Microsoft is one of the world’s most ubiquitous technology company. Best known for their Microsoft Windows operating systems the company also produces office suites, games, Azure, Xbox and many more. However, Microsoft is also at the forefront of developing artificial intelligence technology. As well as the potentially revolutionary Project Hanover, Microsoft AI applications are transforming businesses, services and how we interact with the world.
However, this has not always been the case.
Microsoft was initially slow to realise the potential of AI.
In the course of this article, we will explore how Microsoft, despite being slow off the mark, has transformed themselves into a leading AI-first company.
We will also highlight some of Microsoft AIs most exciting innovations.
By exploring some of Microsoft AIs leading initiatives we will see how then have been able to evolve into a leading AI company.
The History of Microsoft
Microsoft was formed in Albuquerque, New Mexico on the 4th of April 1975 by Paul Allen and Bill Gates.
The company’s name, Microsoft is a portmanteau of the words microcomputer and software.
Success was swift, as was early growth.
At the end of 1979, the company opened its first international office in Japan.
Following a move to Washington in 1979, the company restructured in 1981 becoming an incorporated business.
Early products were variations of the Microsoft BASIC, the dominant programming language of the period.
This meant that a Microsoft operating system would come as standard with IBM computers.
Alongside this Microsoft continued working on their own operating system.
The first publicly released Microsoft operating system was a variation of Unix.
This was launched in August 1980.
It would eventually be home to the first edition of the companies word processing software, Microsoft Word.
By the 1990s Microsoft Windows had over 90% of the market share of personal computers.
However, the company was not content to rest on its laurels.
In 1991 they established Microsoft Research.
This branch of the company is dedicated to researching and developing applications of computer science.
Much of its work paced the way for the development of many Microsoft AI applications.
Microsoft was Initially Slow to Realise the Potential of AI
Despite the formation of Microsoft Research, the company were initially slow to embrace the possibilities of the internet.
Instead, they focused on continuing to develop Windows and Office applications.
Additionally, the company developed security and anti-piracy software as well as the Xbox gaming console.
This focus, on applications away from AI, remained as late as 2014.
This lack of focus meant that other companies were beginning to usurp Microsoft’s established position as an industry leader.
It was also threatening Microsoft’s market share.
This changed in 2014 with the appointment of CEO Satya Nadalla.
Since then Nadella has focused Microsoft into transforming its AI strategy.
Nadella has stated that his mission is to rebuild Microsoft, getting it back to the pioneering company that it originally was.
Nadella’s change in focus has led to the development of a number of Microsoft AIs innovations.
These, such as the cloud suite Office and Azure, have helped to re-establish the company at the forefront of the industry.
2002 saw Microsoft AI launch .NET alongside a number of updated versions of established products.
This included a new programming language C# and API for Windows.
Microsoft AI is not only focused on evolving the companies established operating systems and developing new applications.
The company has also acquired other AI-powered companies such as Skype Technologies.
Microsoft purchased Skype for $8.5 billion USD in May 2011.
This evolution has seen Microsoft’s net profits climb to $16.6 billion.
If growth continues at this rate the company is expected to break the trillion-dollar mark in the next few years.
This transformation is attributable in large part to the focus of acquiring and developing Microsoft AI applications.
Microsoft has a constant desire to evolve and stay at the forefront of technology.
Consequently, this means that Microsoft AI is at the forefront of the majority of their innovations.
This drive has helped the company to remain a world leader.
It has also led to Microsoft being described as a trillion-dollar company alongside Apple and Amazon.
The Microsoft AI Lab
The Microsoft AI lab is dedicated to promoting and developing AI-based initiatives and applications.
The stated aim of the Microsoft AI lab is to design AI applications that are trustworthy and reflect the ethical principles of the company.
This means that the Microsoft AI lab promotes principles such as making AI understandable and accessible.
It also aims to develop applications that are secure, reliable and safe.
Microsoft believes that by providing a safe and accessible means of accessing AI more people will be able to benefit from its possibilities.
To further their aims Microsoft AI Lab initiatives are developed alongside a number of respected partners.
These include the machine learning department at Carnegie Mellon University.
Microsoft AI Lab focuses its efforts in a number of different areas.
The belief is that while progress is made in each separate field, lessons can be learnt and applied across the board.
This harmonious approach to development helps to drive initiatives forward, meaning more complete developments can be made quickly.
Microsoft AI in Action
The Snow Leopard Trust works to protect the endangered snow leopard.
Living in remote, hard to access areas snow leopard numbers are monitored by camera.
Using motion-sensitive trail cameras the Trust captures over 1 million images each year.
However as the cameras are motion-sensitive, many images are of grass, sheep or other items.
Other images may be blurred or difficult to see.
As little as 5% of the images taken could be of snow leopards.
Manually reviewing each image for signs of the snow leopard can be time-consuming and difficult.
With help from Microsoft AI lab, the Snow Leopard Trust have developed intelligent systems that can read these images.
The algorithms that drive this system can sort through the images taken with a reported 95% accuracy.
As technology improves, and the algorithms continue to be refined, this will only improve. Not only is this application reliable, but it is also quick. A workload of images that would take a team about 10 days to sort takes the system around 10 minutes.
Koustubh Sharma is a senior regional ecologist with the Snow Leopard Trust.
Sharma has said of this partnership that “By automatically analyzing the images and creating a database for us, Microsoft AI is providing our small team with the time to do more surveys and collect better data.”
Microsoft AI Unversity
One of the most difficult challenges for any AI-focused initiatives is bridging the skills gap. To solve this problem the company has established an internal university to help train staff in artificial intelligence.
Chris Bishop is the director of the Microsoft Research lab based in Cambridge, UK.
In an interview, Bishop explained that Microsoft AI university “is an internal education programme so that people who are incredibly smart and capable but trained in a different domain can quickly learn about machine learning both in a foundational sense but also in a practical sense of how to use it.”
As AI continues to evolve and improve it is increasingly integral to tech companies strategies and products. Even the smallest advances can have a major impact on product features.
Microsoft is Also Recruiting the Next Generation of AI Talent
Microsoft’s AI policy is not confined to developing the skills of their own staff.
The company is also attending technology conferences, identifying potential employees and sponsoring students through university. Once the student graduates they are then guaranteed a job within the Microsoft AI operation.
Bishop sees the recruitment of professors and established experts as a short term approach.
“I don’t think it serves even the industry itself very well, let alone academia or the nation, to take that rather short term view.”
That is why Microsoft also invests heavily in training their existing staff body as well as recruiting the next generation.
Microsoft is not the only company focusing on developing the next generation of AI experts.
DeepMind is also working alongside universities, giving AI and machine learning lectures to students.
Project Hanover to Revolutionise our Understanding and Treatment of Cancer
While Microsoft is best known for its operating system, Microsoft AI applications are impacting in a whole range of areas.
This is a consequence of the company’s increased focus on exploiting the potential of AI.
One of the most potentially world-changing applications is Project Hanover.
This is Microsoft AIs attempt to use computer science to solve cancer.
Trained biologist Jasmin Fisher who works in Microsoft AI’s Cambridge lab programming principles and tools.
She explained that “we are trying to change the way research is done on a daily basis in biology”.
To hasten the process different teams work on different areas of AI and potential applications.
Microsoft AI does this by sorting through the available research data and presenting it in a categorized, useable way.
Meanwhile, another Microsoft AIs research team is using computer vision and machine learning to develop insights into how tumours progress.
Many of these applications are already being used or trialled practical.
However, with the long term in mind, some teams are working on more advanced ideas or moonshots.
One of these so-called moonshots aims to develop a way of programming cells to fight diseases.
Microsoft AI Adopts Two Basic Approaches
Jeannette M. Wing is Microsoft’s corporate vice president and is in charge of the company’s basic research labs.
She has explained how Microsoft AI’s approach to solving cancer focuses on two approaches.
The first approach works from the concept that cancer, and other diseases and biological processes are natures information processing systems.
Here Microsoft AI researchers utilise software, tools and applications that conventionally model and reason computational processes such as programming languages.
These processes are trained to model and reason biological processes.
The second approach attempts to apply machine learning and similar tools to analyse biological data.
By analysing information with sophisticated tools Microsoft AI researchers try to get a better understanding of how cancer develops and works.
This can then help to inform treatment plans.
Wing believes that “the collaboration between biologists and computer scientists is actually key to making this work.”
Project Hanover has led to Microsoft investing in larger concepts such as cloud computing.
This allows Microsoft AI researchers access to more computer powers, meaning that they are better able to tackle large problems.
Wing has also explained that it makes sense for Microsoft to invest in developing tools that can operate on any computing platform, even a living cell.
She said, “If the computers of the future are not going to be made just in silicon but might be made in living matter, it behoves us to make sure we understand what it means to program on those computers.”
Changing Approaches to Treatment
Microsoft AI’s Project Hanover comes at an interesting time in genetic research and understanding.
David Heckerman, a scientist and director of Microsoft AIs genomics group explained that “We’re in a revolution with respect to cancer treatment.”
Heckerman explained that not so long ago if a patient was diagnosed with cancer they were treated for cancer.
However now, “we know it’s just as, if not more, important to treat the genomics of cancer, e.g. which genes have gone bad in the genome.”
The mapping of genetic material such as the human genome has helped to increase this understanding.
As has the ability of AI tools, in particular, machine learning, to organise and process large volumes of data.
Microsoft AI and Project Hanover are helping to drive these developments forward.
Microsoft’s Bio Model Analyser
Known as the BMA for short, the Bio Model Analyser is a Microsoft AI developed cloud-based tool.
It aims to accurately recreate the connections that cells make.
This allows biologists to model the way that cells interact with each other.
BMA works by creating a computerised model.
This allows the user to compare the processes of a healthy cell with those of an abnormal cell.
This approach allows scientists and researchers to see how interactions between both genes and proteins, can lead to cancer.
Understanding this relationship will help scientists personalised patient treatment plans that are effective but not invasive.
It can also be used to predict and identify at which point cancer will become resistant to treatment.
Ben Hall is a Royal Society University Research Fellow based in Cambridge, UK.
Hall has worked alongside the Microsoft AI team that developed BMA.
He explained that “I use BMA to understand cancers – understand the process of becoming cancers, understand the communications that are going on.”
The Literome Project can Improve Research
One of Project Hanovers first developed tools was Literome.
This is a computing system based on the cloud.
It is capable of sorting through huge amounts of data, such as research papers.
Microsoft AI developers combined machine learning tools with natural language processing systems.
This enabled them to create Literome, a sophisticated research model.
It is capable of conducting complex searches and identifying the most relevant documents.
For example, Literome can be used to identify genomic research that may be useful to individual diagnosis.
Without Literome this could be a time-consuming process that is prone to human error and inconsistencies.
Giles Maskell is a radiologist and president of the Royal College of Radiologists.
Maskell has observed that at one time a CT scan would produce around 200 images.
Now, the technology behind the CT scanner is more advanced.
This means that it may produce 2,000 images.
“The fine detail far exceeds our ability to understand it all and to actually process it into something that is meaningful,” said Maskell.
Applications such as Literome can be used to sort through this information.
This allows professionals to accurately read and interpret the information, helping to improve diagnosis and treatment.
Making Powerful Computers Accessible to all
As well as machine learning tools, Azure cloud computing platform is also helping researchers.
This platform hosts tools and applications that can be accessed by biologists and medical experts the world over, even if they don’t have a powerful computer.
Microsoft is proud of its reputation as a software innovator.
For this reason, Microsoft AI and other applications are developed with the aim of being as easy to use as possible.
The more understandable, and accessible the tool, the more likely it is to be used by physicians and researchers.
How Azure is Making the Power of Cloud Computing Accessible to All
One of the driving forces in Microsoft’s re-emergence as a major AI player is Azure, their cloud computing service.
Originally conceived under the name Project Red Dog.
Azure aims to provide a platform for building, testing, and using a range of applications via Microsoft’s own data centers.
Microsoft AIs Azure is capable of a range of functions.
Capable of supporting different programming languages, Azure also supports tools, frameworks, systems and both Microsoft and third-party software.
Additionally Azure provides Saas (software as a service), PaaS (platform as a service), and Iaas (infrastructure as a service).
This wide range of capabilities makes it a very useful tool.
Microsoft AI’s Azure allows users to process huge amounts of data.
This data can be fed into apps that convert the information into something useful, such as a forecasting API.
The data Azure processes can be used to predict trends, such as a buildings energy needs.
It can also help a company identify which products it should produce and market.
Joseph Sorosh is corporate vice president of Information Management and Machine Learning at Microsoft.
He explained that Azure allows the user to use data intelligently.
“That’s really where machine learning comes in.
Machine learning is really about looking at historical data patterns and being able to predict ahead.
It allows you to take the past and peer into the future.
So instead of looking in the rearview mirror, you’re looking forward.”
An Easy to Use Tool With Real World Applications
As with other Microsoft AI applications, Azure is easy to use.
Alongside the accessible visual interface, users can access a range of starter templates and drag and drop workflows.
Microsoft AI developers have also made the process of loading data easy.
This all means that you can use Azure with little to no programming knowledge.
Microsoft AI’s Azure platform is already being used in a range of real-life scenarios.
For example, ThyssenKrupp uses Azure to power predictive maintenance software.
This software can highlight a potential problem in the machinery before it affects the working process of the elevator.
By fixing the problem before it becomes a problem, it prevents a catastrophic breakdown.
Predictive maintenance, when used correctly allows machinery to become even more reliable.
Musthawu Ahmed is the chief operating officer of JJ Food Service Limited.
They are one of the UK’s largest food delivery service companies.
The company uses Microsoft AI’s Azure to predict orders before the customer makes them.
This helps to speed up the process, and also helps JJ Food manage stock levels.
Ahmed said “By using Azure Machine Learning, we can now make recommendations to customers ordering a particular item.
This feature is vital to promote new products or bring customers’ attention to products that they currently go elsewhere to purchase.”
Azure is Also Improving Microsoft’s own Services
It is not just outside companies that are using Azure and machine learning.
Microsoft AIs developments are also powering some to the company’s most popular applications.
These include Cortana their personal assistant and Bing, Microsoft’s search engine.
In the latter example, machine learning informs targeted advertising and also returns the most relevant results following a search.
Finally developed apps and solutions can also be shared in the Gallery or Azures own Marketplace.
This helps developers to monetise their solutions while making them available to a wider range of potential clients and users.
Microsoft AI and Azure are Powering Healthcare NExT
The Microsoft AI focus is not only impacting on the company’s output.
The applications that they are developing are also helping progress to be made in a number of other fields.
One such area is in healthcare.
Microsoft AI’s Azure cloud computing software is driving the company’s Healthcare NExT initiative.
Healthcare NExT aims to use Microsoft AI and Azure developed resources to improve healthcare provision.
This project has seen Microsoft AI integrating voice recognition software, robots and cognitive services into collaborative healthcare-focused applications.
These are intended to help medical providers deliver more personalised care and automate outpatient care plans.
These applications can also speed up medical record access time, streamline patient triage processes and make data entry simpler.
The University of Pittsburgh Medical Center is one of the United States largest healthcare delivery networks.
They were also the first healthcare providers to make use of Microsoft Ai’s initiative.
Healthcare NExT allows healthcare providers to access Microsoft Genomics.
This is Microsoft AIs attempt to enhance the sample-to-answer process.
Here gene testing methodology is available through an Azure-powered genome analysis pipeline.
Healthcare NExT is Transforming Cancer Treatment
Another Microsoft AI Healthcare NExT initiative is Project InnerEye.
This is a Microsoft AI-driven research-focused software tool for planning courses of radiotherapy.
InnerEye intends to provide oncologists with the ability to 3D contour the planning scans of a patient in a matter of minutes.
Without this application, this planning and contouring could take many hours.
This initiative is already being used in Addenbrookes Hospital in the UK.
Finally, Microsoft AI has developed a health chatbot.
This enables third-party software developers to construct their own conversational healthcare tools and applications.
This feature and other applications are being further improved by Microsoft’s acquisition of Skype.
Microsoft AI is incorporating Skype for Business into a number of virtual healthcare templates.
This is aimed at increasing versatility and functionality, making Healthcare NExT easier to use.
Helping Save Patients From Heart Disease
In India Microsoft, AI is being utilised in Apollo Hospitals, one of India’s largest private healthcare companies.
Here Microsoft AI solutions are being used to improve the detection rates of cardiac disease, highlighting patients most at risk from heart disease.
Previously prediction models in this area have been based on studies conducted in North America or Europe.
These models were inaccurate when used on an India patient base for a number of reasons.
For example one of the main causes of heart attacks in the west is high levels of LDL cholesterol.
Indian patients rarely suffer from high levels of LDL cholesterol, meaning that testing for it is largely useless.
Instead, Indian patients heart attacks are caused by other factors.
Microsoft AIs cloud computing abilities have been able to develop a more effective model.
This allows Apollo Hospitals to better identify which are most at risk of a heart attack.
Microsoft AI is Pioneering Motion Sensor Technology
Microsoft Ai’s motion-sensing input device Kinect was originally designed to be a video game controller for the company’s Xbox console.
The latest version, the Azure Kinect Developer Kit which was unveiled in 2019, is focused firmly on business applications.
Microsoft AI developers conceive the latest version as a tool able to build corporate applications.
These then seamlessly plug into Azure, Microsoft’s massive cloud computer.
Cameras and sensors allow Azure Kinect users to create Microsoft AI-powered applications based on facial recognition or body tracking.
Combining a series of these sensors can also allow the user to create a 3D map of a room.
Healthcare providers Ocuvera are already using this technology in their clinics.
Here it is used to identify potential patient falls in the hospital.
Hospital falls can cause patients further injuries and are, in some cases, fatal.
Microsoft AI’s DK can identify if a patient is about to fall and alert a nurse.
This means that the patient can be aided before any serious damage is done.
AVA Retail is also using Kinect technology.
Here it is being applied to power self-checkouts, allowing customers to grab-and-go.
A Vision for Augmented Reality
HoloLens is the name of Microsoft AIs developed mixed reality smart glasses.
It uses tracking technology that was first developed for Kinect.
HoloLens was the first head-mounted display to use Windows Mixed Reality platform.
HoloLens is a self-contained unit, employing eye-tracking technology alongside a large field of view and hand tracking software.
Since 2018 Microsoft AIs HoloLens has been supplied to the United States military.
Here the technology is used to “increase lethality by enhancing the ability to detect, decide and engage before the enemy.”
Initially available to purchase, HoloLens was subsequently made available for rental.
This was done in partnership with Abcomrents.
In 2019 at Barcelona’s Mobile World Congress Microsoft announced HollowLens 2.
This is a retooled version of the original HoloLens.
Initially, HoloLens was conceived as a game-playing tool, with enterprise possibilities.
HoloLens 2, while making the most of Microsoft AIs latest developments, has a more practical focus.
Microsoft AI intends to use HoloLens 2 as an AR tool to help people get things done.
This means that it seamlessly connects to Microsoft’s cloud services as well as android apps.
The technology of the Future
HoloLens 2, like its predecessor, makes use of a large field of view and hand and eye-tracking technologies.
However, now Microsoft AI has a second use for the eye-tracking cameras, biometric security.
Users of HoloLens 2 have their irises scanned automatically.
This allows the software to automatically sign in to their personal account, or remember their preferred settings.
Currently, this technology is more advanced than its potential client base.
Only the most advanced or forward-thinking factories and organisations are currently ready to make use of HoloLens 2.
One is the US Army, who were previous users of the original HoloLens.
As the world catches up the market will only grow.
Crossing the Platforms in Collaboration with Amazon
In 2017 it a partnership between Microsoft and Amazon was announced.
Jeff Bezos said of the collaboration, “I want them to have access to as many of those A.I.s as possible”.
This means Microsoft AI developed tools and services such as Cortana, Office 365 will integrate with Amazon services such as Alexa.
This partnership will, it is believed, be mutually beneficial.
For example currently, Amazon owns the shopping and buying domain, or Voice Commerce through Alexa.
Over time Cortana users will also increase access to the Amazon platform.
This increased usage not only increases Amazon’s profits but also benefits Microsoft as an Amazon affiliate.
The partnership also helps to cement Microsoft at the forefront of AI development.
Placing Microsoft and Apple at the forefront of the Voice First revolution.
This collaboration may pose an existential threat to both Google and Apple.
This will see Alexa, and Amazon, become more widely available on Microsoft AI-powered platforms.
Consequently, hardware, such as iPhones, and OS, may become obsolete, or at least less important.
The collaboration allows Alexa and Cortana, or Microsoft and Amazon, to reach across the limitations that come with platform domains.
Instead, this synergistic approach allows them to benefit from the power of each separate platform.
In the future, we may even see skills or apps being constructed that unite both platforms.
These apps will use features from both Microsoft and Amazon to create a powerful, overarching solution.
Currently, a customer uses one platform or device.
In the future, thanks to collaborations such as this, there may be only one, easy access, universal platform.
Microsoft AI is delivering Complex Analytics to Financial Service Providers
Here they are developing an AI-powered platform called Next Generation Complex Analytics.
This tool can simulate financial market behaviour as well as transporting networks and other relevant environments.
Next Generation Complex Analytics is useful for predicting financial market patterns.
These tools are also able to generate financial insights, highlighting both opportunities and risks.
Making the most of Microsoft AI developed tools, this has been tested on a number of NatWest’s services.
In one case these Next Generation Complex Analytics was used to analyse the Buy-to-Let housing market.
Here the tool will focus on house price fluctuations, monitoring how they affect mortgages and the demand for small business loans.
This information can be used by staff to decide the best course of action for improved financial performance.
Kevin Hanley is Director of Innovation and Solutions at NatWest.
He said of the technology, “By allowing us to better predict future outcomes, risks and trends, the implementation of this technology could be of significant value to our customers and shareholders over the coming years.
For the first time, we’ll be able to deliver an aggregated, forward-looking view of the world around us, ultimately helping us build a stronger, safer bank.”
Project Brainwave Delivers a High Powered Chip
Project Brainwave is a Microsoft AI-driven initiative aimed at accelerating machine learning algorithms to work in real-time.
It is also helping to accelerate Microsoft AI technologies that power the Bing search engine and Azure platform.
This sees Microsoft AI tools using FPGAs, programmable processors, to power sophisticated machine learning algorithms.
Project Brainwave technology can be programmed straight onto a chip which, when installed in hardware, allows the hardware to function as a processing unit for a deep neural network.
Microsoft Ai is so confident of the power of these FPGA chips that they are already installed in many of the company’s data centres.
Microsoft AIs FGPA chips and tools use machine learning to drive a number of the company’s larger products, such as Cortana, the virtual assistant or the chatbots that operate on Skype.
Here machine learning is used to help the applications improve and develop the more that they are used.
Bing and Office 365 also make use of this intelligent functionality.
Currently, AI, machine learning and similar tools are ideal for set, or restricted tasks.
Microsoft AI intends to develop further in future. The aim is to produce intelligent machines that have AI capabilities.
These machines will be capable of completing any given task.
Harry Shum, executive vice president of Microsoft AI and Research, said “Computers today can perform specific tasks very well, but when it comes to general tasks, AI cannot compete with a human child.”
Microsoft AI Initiatives and Applications Have Helped the Company to Remain at the Forefront of Technology
In recent years Microsoft has transformed its model.
Using Microsoft AI-focused initiatives has allowed the company to reassert itself as a market leader.
Adopting a forward-thinking approach allows Microsoft AI to drive innovation and progress in a number of different ways.
These include transforming healthcare and cancer treatment, via Project Hanover, to platforms such as Azure and HoloLens technology.
Throughout their history, Microsoft has prided itself on delivering accessible, useful software applications to a wide market.
Even today, Microsoft’s most advanced applications are made to be as accessible and usable as possible.
Microsoft has long been a major presence in the technology world.
Their newfound drive to innovate means that they look set to remain a major presence for a long time to come.
Images: Flickr Unsplash Pixabay Wiki & Others