Tag

artificial intelligence

Browsing

Recommendation systems – the systems that help services from Amazon to Netflix suggest things you would might like based on your past choices – have a bias problem that is largely because the data comes from users who put their biases into action when interacting with streaming technology. To combat it, more than a decade ago, Netflix announced The Netflix Prize to attract scientists and programmers to write predictive algorithms with the goal of making the company’s recommendation engine 10% more accurate.  The winning prize was $1 million. Team Belkor was the winner due to their inclusion of new factors that recognized a film’s popularity might change over time and that users have certain biases causing them to rate movies differently at different times. The frequency at which people rate movies was also added as a factor. Addressing bias issues to make recommendations better will continue to be a major goal for many in the tech world as it has since The Netflix Prize was awarded.

American Airlines has just launched facial recognition technology as an alternative to boarding passes. They are not the first to use the technology. According to the Washington Post, as of 2018, 14 airports were using facial recognition technology for the boarding process. It’s being used by Lufthansa at Los Angles’ LAX, at Singapore’s Changi airport and London’s Heathrow. Lufthansa commented that it provided “faster and more convenient processes for guests to move through airport.”  While Heathrow Airport claimed that after its 2018 trial, it could be used to “reduce average passenger’s journey time by up to a third.” American Airlines is using facial recognition as an option at Dallas/Fort Worth International Airport.  Passengers will still need a boarding pass and an ID to get through security. Currently the technology is only being used for international flights at Terminal D. The airline called the new service an “additional convenience during the departure process,” though if you still need your boarding pass and ID, how is that more convenient? Beyond convenience, many people have expressed concerns about the technology’s potential for violation of privacy rights. American Airlines said that the photos of people’s faces won’t be saved, but that scanned images will be sent to US Customs and Border Protection. Alarm bells. Debates about the technology will continue as it further integrates into our lives.  Laws that limit potential abuses will be part of the debate
In the never ending search for kid-safe products, the Amazon Echo Dot Kids’ Edition seems to be one that’s carefully designed and surprisingly attractive. This model comes with parental controls, games and skills designed to encourage young imaginations to run wild, but with everything under control. The $70 Echo Dot Kids Edition is the same device in all respects as the third-generation Echo Dot, but it includes an annual FreeTime subscription, Amazon Parental Service, thousands of audible books, hundreds of Alexa games and child-friendly skill-development tools, as well as Amazon Music without the advertising. The FreeTime package costs $36 for Prime members and $60 for non-members. Considering that the standard Echo Dot costs $50 before discounts, you get more for the money with the Kids Edition. Also, note the books and movies available through the FreeTime subscription can be accessed on an Amazon Fire, Kindle, iOS or Android device. The Echo Dot Kids Edition is guaranteed for two years compared to 90-days for the standard Echo Dot.  Good deal.

1Issues such as performance inhibiting gamer-score limits and the inability of players to buy and sell games easily were the least of the problems for Xbox One. What really annoyed users was that conversations at home were recorded and heard by Microsoft’s “quality control” people as revealed in a recent report from Motherboard.

A former contractor from Microsoft informed Motherboard that most of the recordings classified as “personal conversations” were inadvertent – like picking up unintentional Skype calls – and were triggered by players who had given inadvertent orders to Kinect, the voice command system of Microsoft’s Xbox One or Cortina, the voice assistant.

It’s not as if many people at the product’s launch hadn’t voice concerns about privacy. They had, but their concerns were allayed by claims that Kinect for Xbox 360 was designed and implemented with the best possible privacy checks.  Blah, blah. A Microsoft spokesperson has responded to the issues raised since the release of Motherboard’s report, explaining that interception of audio recordings is no longer a necessary quality control step for the company. A few months ago, Microsoft stopped listening to Xbox One audio.  Additionally, the company appears to be working on updated privacy policies to give users more control over how their personal information is stored.

            For many years artificial intelligence wasn’t producing many breakthroughs. However, in recent times AI has finally produced major success stories. AI has already changed the world and its impact won’t slow down anytime soon. AI is set to dramatically transform the global economy and to disrupt lifestyles around the world. Here’s a short list of ten major changes that are expected to come soon because of the AI revolution.

1. Increased Economic Productivity

AI machines are better at certain skills than humans and can get certain tasks done faster. This means that the productivity of labor, business and entire countries is expected to increase with the AI revolution. Given the slow growth rates of major countries in recent years, this is good news.

2. Increased Unemployment and Declining Pay

Because AI is set to automate certain skills, companies are expected to increasingly automate many positions leaving many workers with less work, less pay or even full unemployment.

3. More Time for what Humans are Good At

AI is expected to give humans more time for what we are good at; social leadership and creative skills are not expected to be automated. AI should free up our time for these types of skills.

4. Higher Levels of Inequality

There are significant gaps between different nations, companies and industries when it comes to AI readiness. Companies are expected to make a lot more money while hiring and paying workers less. All of these factors are expected to increase economic inequality between individuals and between countries.

5. Increased Predictive Maintenance

AI is good at making predictions. This means that tasks like medical diagnostics will increasingly be done by AI, while doctors will have more time for human interactions with patients.

6. Increased AI Related Jobs

More jobs than ever before will be using AI. For many occupations, only certain tasks will be automated. So, if you’re trying to prepare for the coming years of career life, you should start working together with AI.

7. More Partially Automated Occupations

According to Mckinsey & Co. research, about 30 percent of activities in 60 percent of all occupations could be automated; in contrast, in 5 percent of occupations nearly all activities are automatable. This means that working hours might be fewer with remaining hours being spent collaborating with AI.

8. Increased Security 

AI drones will increasingly help us deal with emergency situations by delivering care packages or emergency responses to those in need.

9. Better Healthcare

AI is already better than human doctors at diagnosing certain illnesses; this means that AI should set up humanity for a healthier future.

10. Routine Jobs and Data Maximization Jobs are expected to be Automated

AI is good at routine tasks with stable and predictable environments. It is also good at using data for maximizing results. Jobs involving these tasks are expected to be automated soon.

According to MIT Technology Review, one of Google’s deep learning artificial intelligence (AI) programs called Lymph Node Assistant has already been better at diagnosing breast cancer than human doctors since October 2018. Google trained an algorithm to understand which breast cancer tumors have spread. 90 percent of breast cancer deaths occur because of the spreading of breast cancer tumors which is called metastasis. Now Google AI can detect and diagnosis tumors in real time through its “augmented reality microscope.” This device is a microscope with a camera and running AI software inside. The camera continuously takes photos of body tissue under the microscope. Allegedly the program then provides real time insight onto a screen about whether areas are benign or malignant. The details of this microscope have been published in a paper in Nature recently. This new breakthrough will help doctors treat patients in ways that are faster and more successful.
            A new artificial intelligence (AI) assistant called Tiki was recently invented by a team at UC Berkeley. The invention was based on the goal of creating a virtual assistant that is capable of video streaming things from text or voice questions in order to answers those questions. Some of the questions might be about objects in an image or about the weather. In the future, this new technology might be used for household surveillance. Tiki is very easy and fast to implement. Visual questions and answers (VQA) is a new field of research which uses both computer vision and natural language processing in order to answer questions. This type of service is expected to be increasingly common across digital devices very soon.

            Hearing is one of the five senses which technology has successfully aided. Hearing aids are common and successful technologies. Eyeglasses designed to help people to see well have existed since the Middle Ages. Recently, an artificial tongue has been developed by the University of Glasgow. So far, our sense of smell has been harder to enhance. However, a tech start-up called Cyrano Sciences is commercializing an electronic nose designed not just to assist human noses but also to assist sniffer dogs who are used for security purposes.

A New Zealand wine dealer has partnered with an artificial intelligence company called Spacetime in order to match wine drinkers with wines they would probably love. American companies IBM and McCormick & Company are trying to use artificial intelligence to learn about people’s taste preferences. There are many philosophical questions about how these new technologies might best be used. One concern is that enhanced tasting experiences might addict people to the foods they already like and prevent them from trying new foods that they don’t currently like; oftentimes people learn to like foods that they used to not like through exposure.

            An article published in the Economist in 2017 said that data is the new oil. Recent history has seen massive amounts of data becoming available on the internet, especially through social media. The data created in the social media universe can be accessed by data miners and artificial intelligence (AI) creators. A 2019 global report on internet use by Tech Jury claims that two hours and twenty-two minutes a day is the average amount of time spent socializing online. AI requires vast amounts of data in order to be successful. This is because it uses massive amounts of data to make predictions. The more data available, the better AI programs will function. As an internet user, you are creating AI just through your data alone. Facebook recently released 87 million profiles to the data analytics firm that was hired by the leave campaign advocating for the UK to remove itself from the EU. Allowing personal ownership of one’s data might be a way to overcome some of the privacy issues which emerged with the digital age. Recent companies such as Wibson have allowed users to own and sell their data on their apps. Time will tell if the public gets tired of the lack of privacy on sites like Facebook and turns to alternatives.

A group of Chinese researchers at Tsinghua University recently created a chip that combines conventional artificial intelligence (AI) with an AI inspired by the human brain. The team claims this breakthrough could lead to a more naturalized AI. The researchers demonstrated the capabilities of the new chip in a video of a self-driving bicycle that stands-up, balances itself, avoids obstacles, tracks objects and reacts to voice commands. This breakthrough is happening in the context of a US-China technology competition and trade war. Last May, the US put Chinese telecommunications giant Huawei on a trade blacklist to prevent US companies from selling components that might be used to threaten US national security. China is behind in traditional chip making, but AI chip making presents a huge opportunity for China to advance in the technology race.  For China, this puts them one step closer to realizing its goal of being the world leader in AI by 2030.