Bringing AI to everyone
Microsoft’s approach to making artificial intelligence mainstream
Microsoft, one of the most valuable companies on this planet, is heavily focusing on AI as the go-to technology for their future success. In the following blogpost we will try to unpack
- whether the claim “bringing ai to everyone” can and should be taken literally, and
- what the strategic considerations behind the positioning towards AI are.
Background
AI is currently praised as a secret source and a universal remedy by almost all software and IT platform providers in terms of marketing. It is often difficult to separate the flowery words and descriptions from actually valid facts and independent performance characteristics.
“We want to pursue democratizing AI just like we pursued information at your fingertips”
Satya Nadella, CEO
In a previous blog post, we already gave an overview of the various cloud platforms claiming the topic of intelligent automation and process integration for themselves. After the first wave of digitization, where the focus was primarily on data storage, access options and communication, the key technological component now is machine learning, which naturally benefits greatly from the enormous data volumes and immediate availability in the cloud and has made considerable progress in recent years.
In 2014, Microsoft CEO Satya Nadella announced the cloud-first strategy for the company to prevent another missed entry as seen before in the mobile business. Furthermore, the aim was to transform the company from an outdated single software license business for operating systems and productivity applications towards SaaS business models. Since then, market capitalization has roughly quadrupled within 6 years, making Microsoft one of the most valuable companies in the world. After Amazon Web Service, Microsoft’s own cloud platform Azure is the second largest cloud provider with well-known partners and clients that contributed almost one third of total sales in 2019 with around US$ 39 billion.
If you as a user or decision maker are thinking about using AI in your company or want to transfer your first experiments into a scalable cloud infrastructure in order to work productively with machine learning, Microsoft is most likely a partner you should take a closer look at. The reason is obvious: Almost every company already uses products in some form. But let’s take a closer look at what the company wants to achieve in order to find out whether it can be the right partner:
Microsoft’s AI Vision
Microsoft’s vision statement regarding AI can be summarized as follows:
“Our vision for the enterprise is to enable every company to transform by bringing AI to every application, every business process and every employee”.
Clare Barclay, COO of Microsoft UK, is even more specific in terms of the time frame:
“Based on the progress we’re seeing, we believe that every company will be an AI company in five years”
If one follows this hypothesis, it is not even an ambitious goal — as the market leader for operating systems and productivity tools in the professional and enterprise segment, it is inevitable to adapt to technological change in order to stay relevant. Developments in the mobile and cloud sector have shown that technology can penetrate markets at high speed and turn them upside down, even though penetration is still not 100%.
Without having insider information about the detailed execution plan to bring this vision to life, there are four obvious business clusters that will play an important role on this journey:
- Strengthening the Azure Cloud Platform as AI infrastructure and central growth engine of the business model
- Incremental improvement of the widely used office, productivity and communication tools in the private and professional context
- Fostering the use of analytics and database tools
- Developing new “AI-first” business areas through cutting-edge research and development
Strengthening the Azure Platform
Placing Azure as the center of the AI tools and services is obvious. The cloud market as a whole is growing at double-digit rates and this development will continue in the coming years. The main drivers of the hunger for virtual resources (storage and computation) are IoT and multimedia data (streaming) being created and analyzed in huge amounts.
This business segment, which already generates one third of the company’s revenue today will in future account the lion’s share of income. Machine learning naturally plays an important role and is an absolute must in order to maintain a relevant market share in the cloud area. The most exciting question here is how the various players (including AWS, Google, Alibaba) position themselves, what target groups they focus their offering on and how they manage the interchangeability of pure cloud business in terms of customer loyalty.
Improving the SaaS product portfolio
The Office 365 and the Dynamics 365 product portfolios, which specialize in enterprise solutions, are central components of Microsoft’s SaaS offering. In order to strengthen customer loyalty, remain competitive, and to ensure price stability, the infusion of AI components plays an important role.
This is not about disruptive new product innovations, but rather about the smooth integration for example of
- translation services into MS Word
- extension of MS Excel with basic machine learning applications for small and medium data volumes
- optimization of spam filters and auto-responses in Outlook
- general searchability of files and folder systems
- smart tagging of image and video content
- efficiency optimization or automation of repetitive routine tasks detected by a personalized background AI assistant.
The strategic orientation of Microsoft in this field is certainly to play defense: consolidating one’s own dominant market position and securing price stability against new competitors.
Fostering analytics and database tools
Almost all projects that aim to generate insights from data rely on database infrastructure and reporting tools. Very often relational database management systems such as Microsoft SQL are in place and used in combination with reporting tools such as PowerBI to gain insights that do not necessarily have anything to do with machine learning, but are important to keep the day-to-day business running. AI will drive companies towards more data-driven decision-making and to invest in their data infrastructure. For this reason alone, Microsoft’s business intelligence and database management system tools will very likely see positive impact — independently of the overall cloud buzz.
Development of new business fields
Microsoft already serves 90% of the Fortune 500 companies and has the pressure from investors to tell more and more growth stories. For this reason alone, the company cannot rest solely on existing business areas, but must make enormous efforts to open up new business areas. Since the entire structure of the company is geared towards the mass market, future niche applications for certain verticals are unlikely as they are simply not in the DNA of the company.
Specialization only makes sense in very selected areas such as gaming (with Xbox as core brand asset) as well as the healthcare sector, which turns out to be one of the most promising growth areas and an active research field of the MSRAI (Microsoft Research Labs for Artificial Intelligence).
Acquisitions like LinkedIn and GitHub underline the course of successively adding suitable business areas of special interest and moving quickly towards general interest or at least a broader audience. Nevertheless, the data treasure of these platforms offers potential for new business models, such as the early recognition of content trends and technical development topics. Potential M&A targets include Adobe and DocuSign, both of which would fit the profile well but would require high investments. A further segment of interest is the field of Robotic Process Automation with suppliers like UiPath and BluePrism.
Partner Network
A technology leadership, that Microsoft strives for according to its own vision, always includes a strong network of supporters and partners who act as promoters and specialized solution providers.
Especially for small and medium-sized customers, personal support, contact persons and advice on implementing new technology are essential. Based on its existing product portfolio, Microsoft already has a high penetration in companies of all sizes and maintains a professional network of business partners like almost no other company.
Implementation partners range from medium-sized IT-consultancies to market leaders such as EY, Accenture, KPMG and Capgemini. In the area of process automation, Microsoft has established partnerships with Ui Path and BluePrism, in the hardware area, relationships are maintained with chip manufacturer Nvidia. Even for industrial applications, ABB is a prominent partner for the corresponding sales segment.
Moreover, Microsoft is active in a variety of research collaborations and non-profit engagements to foster the development, awareness and security of the technology on a global scale:
- Partnership on AI (study and formulate best practices on AI technologies, to advance the public’s understanding of AI)
- Center for Brains, Minds + Machines (interdisciplinary study of intelligence)
- Carnegie Mellon University’s Machine Learning Department
- The Amsterdam Machine Learning Lab researches large-scale modelling of complex data sources
- ONNX Microsoft and Facebook joint initiative (Open Neural Network Exchange) to move models between state-of-the-art tools
Responsibility and Ethics
Putting the pure business aspects aside, a transformative approach to focus the entire company on a new guiding technology also requires a strong shift in the mindset of the people. And it is also necessary to attract new talent driving the change and taking the company to a new level. Ethical principles are, in this context, an important part of being able to conduct cutting-edge research in the field of AI, attract the right talents and give the growing customer base confidence to handle their highly sensitive data responsibly.
Especially with the goal of many AI researchers to advance the development towards Artificial General Intelligence (AGI), it is necessary to deal with the scope of technological progress in this area and the responsibility for potential misuse seriously. Microsoft approaches this issue with a code of conduct, which it places at the center of all AI-related efforts — the six principles:
- Fairness
- Inclusiveness
- Reliability and Safety
- Transparency
- Privacy & Security
- Accountability
Another dimension in this context is social responsibility and the possibility of using technology to improve the world. This, of course, in on the one hand part of corporate responsibility for companies that use infrastructure and resources all over the world for its business success. On the other hand, it is a very effective employer branding tool offering increasingly purpose-driven employees the opportunity to work on applications that do not solely pursue capitalistic, but also pursuing humanitarian goals. Certainly at least partly inspired by the philanthropic efforts of Microsoft founder Bill Gates, the company participates in the global “AI for Good” initiative, which is a platform initiated by the UN and ITU to foster AI application for improving the world. Many tech companies and researchers along with Microsoft participate in this program taking on challenges in certain areas such as environment, healthcare, education and others.
Education and AI Readiness
Microsoft faces two key challenges in developing and focusing on AI. The first is to transform its own company, its employees and structures according to the new technology, changed working methods and required competencies. This means new employees, creating attractive working environments, promoting training and development of existing employees, and creating independent, agile work organizations that are able to create innovations without bias.
“There is a gap between what people want to do and the reality of what is going on in their organizations today, and the reality of whether their organization is ready”
Mitra Azizirad (Corporate Vice President for AI Marketing)
The second challenge is that the “product” AI cannot simply be sold out-of-the-box like the Windows operating system. The use of AI currently includes intensive examination of the topic on the customer side — finding suitable use cases, building up competencies and adapting organizational structures. Only then customers are able to accept solutions and infrastructure from a service provider like Microsoft on a relevant scale.
For this reason, Microsoft has probably decided to establish the AI School as its own further education platform and presales tool. It’s a cross-media conglomerate of content from videos to tutorials, code snippets, exercises and documents. The modular content library is designed to effectively guide users through specific tasks, expand their general topic knowledge as well as outline the importance of organizational and cultural adoption for the productive use of AI applications. The latter is stressed in the AI Business school, which strongly focuses on the importance of having a clear corporate data and AI strategy as part of the overall business strategy and continuously involve top decision-makers into the process to avoid mistrust or acceptance issues with the final solution.
Review and conclusion
After this detailed outline of Microsoft’s activities to accomplish its communicated vision for artificial intelligence, let’s briefly review these measures and ambitions in the context of competition, customer acceptance and potential failure.
After being not a first-mover in the AI space, Microsoft adapted its strategy and put out a very comprehensive set of different instruments as you have seen along with some rather bold and optimistic projections regarding the future of AI.
What could go wrong?
1. Technological progress
Assuming that every company will an AI company within 5 years is very bold. But it is important to mention that we are not talking about futuristic utopia here. The focus overall is very much tailored to Horizon I which — if you somehow missed the McKinsey bootcamp — refers to:
Ideas that provide continuous innovation to a company’s existing business model and core capabilities in the short-term (1–2 years)
Microsoft focuses on the small and almost invisible hacks and improvements as part of everyday tools like translators, search and recommendation engines, etc. which are as of today already infused by machine learning algorithms. The goal is to tackle the minor day-to-day problems and inefficiencies. This is completely feasible with today’s technology, and does most likely not rely on cutting-edge research results. That’s not a negative assessment by any means as Microsoft is obviously also involved in cutting-edge research. But as an overall business strategy they put more resources behind the challenge of pushing mass adoption.
Bottom-line: Even if AI research progress does not continue to hit milestone after milestone like the past couple of years, Microsoft’s almost secret infusion of AI into everyday’s business processes could work out independently of the overall hype.
2. Customer acceptance
Acceptance is, of course, a very important aspect in order to scale products around a new technology. And it has two major dimensions: The first one is trust in the technology itself and all its implications including security, data misuse, privacy, necessity, ROI, along with subjective fears about gloom and doom scenarios.
The second dimension is acceptance of the products and services in comparison to alternatives in the market. In this context Microsoft is as trusted business partner for the last decades ahead of competition and does a lot to preserve this position.
But: Microsoft has also a tradition of being not the sexiest alternative in peer-discussions of developers and tech-nerds. As a platform for “everyone”, it is just natural not to attract innovators and first-movers to a large extent, who do not value convenience but are rather excited by the complexity and trial and error progress of doing something completely new. It is clearly a branding issue, and it has already improved in the past as Azure turns out to be an open and reliable platform to build applications on.
Bottom-line: With the AI school and ethical principles, Microsoft seems to be aware of the technology acceptance risk and tries proactively to improve the situation. In terms of the branding issue it might make sense to separate core developer tool-kits together with assets like GitHub from the everyday business and productivity products.
3. Making no money
Last thing that could negatively impact the ambitious plans is strong competition that inhibits the necessary monetisation. In the end, money talks. As of today, many new vendors enter the space of AI tools and cloud solutions and like no other technology AI is driven by a large open-source community. Frameworks like TensorFlow, PyTorch, Keras and others are major building blocks for the development process. They can be seemlessly integrated, but do not generate revenue.
All cash comes from cloud computing and storage along with some useful services for data engineering and warehousing. Prices go down constantly, performance goes up constantly. If the resource demand continues to increase at the same pace, that’s not a big deal but the sheer scale of the market leader could at some point be a problem for very attractive margins of all the others. Amazon has played this “squeeze-out” game already in commerce, they could very aggressively do the same in cloud computing.
Other more specialized vendors like SAP, Salesforce, Oracle also enter the space to provide solutions around their core SaaS products. Alibaba, Baidu and Tencent invest extreme amounts of money, with which they also drive international expansion. Their big move towards international markets is yet to come but not far away. And then there are the most obvious competitors in the field of cloud generalists: AWS and Google along with IBM. Currently, Microsoft and Google take on the broadest position while AWS and Alibaba focus more on professional users, providing the highest possible degree of customizability. SAP and Oracle have — by the nature of their core products — already specialized users and developers, while Salesforce and IBM address a broad and partly non-tech user group but try to keep their focus on applications close to their core business.
Bottom-line: Trying to be the universal solution for literally everyone in terms of AI is a high risk move that could backfire as other vendors get better in their specific domains. Nevertheless, Microsoft’s overall strategy seems very solid and the incremental step-by-step infusion of the existing tool and service landscape totally makes sense. The question will be whether this enough to keep the pace of its competition and tell compelling stories to customers, investors and partners.
