artificial intelligence News for August 27 2017

Google CEO Sergey Brin, “I didn’t see Artificial Intelligence coming. How wrong was I?”
http://www.youtube.com/watch?v=30rx3dBPbIs
In Davos 2017, “I didn’t pay attention to AI at all, to be perfectly honest. Having been trained as a computer scientist in the 90s, everybody knew that AI didn’t …

Our fear of artificial intelligence? It is all too human

The classic sci-fi fear that robots will intellectually outpace humans has resurfaced now that artificial intelligence is part of our daily lives. If we are to worry about a robot takeover it is not because artificial intelligence is inhuman and immoral, but rather because we are coding-in distinctly human prejudice. As the Tay disaster revealed, artificial intelligence does not always distinguish between the good, the bad and the ugly in human behavior. The type of artificial intelligence frequently used in consumer products is called machine learning. If we wanted the artificial intelligence to correctly identify cars, then we’d teach it what cars looked like by giving it lots pictures of cars. If all the pictures we chose happened to be red sedans, then the artificial intelligence might think that cars, by definition, are red sedans. If we then showed the artificial intelligence a picture of a blue sports utility vehicle, it might determine it wasn’t a car. When there is bias in the data used to train artificial intelligence, there is bias in its output. In turn, the artificial intelligence learned to call black defendants criminals at an unfairly higher rate, just like a human might. That algorithm-fueled artificial intelligence amplifies human bias should make us wary of Silicon Valley’s claim that this technology will usher in a better future. I walked into room the other day to a man yelling, “Alexa, find my phone!” only later to realize he was talking to his Amazon Alexa robot personal assistant, not a human female secretary. Adding diversity to product teams alone will not counteract the systemic nature of the bias in data used to train artificial intelligence. Careful attention to how artificial intelligence learns will require placing antibias ethics at the center of tech companies’ operating principles – not just an after-the-fact inclusion measure mentioned on the company website. Now that artificial intelligence makes serious, humanlike decisions, we need to hold it to humanlike moral standards and humanlike laws.
more…

Elon Musk Predict Exactly When Artificial Intelligence Will Overtake Human Intelligence

Given the speed at which researchers are advancing artificial intelligence, the question has become not if A.I. will become smarter than its human creators, but when? During May and June of 2016, they polled hundreds of industry leaders and academics to get their predictions for when A.I. will hit certain milestones. The findings, which the team published in a study last week: A.I. will be capable of performing any task as well or better than humans-otherwise known as high-level machine intelligence-by 2060 and will overtake all human jobs by 2136. Those results are based on the 352 experts who responded. Monday night, Elon Musk, who’s been a consistent A.I. fear monger, chimed in on Twitter. The entrepreneur followed up his tweet with an ominous, “I hope I’m wrong.” Musk has been a vocal critic of A.I. the past several years, painting nightmare scenarios in which it becomes weaponized or outsmarts humans and leads to their extinction. He co-founded OpenAI, a non-profit that aims to ensure A.I. is used for good, in 2015. Musk’s own firm, Tesla, is one of the companies leading the charge in creating self-driving vehicles. The trucking and taxi industries employ about 2 million Americans, all of whom could soon find their jobs obsolete should vehicles become fully autonomous. The experts polled in the study predicted that A.I. would become better at driving trucks than humans in 2027. A.I. will surpass humans in a number of other milestones, the experts suggested: translating languages, writing high-school level essays, and performing surgeries. An A.I. system created by scientists at Carnegie Mellon won $2 million from top poker players in a tournament in January. It’s worth noting that the predicted timelines did not vary based on the experts’ levels of experience with artificial intelligence. One variable that did correlate with the predictions was the location: North American experts thought A.I. would outperform humans on all tasks within 74 years, while experts in Asia thought this would take only 30 years.
more…

5 unexpected sources of bias in artificial intelligence

The reality is that not only are very few intelligent systems genuinely unbiased, but there are multiple sources for bias. These sources include the data we use to train systems, our interactions with them in the “Wild,” emergent bias, similarity bias and the bias of conflicting goals. For any system that learns, the output is determined by the data it receives. Most recently, this kind of bias has shown up in systems for image recognition through deep learning. Learning systems used to build the rules sets applied to predict recidivism rates for parolees, crime patterns or potential employees are areas with potentially negative repercussions. While some systems learn by looking at a set of examples in bulk, other sorts of systems learn through interaction. As we build intelligent systems that make decisions with and learn from human partners, the same sort of bad training problem can arise in more problematic circumstances. What Tay taught us is that such systems will learn the biases of their surroundings and people, for better or worse, reflecting the opinions of the people who train them. Decisions made by systems aimed at personalization will end up creating bias “Bubbles” around us. As we look to social media models as a way to support decision making in the enterprise, systems that support the emergence of information bubbles have the potential to skew our thinking. Sometimes bias is simply the product of systems doing what they were designed to do. Imagine a system, for example, that is designed to serve up job descriptions to potential candidates. The system generates revenue when users click on job descriptions. By understanding the bias themselves and the source of the problems, we can actively design systems to avoid them.
more…

Artificial Intelligence Deals Tracker

The definitions of what artificial intelligence is and what it’s capable of accomplishing have been constantly changing over the years with corresponding changes in technological capabilities. Access to massive amounts of data and advanced hardware processing capabilities have ushered in a new era of AI applications. Most of it is still a lot of number crunching, taking into account more variables than a human possibly can to categorize new data and predict trends, among other things. If it only follows a command to do a specific task, it’s not a self-learning algorithm. If it is constantly learning and improving its answers as you interact with it, you can call it an AI-powered bot. Is Microsoft’s Clippy an AI bot? Maybe 20 years ago, it seemed like an intelligent interface. Machine learning: Machine learning is a set of algorithms used to make a system “Artificially intelligent,” enabling it to recognize patterns from large datasets and apply the findings to new data. Deep learning is a subfield of machine learning, which uses several layers of neural networks. Machine learning can be used to train computers to understand and analyze human language, including text and voice, to identify and analyze images, or for time series analysis, among other things. The algorithms then recognize the object, its attributes, look for other images with similar attributes and then give you suggestions on where to buy it from. This may seem like a very simple, “Is this really AI” kind of task, but here are some slides on Google’s deep learning evolution that explain how its image recognition abilities improved over time. Other advanced vision systems are used in robotics and autonomous cars for object detection and collision avoidance. Natural language processing/generation: Understanding and/or interacting in human language. Apart from chatbots mentioned above, NLP is used in voice-enabled smartwatches and smart home applications, context-specific searches, finding semantic similarities of words and phrases, among other things.
more…

7 trends for artificial intelligence in 2016: ‘Like 2015 …

We’ve seen startling moves in artificial intelligence in 2015. Robots are doing the grunt work in factories. Driverless cars have become a reality.
more…

The disruption and promise of artificial intelligence | CIO

There’s no shortage of books, news articles and comments in social media about how artificial intelligence (A.I.) is shaping our future. Although it’s …
more…

Artificial Intelligence Now :: AI Now

AI Now will be hosting our second annual symposium, July 10th 2017 at MIT Media Lab. In the meantime, we invite you to peruse our archive of video, topic primers, and more from AI Now’s 2016 Symposium, which AI Now co-hosted with NYU and the Obama White House’s Office of Science and Technology Policy.
more…

Artificial intelligence and cognitive computing: what, why and where

Although artificial intelligence is here since a long time in many forms and ways, it’s a term that quite some people, certainly IT vendors, don’t like to use that much anymore – but artificial intelligence is real, for your business too. Instead of talking about artificial intelligence many describe the current wave of AI innovation and acceleration with – admittedly somewhat differently positioned – terms and concepts such as cognitive computing or focus on several real-life applications of artificial intelligence that often start with words such as “Smart”, “Intelligent”, “Predictive” and “Cognitive”, depending on the exact application – and vendor. Just looking at one context where AI and cognitive are used, Intelligent Document Recognition, there are several forms of artifical intelligence such as semantic understanding, statistical clustering and classification algorithms such as SVM, Bayes and Neural-Net, as Roland Simonis explained in part three of a blog series for AIIM, reposted here, where he tackles how AI helps solve the information and Big Data challenge. Today’s artificial intelligence wave is one of rapid adoption of AI technologies in new applications, driven by, among others the mentioned 3rd platform technologies, including the cloud, faster processing capabilities, scalability, Big Data, the push of various companies in a space where technologies continue to be refined across several applications and industries and, last but not least, market demand for smart and intelligent technologies to leverage the potential of new technologies, information and digital transformation. There are many definitions of artificial intelligence, just as there are many definitions of intelligence.
more…