will continue to make our lives easier, we must exercise control. Many eminent technologists, most notably Elon Musk and Stephen Hawking, have already spoken of the dangers of AI.
However, the pop culture notion that AI will wipe out humanity seems very far-fetched, at least for now. This is not to say that machines won’t grow smarter. Machines may be able to perceive their own existence soon. After all, any computer smart enough to pass the Turing test (a test for intelligence in a computer) could also be smart enough to know to purposely fail it. Here are some of the advances that experts believe will take place in the field of Artificial Intelligence (AI):
* Language learning algorithms will be at the forefront of commercial AI.
* Programmers and engineers are expected to improve how we positively reinforce AI programs.
* AI’s ability to predict the future will significantly improve in the coming years.
*It can be used to glean new insights from existing data like electronic health records.
Right from tagging friends in Facebook photos to managing your Amazon shopping cart, AI algorithms play a crucial role in making the magic look mundane. This is just the tip of the iceberg — the rate at which the AI sphere is evolving looks to continue, if not increase!
Latest developments
What are the latest AI developments? As AI is an extremely vast field, here are some of the latest research ideas in a few AI sub-fields:
Computer vision: Computer vision is the extraction and processing of content from images and videos. Computer vision, in conjunction with robotics, is being used for a variety of applications, most notably in self-driving vehicles, through a combination of convolutional neural networks for object detection, object collision algorithms and predictive modelling. Some computer vision-related focus areas include computational photography (capturing photographs which cannot be taken with regular cameras).
Reinforcement learning: Reinforcement Learning (RL) refers to learning through trial-and-error, by rewards and punishments, similar to the traditional human learning methods. RL is used in the design of software-based game-playing agents. RL agents essentially observe the environment and choose actions that maximise payoff. Google DeepMind’s AlphaGo, based on RL, in 2016, was the first computer program to beat a professional Go player. There is also a growing interest in inverse reinforcement learning in Markov decision processes, that is, the problem of extracting a reward function given observed, optimal behaviour.
Natural language processing: Natural Language Processing (NLP) refers to the ability of computers to analyse and understand human inputs, both text and speech. One of the biggest challenges in NLP is training computers to understand languages while accounting for domain knowledge and ambiguity.
Long Short Term Memories are being used to create character-level models with reasonably high degree of accuracy.
The future of AI
The possibilities in using AI to make the world a better place are endless. AI could even be instrumental in solving climate change and finding a cure for many diseases. While AI will continue to make our lives easier, we must exercise control. Many eminent technologists, most notably Elon Musk and Stephen Hawking, have already spoken of the dangers of AI.
However, the pop culture notion that AI will wipe out humanity seems very far-fetched, at least for now. This is not to say that machines won’t grow smarter. Machines may be able to perceive their own existence soon. After all, any computer smart enough to pass the Turing test (a test for intelligence in a computer) could also be smart enough to know to purposely fail it. Here are some of the advances that experts believe will take place in the field of Artificial Intelligence (AI):
* Language learning algorithms will be at the forefront of commercial AI.
* Programmers and engineers are expected to improve how we positively reinforce AI programs.
* AI’s ability to predict the future will significantly improve in the coming years.
*It can be used to glean new insights from existing data like electronic health records.
ADVERTISEMENT
(Published 21 August 2017, 16:08 IST)
ADVERTISEMENT
ADVERTISEMENT