https://deepmind.com/about/
This company in the UK founded in 2010 is a world leader in artificial intelligence and research for it. The company is doing research on programing the AI to solve complex problems on its own with reinforced learning. The companies goal is to use this to solve issues from healthcare to climate change. CEO, Dr Demis Hassabis, has a degree in computer science and a PHD in cognitive neuroscience at UCL and postdoc's at Harvard and MIT. The Deep mind AlphaGo program was able to beat the worlds best Go player. It was also able to teach itself to play and win 49 different atari games with just pixels as input this shows the complex algorithms ability to solve problems faster than a human. The NHS that is part of Deep minds healthcare information helps get patients from test to treatment is less time and greater accuracy. It also alerts nurses of patients condition via alerts on a mobile application. All of the patients information is encrypted for data protection and is not connected to any accounts in the cloud. Deep mind has worked with google to help them reduce costs for their data centers electrical bill and energy consumption by 40%. The waveNet technology is used to generate voices for google assistant. This allows humans to talk and interact with machines to give commands to look info up or to do whatever task it is able to do. And the computer will talk back with a response based off of what you asked or told the assistant. This tech has gotten more advanced over the past few years with new emerging tech. The ethics surrounding this technology is important to look at as well. The Rules and regulations for this program are implemented with a few things in mind. That they are for the global betterment and to improve lives. That the program is under meaningful human control and its impact on society is beneficial. Openness, transparency and freedom of thought are all things that guide the projects. The company has over 200 publications ranging from psychology labs for deep reinforcement learning to neural networks and how they work with this technology. A neural network is a computer program that is able to preform a given task and remember the outcome of old data so it is able to learn and adapt from the previous information to make better and smarter decisions after each execution. In addition a differentiable neural computer is used to preform pattern matching in games and other scenarios. This computer uses stored memory to answer questions about complex data and other things that need to be solved. This also helps with the quick decision making process for the computer. These machines were designed originally to create and navigate data structures in complexity autonomously with no human correction needed. There is still more research that needs to be done on this part of Artificial Intelligence specifically. Optimization is a key feature used when the computer needs to use stored memory to answer a question. When the answer is not correct it will compare that to what the desired answer is and do that over and over until the answer the computer is outputting is the closest to the desired one or the one specified. This allows a machine to execute this process without any needed programming. The differentiable neural computers are still under heavy research but this is the route most of it is taking and what the goal is, zero to limited programming or human interaction with the most precise results. There are also projects that are ongoing that are looking into the racial bias that some criminal justice algorithms have and the bigger picture or bigger impact that AI has on the world as far as what could be the consequences if the tech was used in the wrong hands. These networks are a combination of links and connections similar to the human brain with some differences. This is at its core how reinforced learning happens with most AI technologies. They are able to keep useful data from past tasks and apply it to new ones simultaneously. There are many faucets and parts to artificial intelligence and the way we interface with technology is used in research on how to improve upon it. Overall this technology's future will be useful because it will enhance the effectiveness of computers and will take care of user error and will be more intuitive and user friendly to people who are not tech savvy. That is the ultimate goal for AI is to bridge the gap between human error and computers. Also to make life easier for people with the use of technology.
Research and info was found at: https://deepmind.com/ on 2/19/18Artificial Intelligence https://en.wikipedia.org/wiki/DeepMind on 3/22/18Artificial Intelligence http://www.dailymail.co.uk/sciencetech/article-5449483/DeepMind-AI-learning-understand-thoughts-others.html 3/22/18Artificial Intelligence http://www.bbc.com/news/technology-41668701 on 3/22/18Artificial Intelligence
https://deepmind.com/about/
This company in the UK founded in 2010 is a world leader in artificial intelligence and research for it. The company is doing research on programing the AI to solve complex problems on its own with reinforced learning. The companies goal is to use this to solve issues from healthcare to climate change. CEO, Dr Demis Hassabis, has a degree in computer science and a PHD in cognitive neuroscience at UCL and postdoc's at Harvard and MIT. The Deep mind AlphaGo program was able to beat the worlds best Go player. It was also able to teach itself to play and win 49 different atari games with just pixels as input this shows the complex algorithms ability to solve problems faster than a human. The NHS that is part of Deep minds healthcare information helps get patients from test to treatment is less time and greater accuracy. It also alerts nurses of patients condition via alerts on a mobile application. All of the patients information is encrypted for data protection and is not connected to any accounts in the cloud. Deep mind has worked with google to help them reduce costs for their data centers electrical bill and energy consumption by 40%. The waveNet technology is used to generate voices for google assistant. This allows humans to talk and interact with machines to give commands to look info up or to do whatever task it is able to do. And the computer will talk back with a response based off of what you asked or told the assistant. This tech has gotten more advanced over the past few years with new emerging tech. The ethics surrounding this technology is important to look at as well. The Rules and regulations for this program are implemented with a few things in mind. That they are for the global betterment and to improve lives. That the program is under meaningful human control and its impact on society is beneficial. Openness, transparency and freedom of thought are all things that guide the projects. The company has over 200 publications ranging from psychology labs for deep reinforcement learning to neural networks and how they work with this technology. A neural network is a computer program that is able to preform a given task and remember the outcome of old data so it is able to learn and adapt from the previous information to make better and smarter decisions after each execution. In addition a differentiable neural computer is used to preform pattern matching in games and other scenarios. This computer uses stored memory to answer questions about complex data and other things that need to be solved. This also helps with the quick decision making process for the computer. These machines were designed originally to create and navigate data structures in complexity autonomously with no human correction needed. There is still more research that needs to be done on this part of Artificial Intelligence specifically. Optimization is a key feature used when the computer needs to use stored memory to answer a question. When the answer is not correct it will compare that to what the desired answer is and do that over and over until the answer the computer is outputting is the closest to the desired one or the one specified. This allows a machine to execute this process without any needed programming. The differentiable neural computers are still under heavy research but this is the route most of it is taking and what the goal is, zero to limited programming or human interaction with the most precise results. There are also projects that are ongoing that are looking into the racial bias that some criminal justice algorithms have and the bigger picture or bigger impact that AI has on the world as far as what could be the consequences if the tech was used in the wrong hands. These networks are a combination of links and connections similar to the human brain with some differences. This is at its core how reinforced learning happens with most AI technologies. They are able to keep useful data from past tasks and apply it to new ones simultaneously. There are many faucets and parts to artificial intelligence and the way we interface with technology is used in research on how to improve upon it. Overall this technology's future will be useful because it will enhance the effectiveness of computers and will take care of user error and will be more intuitive and user friendly to people who are not tech savvy. That is the ultimate goal for AI is to bridge the gap between human error and computers. Also to make life easier for people with the use of technology.
Research and info was found at: https://deepmind.com/ on 2/19/18Artificial Intelligence
https://en.wikipedia.org/wiki/DeepMind on 3/22/18Artificial Intelligence
http://www.dailymail.co.uk/sciencetech/article-5449483/DeepMind-AI-learning-understand-thoughts-others.html 3/22/18Artificial Intelligence
http://www.bbc.com/news/technology-41668701 on 3/22/18Artificial Intelligence