No other technology was more important over the past decade than artificial intelligence. Stanford’s Andrew Ng called it the new electricity, and both Microsoft and Google changed their business strategies to become “AI-first” companies. In the next decade, all technology will be considered “AI technology.” And we can thank deep learning for that.

Deep learning is a friendly facet of machine learning that lets AI sort through data and information in a manner that emulates the human brain’s neural network. Rather than simply running algorithms to completion, deep learning lets us tweak the parameters of a learning system until it outputs the results we desire.

The 2019 Turing Award, given for excellence in artificial intelligence research, was awarded to three of deep learning‘s most influential architects, Facebook’s Yann LeCun, Google’s Geoffrey Hinton, and University of Montreal’s Yoshua Bengio. This trio, along with many others over the past decade, developed the algorithms, systems, and techniques responsible for the onslaught of AI-powered products and services that are probably dominating your holiday shopping lists.

Deep learning powers your phone’s face unlock feature and it’s the reason Alexa and Siri understand your voice. It’s what makes Microsoft Translator and Google Maps work. If it weren’t for deep learning, Spotify and Netflix would have no clue what you want to hear or watch next.

How does it work? It’s actually simpler than you might think. The machine uses algorithms to shake out answers like a series of sifters. You put a bunch of data in one side, it falls through sifters (abstraction layers) that pull specific information from it, and the machine outputs what’s basically a curated insight. A lot of this happens in what’s called the “black box,” a place where the algorithm crunches numbers in a way that we can’t explain with simple math. But since the results can be tuned to our liking, it usually doesn’t matter whether we can “show our work” or not when it comes to deep learning.

Deep learning, like all artificial intelligence technology, isn’t new. The term was brought to prominence in the 1980s by computer scientists. And by 1986 a team of researchers including Geoffrey Hinton managed to come up with a back propagation-based training method that tickled at the beginnings of an unsupervised artificial neural network. Scant a few years later a young Yann LeCun would train an AI to recognize handwritten letters using similar techniques.

But, as those of us over 30 can attest, Siri and Alexa weren’t around in the late 1980s and we didn’t have Google Photos there to touch up our 35mm Kodak prints. Deep learning, in the useful sense we know it now, was still a long ways off. Eventually though, the next generation of AI superstars came along and put their mark on the field.

In 2009, the beginning of the modern deep learning era, Stanford’s Fei-Fei Li created ImageNet. This massive training dataset made it easier than ever for researchers to develop computer vision algorithms and directly lead to similar paradigms for natural language processing and other bedrock AI technologies that we take for granted now. This lead to an age of friendly competition that saw teams around the globe competing to see which could train the most accurate AI.

The fire was lit. By 2010 there were thousands of AI startups focused on deep learning and every big tech company from Amazon to Intel was completely dug in on the future. AI had finally arrived. Young academics with notable ideas were propelled from campus libraries to seven and eight figure jobs at Google and Apple. Deep learning was well on its way to becoming a backbone technology for all sorts of big data problems.

And then 2014 came and Apple’s Ian Goodfellow (then at Google) invented the generative adverserial network (GAN). This is a type of deep learning artificial neural network that plays cat-and-mouse with itself in order create an output that appears to be a continuation of its input.

When you hear about an AI painting a picture, the machine in question is probably running a GAN that takes thousands or millions of images of real paintings and then tries to imitate them all at once. A developer tunes the GAN to be more like one style or another – so that it doesn’t spit out blurry gibberish – and then the AI tries to fool itself. It’ll make a painting and then compare the painting to all the “real” paintings in its dataset, if it can’t tell the difference then the painting passes. But if the AI “discriminator” can tell its own fake, it scraps that one and starts over. It’s a bit more complex than that, but the technology is useful in myriad circumstances.

Rather than just spitting out paintings, Goodfellow’s GANs are also directly behind DeepFakes and just about any other AI tech that seeks to blur the line between human-generated and AI-made.

In the five years since the GAN was invented, we’ve seen the field of AI rise from parlor tricks to producing machines capable of full-fledged superhuman feats. Thanks to deep learning, Boston Dynamics has developed robots capable of traversing rugged terrain autonomously, to include an impressive amount of gymnastics. And Skydio developed the world’s first consumer drone capable of truly autonomous navigation. We’re in the “safety testing” phase of truly useful robots, and driverless cars feel like they’re just around the corner.

Furthermore, deep learning is at the heart of current efforts to produce general artificial intelligence (GAI) – otherwise known as human-level AI. As most of us dream of living in a world where robot butlers, maids, and chefs attend to our every need, AI researchers and developers across the globe are adapting deep learning techniques to develop machines that can think. While it’s clear we’ll need more than just deep learning to achieve GAI, we wouldn’t be on the cusp of the golden age of AI if it weren’t for deep learning and the dedicated superheroes of machine learning responsible for its explosion over the past decade.

AI defined the 2010s and deep learning was at the core of its influence. Sure, big data companies have used algorithms and AI for decades to rule the world, but the hearts and minds of the consumer class – the rest of us – was captivated more by the disembodied voices that are our Google Assistant, Siri, and Alexa virtual assistants than any other AI technology. Deep learning may be a bit of a dinosaur, on its own, at this point. But we’d be lost without it.

The next ten years will likely see the rise of a new class of algorithm, one that’s better suited for use at the edge and, perhaps, one that harnesses the power of quantum computing. But you can be sure we’ll still be using deep learning in 2029 and for the foreseeable future.

The post 2010 – 2019: The rise of deep learning appeared first on Artificial Intelligence.

Machine learning has grown to have a significant impact on our daily lives: From Amazon’s home assistant Alexa collecting and analyzing information to anticipate our needs, or Facebook suggesting who we should friend, to applications protecting us from credit card fraud and improving online shopping experiences.

Organizations want their data to do the heavy lifting for them, driven by the desire to save on costs, improve consistency and streamline operations. While ML technologies were previously perceived as an excessive expenditure, today they are seen as an investment in the business’ future and a competitive revenue driver.

In order to stay competitive and successful, organizations have to invest in the right technologies and intelligently use the skills and data systems that they already have. The following three tips will help enterprises evaluate ML benefits and investments and make the most of the technology they already have. 

Get quality data and get it organized

For ML algorithms to offer informed judgments and recommendations on business decisions, the underlying database must provide a steady supply of clean, accurate, and detailed data. It’s important to rememeber that more data doesn’t necessarily mean better data. Quality always comes first. When the quality of data is low, insights derived from the data will be less valuable, as will be the decisions organizations make based on the data.

According to a 451 Research report, 22% of the companies surveyed have already implemented ML algorithms in their data management platforms, while 42% are planning to implement one in the next 12 months. This shift in investment, focused on ensuring captured data is of the highest quality possible rather than simply casting the data net as wide as possible, is a stark industry change. Less than a decade ago, dedicated data quality services and tools were a niche service and largely underused by data-heavy businesses. Now, they are front and foremost in the C-suite’s future plans.

As ML continues to progress, organizations need to ensure that they provide support for their data scientists and invest in the necessary technology to process ML algorithms. If data scientists do not have the correct resources, this momentum will falter. Organizations need to have a high-quality database as the first step in preparation for incorporating ML into their business processes.

Embrace Python

For many organizations, predictive analytics is a key motivator for investing in ML. Predictive analytics use ML to mine large datasets and predict the outcome of future events. This predictive analytics function depends on the data scientists’ mastery of the appropriate programming language. And just how does one master anything? By studying, experimenting and learning from others.

Here is where Python, one of the most popular programming languages in the world according to Tiobe Index, really stands out. Python has become popular mostly because of its simplicity, readability, versatility and flexibility. As millions of people around the world learn and use the language, more and more individuals and groups share programs, tips and entire algorithms with each other. Python’s network of users gives organizations hoping to use and experiment with Python countless learning materials right at their fingertips.

Ultimately, having one underlying data infrastructure that everyone across all teams can feed into and take from is the key. For the business intelligence team, this will typically be Structured Query Language (SQL). However, in order to succeed, data scientists must be able to run scripts on the data using their preferred language — notably Python. This standardization and democratization of data means that organizations can apply ML across any and all parts of the business in more creative and experimental ways.

The benefits of hyperscale cloud

Despite on-premise IT infrastructure’s ability to host many open-source frameworks to create ML solutions, many organizations still lack the power and scalability to support them. If an organization is evaluating ML for a project, hyperscale cloud might be a good option to consider, since it offers consumption-based access to graphics processing unit (GPU) compute, which can dramatically accelerate the process of training a deep learning algorithm.

Once the requirement moves from batch analysis to real time, the flow of relevant data must keep pace with ML algorithms working in near real-time. Ensuring that workloads are supported throughout a project’s lifecycle and organizations have the ability to experiment with ML capabilities is essential, and cloud elasticity can be used to address that.

It has never been easier for organizations to expand into the cloud, as the big three public cloud providers — AWS, Google and Amazon — all fight for ML business. Despite this, organizations still lag behind in exploiting the elastic scalability of the cloud to derive value from their organization’s data with ML.

While ML may seem overwhelming and complicated, creating an infrastructure for ML projects is more achievable than many organizations think. In fact, most organizations are already using the technologies they need, such as databases, programming languages, and Infrastructure as a Service, to lay the foundation for ML optimization.

The post Three Tips for Laying the Groundwork for Machine Learning appeared first on Artificial Intelligence.

BGSU is gradually implementing artificial intelligence, or machine learning, on campus in ways that should theoretically be able to help with efficiency and compensate for human errors while supplementing our strengths.

Machine learning is the capability of computers to learn new information without someone giving specific instructions to them. They learn by analyzing patterns and data, rather than someone programming them. Artificial intelligence can potentially simulate human-like patterns of thinking, but emotions are thought to be exclusively-biological. John Ellinger, BGSU’s chief information officer, says that machine learning is the technology that is by far the most likely to change higher education in the next ten years.

“It may be 2028 or 2030 before we completely implement it, but the gains that are being made in this area are striking,” he said. “I just read an article that higher education across this country would be replacing up to as much as 10 per cent of our tasks with artificial intelligence.

Computer science professor Venu Dasigi knows one of BGSU’s current uses of fairly primitive AI is identifying students that may need help to avoid failing classes or dropping out.

“What I know of is that the AI we use, where our systems learn from big data, we try to use an advising program where we work with students and so on. The university makes use of whatever it has seen from student patterns in the past,” Dasigi said.

Amazon and some of its products utilize a form of artificial intelligence designed to be consumer-oriented.

“Alexa is nothing more than something attached to the internet, which has programming behind it to let you ask questions, and it goes and finds an answer. Another example is Amazon’s shopping experience. It lists the things that you might be interested in, so they are using predictive analytics, and lots of companies do that now,” Dasigi said.

Ellinger and Dasigi agree the most physically-tangible application on campus for machine learning currently, is the robotics lab in the basement of Hayes Hall.

“It’s a lot more about using automatic manufacturing. The robot motion needs to be optimized because the robot needs to figure out optimal trajectories for picking stuff out and things like that,” Dasigi said.

“We are at a point with AI where we probably will not be able to fully replace humans, at least for the next few years, in most cases. There are a few scattered cases, maybe, but most help desk cases or online check cases, most implementations, they should probably be supplementing a human and not replacing them,” Porco said.

The way that AI is rolled out at BGSU and other colleges across the nation spells out some clear advantages for the effectiveness of web-based operations. With that said, it isn’t a bad thing to stay vigilantly informed of how far it can intrude into people’s everyday lives.

The post How colleges are utilizing artificial intelligence appeared first on Artificial Intelligence.