artificial intelligence • News March 2021

Google Magenta Project Releases First Song

June 3, 2016 By Lonnie Davidson Leave a Comment

Google Brain, one of the tech giant’s many units, unveiled last month the Magenta Project, which would be using Google’s TensorFlow, the open source artificial intelligence platform.

With TensorFlow’s help, the Magenta Project was supposed to be able to generate art and music through algorithms. And now, the trained neural network has released its first masterpiece – a 90-second piano melody.

As the first tangible product of the project, the song works only with four notes provided at the start, but it’s still an impressive breakthrough.

So can neural networks write songs? Yes. But are the songs any good? As with any other form of art, the answer to the second question depends on the listener. If you love hearing your toddler relentlessly smashing on the keys of an electronic keyboard, this is for you.

But for everyone else, the melody generated by Google’s Magenta is not exactly a delight for the ears. In spite of being no good, the breakthrough still matters, mainly because it proves that machines can be taught how to create art.

According to a blog post by Douglas Eck, a Google research scientist, Magenta has two ultimate goals. For one, the project should help advance the development of algorithms that will allow machines to learn how to create music and art in general.

On the other hand, the second goal is broader and in the long term; it deals with Magenta’s ability to create a community comprised of coders, machine learning researchers, and artists who will come together to develop machine-generated content.

As with any creative project, Google will be releasing Magenta’s work to collect feedback; so if you’re a researcher that fits the profile or an artist, you are invited to join the project community. One of the practical things you will do is check out its code and feed more data to Magenta’s database.

Google has a history of launching projects that look to teach machines how to generate art. DeepDream was before Magenta and worked as a free visualization tool enabling users to create their own images inspired by neural networks.
Image Source: Next Geekers

Amazon’s Alexa Comes To Desktop With Echo Simulator

May 28, 2016 By David Kellen Leave a Comment

One of the most recent unveilings of Amazon is the Echosim.io, a site that mimics the functionality of an Amazon Echo speaker. In others words, the Alexa voice assistant technology is now available on desktops.

Amazon’s Alexa has already charmed thousands of users in devices like the Amazon Echo or its smaller siblings, Amazon Tap and the Echo Dot. But for those who don’t have any of Amazon’s gadgets were missing out – but not for much longer.

The idea behind Echosim was first developed by Sam Machin during a 2015 hackathon; later, it was dubbed Alexa in the Browser. How exactly does it work, though?

Tapping into Alexa’s potential on desktop systems is really easy. You just have to access Echosim.io and you don’t even have to create a new account. Sign in with your Amazon username and password, hover your mouse over the microphone button, and simply talk to the voice assistant.

Evidently, throwing questions at the AI (artificial intelligence) requires a working microphone, but that shouldn’t be a problem in this day and time. But the integration will benefit more than just the customer sector.

“Developers worldwide can use Echosim to experience Alexa,” said Glenn Cameron, the developer marketing manager for Amazon Alexa. And, truth be told, the app has landed at the perfect time.

Why? Because it’s a worthy competitor for Apple’s Siri software development kit (SDK), which the company has already announced. The SDK is a tool that will help coders improve their apps with the help of Siri’s features.

Furthermore, inside sources from the company reveal that Apple will soon enable Siri on the Apple TV set-top box, which would turn out to be a strong counterproposal to Amazon’s Echo.

Amazon is still in the lead, however, because Siri will be landing on OS X only some time later this year. More than the two major tech players, other companies have started tapping into speech recognition abilities.

Google, for instance, has also made this feature available on desktop systems. Some versions of Google Docs already support the Google Assistant tech, and the search giant is looking forward to giving third-party developers the keys to its treasure chest.
Image Source: Amazon

Robots Step Aside as Humans Take the Lead in Mercedes’ Next S-Class

February 26, 2016 By Joseph Decker Leave a Comment

Robots have fascinated humans for years now. We’ve made futuristic movies about them, we’ve advanced our technologies so that one day we might be able to get these large stacks of metal to do everything for us: shop, clean, work and even drive us around.

But there is one company who doesn’t think artificial intelligence is quite that suitable for all their operations and that is Mercedes.

According to Mercedes’ head of production Markus Schaefer, robots cannot handle all the different degrees of individualization, as well as the multiple variants that the company has these days.

By turning to humans for some parts of their activity, Mercedes will eventually save money and protect the world’s future by giving people a place to work.

As it turns out, certain features presented on the new S-Class, such as cooled and heated cupholders or the different kinds of tire valve stem caps, cannot be developed by automotive assembly robots, which is why Mercedes has already started replacing them with humans, who are better suited for this kind of work.

This means that when it comes to vehicle customization processes, robots cannot provide the flexibility and detail-oriented skills that humans possess. While these intelligent machines are extremely apt when it comes to performing repetitive and well defined tasks, they cannot adapt to other variables.

Nevertheless, this does not mean that the German automaker will kick all their robots to the curb. There are still a great number of machines at their assembly plants that contribute with a number of tasks, such as painting, welding and even engine installation.

The company’s biggest plant, located in Sindelfingen, rolls out some 400,000 vehicles per year, processing some 1,500 tons of steel per day, so obviously automation plays a big role in their operations.

Since robots are highly reliable and efficient, that makes them a valuable resource when it comes to streamlined production. However, the lack of versatility of these machines is a problem for the German company, who is now looking to cut the hours need for producing a car from 61 to 30.

Even though robots will not fully vanish from the Mercedes manufacturing plant, they will eventually have a reduced size and function alongside human workers in order to improve their responsiveness.

According to the International Federation of Robotics, there are about 1.3 million industrial robots which will start operating within the next two years.

Image Source: Mashable

Microsoft Decides to Apply AI Technology to Offline Translations

February 19, 2016 By Janice Bower Leave a Comment

In an interesting move which is bound to make their application a powerful competitor for Google Translate, Microsoft decides to apply AI technology to offline translations.

The company has recently introduced some updates for the Android version of their Translator app, which will include access to offline translations, powered by the same artificial intelligence technology that Microsoft used in Skype Translator and Bing Translator.

It’s called a Deep Neural Network and it guarantees that the offline translations will be extremely close to their online versions in terms of accuracy.

This will reportedly make Android users very happy, seeing how many of them are constantly on the move and they don’t want to use up too much of their mobile data.

This means that if you are traveling abroad and you want to find out what a particular phrase means, you can simply download the offline translation pack and use it for translating both text and voice.

Uers can find offline language packs available in Chinese Simplified, Italian, German, French, Russian, Polish, Portuguese, Spanish and Vietnamese.

Microsoft has also incorporated a special option for the iOS version and that is the image translation feature. In technical terms this is known as the Optical Character Recognition (OCR). Confused about how it works? It’s pretty simple actually.

All you have to do is put your camera on whatever you want to translate and then the application will automatically generate a translation, replacing the previous text from the camera.

Even more interesting is the fact that you can use this tool in order to translate a text from an image obtained from a social media or messaging platform. For this, you can use the same procedure as before and it will automatically show you the translation.

Nonetheless, you should know that the image translation feature does not work for all languages. It currently supports only 21 of them, so make sure to check out if the language you are interested in is being supported by this application.

However exciting this news may be, we have to keep in mind the fact that all of these features have existed for several years now, both in Windows and Windows Phone platforms. This makes Android and iOS the highly anticipated latecomers to this party.

Regardless of all the competition, Google Translate still remains the most important application in this field, supporting over 103 languages over a whole range of operating systems.

Image Source: AndroidPolice

Drones May Help Search for Lost Hikers

February 11, 2016 By Lonnie Davidson Leave a Comment

Every year authorities in Switzerland have to address over 1,000 calls regarding injured or missing hikers. This number could go way down now that drones may help search for lost hikers.

A team of researchers from Switzerland have started to teach drones how to navigate forest trails and from the experiments that they’ve conducted so far it seems that these unmanned aerial vehicles can predict the correct path 85% of the time.

This could be the answer that most people have been looking for, seeing how man-led search and rescue operations are not only costly, but they also take a lot of time. It is important to keep in mind that time is the primary factor taken into consideration when trying to locate a lost hiker, due to the unforeseen dangers that they may be exposed to or the serious injuries that they could be suffering from.

From this perspective, it is a safe bet to say that a drone is less expensive and more efficient, because the automated process it uses allows it to conduct a search faster and more in-depth.

In order to teach drones how to overcome a variety of obstacles that they may encounter, computer scientists from the Dale Molle Institute for Artificial Intelligence and the University of Zurich have created a special navigation software. The algorithms of this software will allow the drone to recognize a vast array of forest landscapes and to specifically follow man-made trails.

As Alessandro Giusti, researcher at the Dale Molle Institute, points out, this is an incredibly difficult task, because often times not even humans can follow the trails designed for them, so to ask a computer to navigate such elaborate surroundings is definitely a challenge.

The team of researchers managed to train the drone by feeding it 20,000 images of hiking trails from the Alps, taken with the help of a helmet cam. Having all this information available and the right software to process it, the drone was capable to predict the correct trail 85% of the time, whereas humans had an 82% success rate.

Juergen Schmidhuber, director of science at the Dale Molle Institute for Artificial Intelligence says the method that was used for these drones has been under development in their lab since the early 1990s and it’s been used not only for drones, but also for various applications, such as speech recognition products and lightweight robots.

Image Source: Gizmag

Marvin Minsky Died at the Age of 88

January 26, 2016 By Grant Hamersma Leave a Comment

Marvin Minsky, one of the brightest minds in science has sadly passed away on Sunday.

Marvin Minsky died at the age of 88. The famous cognitive scientist died on Sunday and the news of his death came from the MIT Media Lab founder, Nicholas Negroponte. Negroponte has praised Minsky and said that he was one of the best in the science domain. He also said that Minsky brought deep thinking, humor and a different perspective to the Media Lab.

In 1951, Minsky developed the first neural network simulator, called SNARC. In 1956, he participated in a Dartmouth symposium that lasted two months, which is considered to be the event that founded the concept of artificial intelligence. After the symposium, Minsky wrote books that helped the scientist better understand the challenging that occur when machine intelligence is trying to be developed. These books include The Emotion Machine, Perceptrons and The Society of Mind.

Some of his most famous inventions include robotic devices, mechanical arms and hands, the Confocal Scanning Microscope and the musical variations synthetizer called the Muse. He was a member of NAE and Argentine NAS. He received a lot of awards for his extraordinary work, which include the Rank Prize, the ACM Turing Award, the Benjamin Franklin Medal and the IJCAI Research Excellence Award.

Marvin married his wife Gloria, a psychologist, in 1952 and their home became a repository full of valuable icons and artifacts. According to some experts, the house should be considered a national historic site. Some of the most interesting objects found in their home are a vintage robot, a bongo that Richard Feynman played when he visited, a vintage jukebox and a trapeze.

During an interview given to the Post last year, Gloria said that during one of her first conversations with Marvin, he said that he wanted to study the way in which the brain works. When he revealed that to her, she thought that he was either very wise, either very dumb but the years demonstrated that he was indeed extremely smart. The couple had very famous friends like Stanley Kubrick, Robert Heinlein, Arthur C. Clarke and Isaac Asimov. Minsky also had conversations with John von Neumann and Albert Einstein.

Marvin Minsky died at the age of 88 due to a brain hemorrhage and the world lost one of the brightest scientists there ever was. All of his life he dedicated to research in the artificial intelligence field.

Image Source: www.upload.wikimedia.org

The Killer Robots Could Endanger the Human Species

January 25, 2016 By Grant Hamersma Leave a Comment

Robots developed to kill could put the entire human race at risk.

Even though it might seem like something that came out of a futuristic movie, the killer robots could endanger the human species. An elite gathering was held in the Swiss Alps, at a ski resort in Davos in order to warn influential people about the risks of further developing robots with artificial intelligence for wars. The meeting lasted for four days, from January 19 until January 23.

One of the speakers at the Davos meeting, Angela Kane, who is the German representative of the Disarmament Affairs said that the world didn’t take good measures when it developed robots with artificial intelligence. She warned that it might be too late, that the damage has already been done. She also stated that many countries that are not very developed don’t even know about the risks that they could face in the near future.

A lot of scientists who were present at the meeting agreed that autonomous weapons could mark the beginning of a new, more dangerous warfare era. Another speaker at the meeting was Stuart Russell, who is a computer science professor at the University of California. He said that the biggest concern of the human race right now should be the artificial intelligence weapons. These weapons could locate and kill the targets, without any intervention from humans. Russell also said that the wars fought by robots for humans could never end well.

Russell said that he was against developing robots for ethical reasons, but he believes that these types of robots create issues that surpass the ethical border. He hopes that the development of these weapons will be prohibited and stopped, as it happened with the biological weapons.

Alan Winfield, who is an electronic engineering professor at the University of the West of England said that replacing humans on the battlefield with robots could have serious consequences. He said that not only will the moral responsibility be deprived from the humans, the reactions of robots will be hard to predict. He said that the robots could act chaotically, when put in chaotic situations.

In July 2015, more than 1,000 scientist including Stephen Hawking addressed the issue of artificial intelligence weapons and called for them to be banned. They stated that they could easily get out of the human control and they could also fall into the hands of extremists that are violent. Either way, the killer robots could endanger the human species.

Image Source: www.upload.wikimedia.org

How Will the World End, According to Stephen Hawking

January 22, 2016 By Grant Hamersma Leave a Comment

Stephen Hawking addressed his fears on how will the world end.

At the beginning of January news on how will the world end, according to Stephen Hawking was shared. Some of the assumptions of the professor at the Cambridge University include global warming, genetically created viruses and a nuclear war.

The theoretical physicist also fears that the artificial intelligence could endanger the entire humanity. He claims that these apocalyptic scenarios can be avoided only if the planet is left behind. He said that for now the chances of such major disasters are low for now, as the years will pass they will only increase. He approximated that in the next thousand years, the human annihilation by means of one of his apocalyptic scenarios is very likely to happen.

Hawking believes that in the next hundred years, self-sustaining colonies will exist in the space and he warns that until then people must be careful at the dangers that surround them. One of his biggest fears is the development of the artificial intelligence. Hawking believes that technology could both save and destroy the human kind. In an interview given to BBC in 2014, he talked about the threat that the full artificial intelligence could pose to the humans. He advises the researchers to stop the development of the artificial intelligence here, where it is still safe and useful. Hawking believes that further development of artificial intelligence could surpass the humans, who have a biological evolution that is slow. Some experts have proposed slowing down the innovation pace, but Hawking believes that is an unrealistic goal. He said that the best thing the experts could do in this situation is to assess the dangers and make sure that they can be controlled or even stopped.

Hawking also talked about other dangers, of which the experts should be more aware of. He said that many clinical trials that involve the genetic modifying of viruses, crops and other organisms are unsafe. Hawking claimed that some of them are done only for profit and some of them don’t follow the guidelines that ensure they pose no threat to humans.

Stephen Hawking also mentions the threat the global warming could have on the humanity. He believes that the day will come when the temperatures will be so high on Earth that human life will no longer be possible. He mentions that the carbon dioxide levels that have been rising from the burning of the fossil fuels is one of the main reasons of the global warming.

How will the world end, according to Stephen Hawking includes a lot of doomsday scenarios, which we hope will never occur.

Image Source: www.upload.wikimedia.org

MIT Algorithm Understands How Memorable Images Are

December 17, 2015 By Grant Hamersma Leave a Comment

map of memorability score created by MemNet algorithm

MIT algorithm understands how memorable images are, feeding into artificial intelligence deep learning and shining new light on memorability.

The research team with MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) created the algorithm. But the team’s plans don’t stop here.The algorithm, available for try-out online could soon be turned into an app. What would we do with such an app? Well, for starters check out how our memorable our selfies and profile pictures are.

It might seem trivial, but the memorability checking algorithm could become a crucial factor in advertising, marketing and other largely imagery dominated sectors. The algorithm is called MemNet if you’re planning to check the memorability score of your own images online.

Aditya Khosla, graduate student with MIT’s CSAIL and lead author of the paper explaining the mechanism behind the MemNet algorithm declared that:

“Understanding memorability can help us make systems to capture the most important information, or, conversely, to store information that humans will most likely forget”.

Simply put, MemNet algorithm functions kind of like an instant focus group that will allow you to gain insight on how valuable a visual message is. MemNet algorithm is based on previous research conducted by the CSAIL team. The most exciting thing about the newly developed algorithm is that is can perform almost at the same level as humans.

Based on deep learning, an area of artificial intelligence, the MemNet algorithm makes use of neural connections to find relevant data and create unique patterns of recognition for each case. Facebook’s photo-tagging feature or Apple’s personal assistant Siri are also created based on deep learning. The neural networks underpinning MemNet algorithm’s functioning are organized in layers and layers of processing units. Each of the units is performing computations on the data fed into the algorithm in succession. The more data it receives, the better and more accurate the final deliberations are.

Thousands of images were analyzed with MemNet algorithm to test its capabilities. When the results were compared to how humans perceived the images in terms of memorability, MemNet algorithm was just a few percentage points behind human performance. The MIT algorithm understands how memorable images are or aren’t for that matter. One thing is certain. MemNet algorithm fared 30 percent above any other algorithm working on the same principle.

Photo Credits: i.kinja-img.com

Elon Musk Wants Us on Mars Before a Third World War or Robot Uprising

December 14, 2015 By Barbara Mast Leave a Comment

SpaceX’s CEO believes humanity will self-destroy

Not being very optimistic about our future on this planet, Elon Musk wants us on Mars before a Third World War or Robot Uprising. The CEO of SpaceX says that human’s colonization of Mars would be like making a backup of life just like we backup the important information on our hard drive, which we wouldn’t like to lose in case of an unexpected event.

Musk is positive that we have more chances to continue life on Mars than we have here, on Earth. SpaceX has no concrete plans to announce in this moment but they will as soon as they’re ready. However, Musk wants to send humans on the Red Planet as soon as possible, fearing that a possible nuclear Third World War would be worse than anything before, not excluding the chance of exterminating life on Earth.

But even if life wouldn’t perish in the eventuality of a nuclear war, it could surely send the technology back tens or even hundreds of years.

Another possible end of life on Earth predicted by Elon Musk would be the development of Artificial Intelligence which could take over the world.

According to Musk, we have to move fast and increase our chances of survival by moving to Mars. However, he wouldn’t have any problem to detonate nuclear bombs at the poles of the Red Planet in an attempt of making it hotter and more habitable.

Musk disagrees with NASA’s claims that nothing that could be contaminated with probes from Earth should be allowed to get near the areas rich in water on the Red Planet. He believes that since there is no evidence of life on Mars and in the best case that could only be found as subterranean microbial life, it would be almost immoral for humans not to colonize the Red planet if that would be our only chance of survival as a species.

Despite its concerns regarding the development of Artificial Intelligence which could rise against humans, Elon Musk has just donated $1 billion to OpenAI, a non-profit research company in the field of artificial intelligence. The concept of OpenAI is to make research without being concerned by generating financial return.

However, Musk claims that there should be some international regulatory organism watching over the development of artificial intelligence, to make sure that nobody makes a huge mistake.

Image source: pixabay

Categories