Why everyone in the boardroom needs AI

A new employee in the boardroom

In the Salesforce boardroom, there’s a new employee who helps the CEO make decisions, solves disagreements and helps the company look to the future. It’s called Einstein.

Similar ‘employees’ are infiltrating boardrooms at companies all over the world, with varying names, objectives and levels of importance. At Hong Kong-based venture capital firm Deep Knowledge Ventures, it’s called Vital (Validating Investment Tool for Advancing Life Sciences). At Merck and Columbia Sportswear it’s called Aera, purchased from the enterprise tech startup of the same name.

Einstein and these other AI solutions help their teams make all sorts of decisions, using vast amounts of external data. Whilst we are a long way off from a world fully integrated with AI, the boardroom is one place that AI can and will flourish in the near future.

How is AI already being used in the boardroom?

At Salesforce, Einstein sifts through massive amounts of data to keep CEO Marc Benioff informed. In an evaluation around the performance of various markets and executives, Einstein is able to find trends and patterns showing strengths and weaknesses, and to indicate where attention is needed. Marc Benioff claims: “Every question I could possibly have, I’m able to ask Einstein.”

Vital at Deep Knowledge Ventures is credited with rescuing the company when it was on the brink of bankruptcy. The venture capital fund was investing in too many ‘overhyped’ projects in the biotech industry, a notoriously difficult one for investors with its very high failure rate. With Vital, they were able to analyse big data that revealed patterns of risk for their investments. Goldman Sachs Asset Management announced a similar tactic to help them choose wise investments, in 2017.

Despite the value and results that have already been demonstrated, companies such as DKV and Salesforce are in a minority of boardrooms currently leveraging AI.

When Harvard Business Review surveyed 3,000 executives to see how businesses currently use artificial intelligence (robotics, autonomous vehicles, computer vision, language, virtual agents and machine learning), the results painted a narrow picture of AI use. Only 20% of their survey respondents said that they used one or more AI technologies as a core part of their business. 41% claim to be experimenting or piloting AI. Looking at these results it’s clear that the intention to adopt AI is there, but we are yet to get to the tipping point of AI appearing in companies at multiple levels.

What will the future of boardroom discussions look like?

Is the traditional company makeup under threat?

At the moment it seems like systems such as Einstein won’t be replacing anyone in the boardroom; rather they will complement their human counterparts, allowing decision-makers to free up their minds and their time to think about the larger strategy: “People and machines will complement each other in a way which will help them achieve what neither could if left alone,” says IBM’s VP of Cognitive Computer Guruduth Banavar.

For the moment it’s clear that AI needs humans as much as humans need AI. Without building the right framework, hiring the right people and using AI in the most optimal way, a technical solution will not add value to a boardroom. It’s unlikely that we will soon see AI replacing CEOs. But as a compliment to the C-suite, AI holds an infinite amount of possibilities.

Understanding risk

‘Safety’ will also become a buzzword when it comes to AI’s involvement in the boardroom. It’s possible  that in the race to get an AI product out, some developers can unwittingly release a product that is biased. Even a product created under exemplary conditions, however, must be used in the right ways, necessitating human involvement at some level. Is the data biased? Should its results be trusted, or analysed further? Does the AI system cancel out an executive’s gut feeling, or are decisions always a mix of the both?

Whilst in an extreme way Elon Musk talks about AI and its risk for the human race as a whole, it is vital that companies understand how to properly use AI for boardroom decision-making. KPMG recently stated: ‘if the risk team is not knocking on the boardroom door to discuss AI right now, either about the risks of using AI…they soon will be’.

What should business leaders be doing now to work toward implementing AI?

Google recently acquired Kaggle, tapping into thousands of data scientists, both amateurs and professionals. If even Google has to partner up to be able to achieve its AI goals, other companies will have to think about doing the same. Implementing AI solutions in-house can be expensive, so companies should think about looking outwards to achieve what they need.

As mentioned at the beginning of the article, companies such as Columbia Sportswear and Merck use AI software created by Aera. Vital was created in-house by DKV and Einstein by Salesforce, but even the latter are looking to integrate their AI solution with IBM’s Watson to expand their product into language understanding. So if you are looking to incorporate AI into your decision-making processes, there are plenty of options outside of creating a new department with in-house data scientists.

The issues with safety have already been highlighted above, but what can a company wishing to implement AI into its boardroom decision-making processes do to temper the element of risk? Feedback will be key. It allows time to identify inaccuracies, either at the beginning or end of the AI solution’ process. Having a human check an algorithm allows for better ethical standards, but technically a computer is a better thinker. This risk element will require some sort of investment: auditing, risk assessment, and interacting on the artificial intelligence solution.

It’s also essential that companies are digitally integrated now. It’s no coincidence that the most digitally-minded companies are the ones that adopt AI the fastest. The HBR survey revealed that there was a 50% higher chance of generating profit in companies that have strong experience in digitization. Laying the foundation as soon as possible will enable AI transformation to have the greatest impact.