The Gap between AI-Haves and AI-Have-Nots Will Be Larger
Jan 25, 2018

He is in charge of the Microsoft search engine Bing, personal assistant Cotana and cloud platform Azure. He comes from Taiwan and is now CVP of Microsoft AI Core and CTO of the AI and Research Division. He is David Ku, the person in charge of integrating AI function into Microsoft’s essential products.

At the 2018 CommonWealth Economic Forum David Ku said that data is the artery of AI and will become a new form of capital. How do businesses turn data into capital and reflect it on actual operational performance?

The following are excerpts of his speech:

When Bill Gates founded the Microsoft research lab 25 years ago, he had a vision that computers can one day see, talk, and understand humans. And it’s with that vision and ambition we have seen prominent progress and investments around AI over the past decades. The development of AI has its setbacks. However, the proliferation of computing and artificial voice recognition has created new possibilities. If you look at scenario like self-driving car, it was actually proposed in the 1990s. It has taken us so long to get to where we are today.

Why is everyone so enthusiastic about AI today? What has created the momentum and energy around AI? There are three major reasons: better algorithms, more data and bigger computing capabilities.

In particular, there is a paradigm shift in algorithms, in the way we think about computing. Now we have the deep neural networks. The more data you give to the model, you can get better results. This is what classical machine learning is not able to accomplish. There is now proliferation on finding algorithmic solutions in areas that was once thought not possible, such as self-driving cars. Moreover, this push on AI has spurred a new industry on developing innovative ways of computing like GPUs, providing a solid ground for AI’s development.

AI Can Improve Business Performance

What does the application of AI mean to businesses? AI has created great opportunities for companies to harness the value of their data. Through collecting data and predictive analysis, their business performances can be significantly increased.

According to the estimation of IDC (International Data Corporation), AI’s application would create over a trillion dollars of net new incremental value in the next 3 years. There is a great momentum on AI because everyone thinks it can unlock the value of data.

How can business make better decisions and properly manage their supply chains? This is also related to data. Collecting data from customers can help business with decision making, reasoning and analyzing. And with good feedback mechanism, they are able to provide users with the most in-time and responsive service.

Most importantly, AI can improve the interaction among businesses, employees and customers. It is capable to help business to solve operating and producing challenges, knowing what can be done in order to optimize the organization.

Innovative applications of AI open new doors for traditional industries such as agriculture and medicine. With sensors and drones, small farmers can now closely monitor and manage their crops to maximize production. AI can also be used to solve medical conditions, addressing the aging populations. For instance, Microsoft has projects where doctors can use AI to create accurate 3D models of cancer or other diseases from MRI and X-rays scans. This optimizes patients’ chance of recovery.

In fact, AI is fundamentally digital transformation. Every company can digitally transform its business. At the industry level, several industries cooperating and sharing information together in the same supply chain would create a better forecast of demand, which ultimately benefit the consumers.

Turning Data into Capital

AI also has great impact on the society level. How will digital transformation influence universities, NGOs or even countries? How do we embrace this strong, furious tide of transformation?

From the standpoint of Microsoft, there are three dimensions to be discussed. It first starts with data. Data is the lifeblood of AI. It is the new means of production. Data is the new capital because it can bring more value and data to be used in models and diagnoses, which create bigger value in the future. Therefore, what businesses should do is take a fundamental look at where their data assets are.

Secondly, once we identify the flows of data and how the data assets are produced, we need to create systems of intelligence that connect these data signals. When we know the causalities and sequences, we are able to create a feedback loop which can be applied to building accurate predictive models or reducing defects. All of these require the data to flow. Data is less valuable if it is stale or disconnected.

Ultimately, this is not about updating computers to a newer version or using new devices. This is about understanding the flow and value of data, knowing what kind of ‘muscles’ you need to build. This is about rethinking how you interact with customers. This is about driving a new digital transformation in your business operation.

In sum, data can give tremendous power to those who can leverage AI in the most effective way, which may create inequities in the process. The distance between the AI-haves and AI-have-nots will become larger.

This decision of digital transformation concerns not only businesses, but also industries, society and the whole country.

Three Major Values of AI

What should we do to create ‘good’ AI, which benefits everyone? Here are three major values that Microsoft is currently pursuing.

The first one is to democratize AI. AI is a tool, a kind of software. We need it to be inclusive, so more people have access to information. The business world should collaborate with each other, even establish data co-op programs. With data sharing, decisions and plans can be executed in a better way. In addition, by contributing back to the providers of data, we can create a healthy value chain.

The second is being proactive in creating responsible AI. We need to think of security and responsibility when developing AI, making sure that it does not harm people. Last year, we started ‘Partnership on AI’ that comprises over 40 companies and 30 NGOs. Our goal is to understand and discuss the nuisances and ethics of AI.

Last but not least, partnership is going to be the key. AI’s success depends on not only technology but also data. However, the owners or drivers of technology do not necessary have the domain expertise or the data. The partnership that brings the technology and domain together will let AI benefit more people in various fields.

AI is a technology revolution. But let us not forget that it is never about the technology alone. It is always about technology enabling people, companies and societies. We want to augment human ingenuity with AI. Moreover, we want to empower every person and every organization on the planet to achieve more, reaching their full potential.

Translated from the Chinese Article by Shawn Chou