Archive for Microsoft

Quantum Computing Gets a New Player

Feynman Diagram

This was originally published on the Amalgam Insights website.

Last week (the week of September 25th, 2017) Microsoft made a huge announcement at its annual Ignite and Envision conference. Microsoft has become one of a small number of companies that is demonstrating quantum computing. IBM is another company that is also pursuing this rather futuristic computing model.

For those who are not up to date on quantum computing, it uses quantum properties such as superposition and entanglement to develop a new way of computing. Current computers are built around tiny electron switches called transistors that allow for two states, which represent the binary system we have today. Quantum computers leverage quantum states that give us ones, zeros, and combinations of one and zero. This means a single qubit, the quantum equivalent of a bit, can represent many more states than the bit can. This is, of course, a gross oversimplification but quantum computing promises to deliver more dense and exponentially faster computing.

There are a number of problems with practical quantum computing. The hardware is still in a nascent stage and must be cooled to a temperature that is quite a bit colder than deep space. This makes it much more likely that quantum computing will be purchased via a cloud model than on-premises. The other inhibitor is that there is no standard programming model for quantum computing. IBM has demonstrated a visual programming model that shows how quantum computing works but is clearly not going to be a serious way to write real programs. Microsoft, on the other hand, showed a more standard looking curly bracket programming language. This application layer makes quantum computing more accessible to existing programmers who are more used to the current model of computing.

When quantum computing becomes practical – I would predict that is at least 5 years away, perhaps longer – it won’t be for everyday computing tasks. The current model is already more than adequate for those tasks. It’s also unlikely that the capabilities of quantum computers, especially the information dense qubit, and costs will have much a place in transactional computing. Instead, quantum computing will be used for analyzing very large and complex data sets for simulation and AI. That’s fine because the AI and analytics market is still new and the future needs are not yet completely known. That future computing needs is what quantum computing is meant to address. Even today’s big data applications can stretch computing capabilities and force batch analytics instead of real-time for some use cases.

Microsoft’s entry into what has been an otherwise esoteric corner of the computing world signals that quantum computing is on the path to being real. It has a long way to go and many obstacles to overcome but it’s no longer just science fiction or academic. It will be years but it is on the way to becoming mainstream.

Microsoft Infuses Products with Machine Learning and the Social Graph

Micsoroft plus LinkedIn Social Graph

This article was recently published on Amalgam Insights.

 

This past week (September 25 – 27, 2017) Microsoft held it’s Ignite and Envision Conferences. The co-conferences encompass both technology (Ignite) and the business of technology (Envision). Microsoft’s announcements reflected that duality with esoteric technology subjects such as mixed reality and quantum computing on equal footing with digital transformation, a mainstay of modern business transformation projects. There were two announcements that, in my opinion, will have the most impact in the short-term because they were more foundational.

The first announcement was that machine learning was being integrated into every Microsoft productivity and business product. Most large software companies are adding machine learning to their platforms but no company has Microsoft’s reach into modern businesses. Like IBM, SAP and Oracle, Microsoft can embed machine learning in business applications such as CRM. Microsoft can also integrate machine learning into productivity applications as can Google. IBM can do both but IBM’s office applications aren’t close to having the market penetration of Microsoft Office 365. Microsoft has the opportunity to embed machine learning everywhere in a business, a capability that none of their competitors have.

For the average knowledge or office worker, having machine learning embedded in the applications that they use every day means help doing their normal daily tasks. Machine learning in of itself is useful for analysis or automation. The real power of machine learning will be most evident when it is available to help with everyday tasks such as scheduling, analyzing data in spreadsheets, managing customers, developing financial projections, enhanced search capabilities, and creating impactful presentations. For corporate workers, this is the type of AI technology that helps them to do their jobs better, making them more valuable rather than obsolete.

The second was the integration of Microsoft 365, Microsoft’s social graph platform, with LinkedIn and Dynamics social graphs. Since Microsoft purchased LinkedIn, the big question has been “how will they leverage it?” The answer is finally here. By combining the Microsoft 365 (contacts internal to the company) and Dynamics (contacts external to the company) social graphs with the personal connections of employees, Microsoft products will help to surface and leverage useful relationships no matter what they are. There are some obvious advantages. First, it means being able to connect the right group of people together to accomplish something no matter how they are linked. By breaking down the walls between the personal, internal, and external relationships, Microsoft will allow knowledge workers to find and foster essential connections that are at the heart of business. The most obvious beneficiaries will be sales and marketing by surfacing paths to and intelligence about people that they need to do business with. Social graphs also represent rich data that can be mined for a variety of other purposes. Since the Microsoft 365 is extensible, other information about people, relationships, and location can be added to the extended social graph. This will create a rich pool of information that can be mined for a variety of purposes. Applications include finding and managing vendors and partners, recruiting new personnel from internal and external pools, identifying better ways to communicate, and seeking out M&A targets.

Microsoft’s announcements are, aside from the quantum computing announcement, more incremental but in a good way. They are taking highly hyped but useful technology and making it relevant to the masses. Both machine learning and social graphs will no longer be primarily the realm of specialized applications. Instead, through inclusion in Microsoft’s most ubiquitous apps especially Office 365, they have the potential to become part of the fabric of everyday work life.