Artificial intelligence and business - from fear to value

June 12, 2018

Interview with AI expert Anstassia Lauterbach

Following her guest lecture "Artificial Intelligence in Business & Society: Quo Vadis?" for the Global Executive MBA, Dr. Anastassia Lauterbach, former top-manager, investor, entrepreneur and bestseller-author, gave an exclusive interview about the topic. The AI-expert talks about how businesses can prepare themselves for the AI-revolution, how AI is going to influcence work and employment and what the 5 most widespread AI myths are.

Picture of a manager with a robotic hand
Is AI a reason to be scared or can it be a contributor to value creation? Anastassia Lauterbach gives answers in the interview.

In your new book called “The Artificial Intelligence Imperative” you provide a practical roadmap for businesses to prepare themselves for the AI revolution. What are the three most important aspects/challenges a business should take into consideration when dealing with AI?

While advising any business I think along three categories: Data, Architecture and Talent. A company needs to develop what I call a Data Value Strategy to crystallize priorities in strategic and operational questions its data might be capable to support. As an example, if you are in a media business, you are well served to understand what content drives the most attention, what readership groups discuss on social media and what your competition is publishing. You can use unstructured data and social listening technologies to understand patterns and identify the things that matter most. You can build new content and levels of engagement based on this input. Of course, you need an IT environment to process this data with the help of Machine Learning or other AI technologies. Here again, a company has to prioritize which IT stack needs updates first, what kind of data processing environment is best, how to build a data lake and who to work with in hardware. Last but not least, there is no one single company to accomplish everything 'in-house'. Every business requires comprehensive workforce education and employment strategy. Businesses will get used to work with freelance experts, interim managers and outside partners, and hire new (and diverse) talent. The ability to embrace new ideas and accept different points of view is crucial.

If a rather traditional company has no experience with AI and is thinking of whether or not implementing AI into its core business processes, what would you recommend?

AI is just a technology. As Picasso said, „Computers are stupid. They can only give you answers“. A company needs a list of strategic and operational questions before thinking about what type of a Machine Learning technology to implement. Besides, it needs to understand the regulations around personal data, discrimination and transparency on technology used. Once these internal and external perspectives are clarified, the current IT environment has been reviewed and target architecture drafted, one can move into Machine Learning. I am very cautious with companies in enterprise software jumping onto the AI bandwagon to beautify their offerings. Sometimes it is better to get partners with healthy industrial expertise, even if they are small. Last but not least, any business needs to understand how it uses cloud technologies powered by AI with Big Tech such as Amazon or Microsoft, and which companies it shares data with. While piloting Machine Learning, companies should adapt a portfolio approach and be not afraid to fail, as not every model will work, and not all suppliers will deliver. Last but not least, AI is a wonderful force to rethink a company’s approach to its data governance and organizational structure. If implemented consequently, every single employee will add data to its source of knowledge, learn and make decisions in real-time. This leads to people empowerment on a very different scale from what traditional companies - with larger layers of financial controllers and middle management - are used to.

Picture of a thinking robot
AI is a wonderful force to rethink a company's approach to its data governance.

How will AI affect the future of employment? What are the most difficult ethical and social issues to be solved in this regard?

AI and employment is about executive leadership. Leadership implies courage and long-term thinking. If a large company does not have a long-term AI strategy, it will disregard how automation changes requirements for skills, what disrupts communities a business is working in, how to adapt an aging workforce to AI, and what diversity truly means. Leading diverse teams is challenging. Cross-disciplinary and -cultural expertise will be a must, multilingual skills required. Passive aggressiveness of those not willing to adapt will accompany change. Thankfully some large funds like Blackrock start to consider long-term strategies of their investment targets. I expect within the next 10 years this thinking will grow, forcing changes in corporate boards and executive teams. The future of workforce along with questions on environment and corporate responsibility will belong to sustainability frameworks of businesses. Not everyone will succeed. Short- and mid-term, we will experience a further raise in inequality, louder discussions around frameworks such as the Universal Basic Income and geopolitical fears of losing the competitive edge to China.

The CEO of Alibaba, Jack Ma, reckons that there will be robo-executives and even robo-CEOs - according to him, robots are more objective and less sensitive than humans. What do you think, will robots be the better CEOs?

According to the latest McKinsey study on Automation, CEO is a job category which AI does not impact a lot. CEO's job should be about leadership and strategy, not about scaling one process or minimizing errors in one task. As Yann LeCun of Facebook says, today’s AI has an intelligence of a rat, recent Chinese research indicates that it is comparable to the IQ of a five-year old. We need very a different approach to research and development of Deep Learning, going beyond twitching ways to get more out of Neural Networks. We need evolved silicon, which might be based either on quantum or neurotrophic computing - both are still in their infancy. Jack Ma might need to wait for another 30 years to come closer to his vision on replaceable CEOs.

What are the five most widespread myths about AI?

  1. A company will be capable to embrace AI without changing its executive leadership and culture

  2. Artificial General Intelligence is achievable within the next 20 years

  3. Harmful algorithms (e.g., in Automotive) will lead to accidents

  4. Europe has still a competitive edge in AI

  5. People without a degree in math or engineering can’t work in AI

Among many others, artificial intelligence and big data are topics of the Global Executive MBA. For more information about the program, please click here.

Share this