03/15/2017 | Data and Analytics | Thought Leadership
In recent meetings with business customers around the world, Microsoft emphasized tremendous advances in artificial intelligence (AI) and expounded on the present and future impacts the technology will have on businesses and society.
The state of AI now is very similar to the cloud in 2008 when Microsoft announced Azure. AI isn’t new technology, just as cloud computing wasn’t new in 2008 (indeed, Microsoft was late to the party as Amazon launched AWS in 2006), but interest in AI is getting serious and Microsoft’s vocal investment to its business customers signals an important shift. There were many naysayers of the cloud a decade ago, but as enterprise solutions gained traction, it established a sense of legitimacy and urgency to leverage the cloud for competitive differentiation. Less than a decade later, the cloud is ubiquitous, and although concerns about privacy and security still exist, such issues are now viewed more as nuisances to be solved than deterrents to cloud technology as a whole. The cloud was the future, just as predicted. We’re seeing something similar now.
Only AI will be bigger.
At a high level, artificial intelligence means computers or machines that can “think” or process complex sets of data to carry out tasks in intelligent ways. AI means that the machine can carry out a specific task with intelligence; an example would be Siri, Apple’s intelligent assistant program, or Google’s famous self-driving vehicles.
Advanced AI is still in the sphere of science fiction—think robots that can feel and solve problems—but AI technology is evolving. The most talked about example is probably Watson, the cognitive machine built by IBM to compete on Jeopardy and is now partnering with General Motors on a line of smart vehicles.
In large part, the advancement of general AI is due to Machine Learning, a subset of AI that gives computers the ability to learn through algorithms that can make predictions by analyzing an influx of data. A simple example of machine learning is the autocomplete function in a search engine, where the tool is able to guess what the user wants to type based on having “learned” from the search strings of other users. Other techniques relevant to AI technology include neural networks, natural language processing, sentiment analysis, and speech recognition.
Currently, only large and niche companies are tapping into the true value of AI. These are early adopters leveraging AI to build augmented reality, virtual reality, and IoT-based applications that are “smart” enough to analyze what users need and act on that data. But this movement will eventually sweep small and medium sized businesses up in a blur of data science and infrastructure services designed to leverage new AI technologies. Companies of all sizes will soon be able to tap into the entire power of AI without having to make huge investments, enabling a stage that might be called “AI for everyone.” The impetus now is on businesses to prepare for AI, because it’s truly about to explode.
The Microsoft Angle
Microsoft has recently emphasized its unique position in the cloud and AI services markets, drawing attention to a combination of massive research and development (R&D) spend, numerous state-of-the-art datacenters, and being the developers of much of the software behind these efforts. The Seattle-based giant spent more than $12 billion on R&D in 2016, often working in open communities on joint projects and programs. In late September, Microsoft announced the Microsoft AI and Research Group, which includes more than 5,000 data scientists and computer engineers dedicated to AI advancements, such as with Bing search engine and the Cortana digital assistant, Microsoft’s venture into robotics.
As is often the case with emerging technologies, there is a lot of guesswork regarding what users will actually want or need from AI. Whatever that may be, Microsoft is prepping for the expectation. Here are a few examples of advancements Microsoft has discussed:
- Predictive Data Support: Microsoft has shown how companies can use Azure-based data and analytics services to perform 1.25 million predictions per second, and not only believes that companies should be doing this, but will need to be doing this soon to maintain business advantage.
- Speech and Translation: In October, Microsoft researchers announced that they’ve reached human parity in conversational speech recognition. One of its systems can transcribe the content of a phone call with “the same or fewer errors” than real human professionals trained in transcription — even when the human transcript is double-checked by a second human for accuracy. In addition, Microsoft announced in December that it has expanded the Skype Translator feature to allow users to translate their speech in real time. Microsoft is already putting these technologies into its products for businesses to leverage now.
- AI to Negate Disabilities: A Microsoft engineer, blind since age seven, teamed up with fellow engineers to develop a real-life application based on the Microsoft Intelligence API that allows him to “see” what’s around him. As shown in a video, through a set of connected Pivothead sunglasses, the application reads a restaurant menu for him and even tells him how many people are around him and their apparent emotions. Applications like this, which were only science fiction in the past, are capable of changing people’s lives in the near future.
Advancements in AI will continue to climb and drive mind-boggling change, altering the way humans behave and what they expect from technology and creating endless possibilities for innovation in the marketplace. As AI services become more ubiquitous, every industry will be confronted with the opportunity to provide innovation in services to current and future customers.
What Should You Do?
AI has already disrupted some industries. The momentum has begun, is gaining steam, and is here to stay.
As a business, start thinking seriously about things you thought weren’t possible a few years ago or weren’t going to be possible for a “long time.” Chances are, they are possible now or will be much sooner than you thought. Your competitors are probably thinking about these things as well, so it’s important to be bold. If you haven’t started to think about how your business can leverage AI, you really need to, and soon.
However, before you can even invest in AI, you have to think about your data and analytics maturity. AI is the next step beyond machine learning, but machine learning is dependent on having more basic analytics capabilities and supporting infrastructure in place. Can you use predictive analytics yet? Prescriptive? Do you have the technology infrastructure in place to support growth in your efforts with data? Is your data properly governed and reliable? If not, this is where you need to start.
- Leverage the cloud for computational abilities if you have not already.
- Get your data and advanced analytics capabilities in line and up to snuff.
- Utilize the APIs available in Azure or AWS-based cloud services (Microsoft Azure Cognitive Services APIs is one of many examples) to build your own applications.
- Leverage some AI-influenced technology already available to you (such as the Skype real-time language translator). Look for ways to use this technology or innovate within your own solution and to offer new services to customers.
- Get started with a small machine learning or AI-related project. Get your feet wet, learn fast, and then use the experience and your own creativity to build upon it.
Feeling lost? Partner with savvy technology consultants or achieving a foundation in machine learning and AI. Contact us today.
