Leaders’ framework to scale AI – What it really takes to scale artificial intelligence

From being a data-hungry threat to our privacy to beating a multiple-time Go world champion, AI has been making headlines for years. AI already creates recommendations for movies we might like, targets us with highly accurate ads, helps doctors detecting cancer, decides our insurance premium, and assists with logistics.

AI has proven to be revolutionary: its market value has been expanding from 58.3 billion USD in 2021 to an estimated 309.6 billion USD by 2026 [1] and, technologically, it has been evolving at a pace that is hard to keep up with. However, many companies are not investing enough in the cultural and organizational challenges required to scale AI to the point of generating meaningful value. To capture the full value AI can give, companies must reimagine their business models as well as the way work is accomplished.

While AI is often applied to individual use cases in company-specific silos, this approach will not bring consequential change to the bottom line of the company’s operations. That also makes scaling AI exponentially more complicated since each team tackles training, change management, data, technology, and other problems individually and without using synergies. On the other hand, a comprehensive overnight change would end up most certainly in failure. The key is to find a part of the business to start with, where the quickest, most efficient but also big enough impact can be achieved to initiate the chain reaction in the company.

In the end, the ones that won’t take full advantage of this technology will be left behind by the ones that will have managed to exploit it. Even currently the outperformers of tomorrow are the ones that know how to use AI with their assets. That can already be seen in all kinds of industries, from investment and manufacturing to logistics and medical.

Here are some practical steps on how to prepare your company to aim high and get the right results:

5-steps-to-scale-AI
  1. Realizing the need for change

Leaders in traditional companies (non-digital native) struggle to move from MVPs to company-wide solutions. For the most part, it is because they don’t see that cutting-edge technology does not only require single talent, but it also demands just as many significant changes in corporate culture, structure, and mindset. Often leaders think that AI is a plug-and-play solution, expecting a big ROI by making small tweaks here and there. Therefore many businesses are not investing enough time and resources in the organizational and cultural change required by AI.

Eventually, companies with strict and conservative mindsets must move to more agile, experimental, and adaptable structures, where data-driven decisions are more important than experience-based ones. Creating the right policies to gather the right data while complying with the local laws should be a top priority. This can be a substantial risk and if not well played can create bottlenecks when developing and scaling AI solutions.

  1. Getting your employees engaged

To get everyone on the same boat, the employees must have a common north and a realistic view of the upcoming challenges and changes. A company-wide communication should address the following points:

  • Address the specific obstacles your company has to overcome. That should be started even before introducing AI. Each company has a unique culture, depending on the type of business, their people, structure, etc.
  • Inspire your organization and explain why AI is important. Every time a new technology is introduced, dialogue should be the catalyst for better change as well as a reassurance that the company’s vision is understood by everyone.
  • Address the changes in the day-to-day operations and how that will affect the workforce. New jobs will be created, other jobs will change, and some people will have to be re-trained and adjust to the core. This should not be scary to anyone but appreciated.
  • No change comes without new challenges, and budget is usually the silent killer to a lot of new adoptions. While specific to each company having a realistic view of the return on investment and budget at the time necessary for every cycle should be consulted and well planned.
  1. Organizing to scale

There are some practical steps you can take to start building your business processes and training your staff to support the scaling of AI as well. There is a lot of debate on whether the AI teams should reside within business units or be a central body. Based on experience with customers scaling their AI initiatives, there is a sweet spot, with a balanced hybrid solution:

The long-term and strategic tasks like managing systems, recruiting strategies, hardware, and data governance are recommended to be under the control of a centralized body, regardless of the company’s location and size.

Furthermore, more granular responsibilities such as setting up, designing, and testing are tasks that should start on the main central body, but should gradually be transferred to the business units with the internal growth of the AI adoption.

Activities such as end-user training, and impact tracking, are almost always best owned by smaller teams located inside the business units.

  1. Educating the organization

Getting the latest and fanciest hardware and the best software doesn’t guarantee more work efficiency unless you learn how to use it the right way. To secure the correct adoption of AI, companies need to educate everybody from top leaders down according to their level.

To this end, companies should launch internal workshops that provide decision-makers with a basic understanding of AI and prepare other decision-makers to use AI tools in their day-to-day work. They must also update and sharpen the skills of the analytic teams, so they keep track of the rapid changes and developments in AI technology.

  1. Tracking and facilitating the adoption

Keeping the momentum of adoption is key when introducing new technology to a company. It helps to change management culture, keeps the interest and curiosity high, and gives the right message that the change is positive. Since most AI transformations take from 12 months up to 3 years to complete, leaders must be the engine that keeps the momentum going. Here are some initiatives that have worked with our customers.

Accountability: Creating KPIs for the projects and checking constantly with stakeholders to align with the AI business units and the front-line workers, to keep track of the adoption.

Rewards: Incentivizing, rewarding, and giving credit to the people that have been key for the success of the company’s AI initiatives.

Leadership engagement: Leaders must act as role models and actively encourage new ways of working. AI requires experimentation, and often early iterations don’t work out as planned. When that happens, leaders should highlight what was learned from the pilots and how to incorporate that in moving forward. That will help encourage the team to work and experiment.

Conclusion:

Revolutionary technologies are like a train leaving the station: the sooner you get in, the more time you have to get comfortable and adapt, and those who did not get on board will be left behind and forgotten sooner than they think. AI is for many the most revolutionary technology since the invention of computers, and it is just in its infancy!

Leaders must break away from the classical ways of decision-making. They must start to listen to and follow the data — if your company does not have a data strategy, then this is a wake-up call to act before this train leaves the station.

A comprehensive change of mindset throughout the companies has to happen. Roadmaps and strategies have to be rewritten and adjusted company-wide. Getting all the employees on board is a must and once everyone has a good idea where the train is going, it is time to educate and re-train the workforce from top tier to the newcomers continuously. When the journey is underway,  tools must be in place to track change, measure KPIs, and make the success visible.

The near future will be in the hands of companies in which the human-machine teams outperform humans or machines working independently.

#Cloudflight

[1] Artificial Intelligence Market [ …] Global Forecast to 2021, retrieved August 8th 2021, from https://www.marketsandmarkets.com/Market-Reports/artificial-intelligence-market-74851580.html

Do you have any topics to discuss with our experts?

Get in touch

Get in touch

Menu