What hampers companies the most when implementing advanced technological solutions?
Many companies are typically reluctant to adopt new technologies, mainly due to the inherent risk. Companies try a technology carefully at first. When things do not go the expected way, decision-makers prefer to wait. Moreover, in the case of AI, there is often no strategy for implementation. But in addition to the state of mind, there is a practical problem, which is capacity. There is a lack of people who can deliver and manage innovation. That is why we help companies identify opportunities and achieve their first goals in a very short time. Along with the first results comes the confidence to implement more comprehensive solutions that will transform the business and keep it competitive.
On the other hand, we see a few innovative companies successfully implementing AI/ML solutions. These companies are typically far ahead of the competition, and they are better protected against disruption.
In which fields has artificial intelligence currently the most visible impact?
Artificial intelligence and machine learning are big topics, and they are still at their beginning of impact. Companies are already aware of their potential, but they are far from exploiting it. AI has wide possibilities in the automotive and manufacturing industries. However, AI also offers huge added value and competitive advantage to companies from the banking/fintech segment, retail, and e-commerce. In this case, AI helps to provide customers new types of services at acceptable prices. Algorithms and technologies today control stocks, sales, and so on.
AI is a popular topic and area. Which technologies do we neglect?
We underestimate the Internet of Things, 3D print in production, or the business benefits of virtual and augmented reality, which could be used much more. At the same time, augmented and virtual reality will soon become a revolution, especially in personal life. People will start using them in everyday lives, and companies must also prepare for that.
There is much more talk about the technological readiness of companies in connection with coronavirus. But was the pressure to digitize really affected by the pandemic?
Today´s companies need to bring new products to market quickly and optimize the supply and production chain to avoid unnecessary costs. Therefore, the pressure on digital transformation was visible even before coronavirus – the pressure stems from customers’ ever-increasing needs and the dynamics of the business itself. In areas such as banking, e-commerce, or retail, for example, customers are putting significant pressure on the quality and availability of services. The pandemic only intensified all this and revealed the (un)readiness of companies. Companies that previously implemented solutions such as automated production planning and logistics scheduling or online communication channels for customer support can now keep up. They are also able to keep lower operational costs and their customers.
What challenges will companies face in the next five to ten years?
Companies mainly need to grasp and adopt new technologies and turn them into their business and processes in the longer run. However, this strategy has often been disrupted by a pandemic. Companies are now working mainly with a time frame of a year or two. They think about how not to stifle planned innovations and at the same time how to save costs that increase with the crisis. At Blindspot, we are helping them through our solutions that can provide returns and savings quickly and open up other ways to maintain business growth.
What do you perceive as an AI challenge? Is there anything around the AI field that you would like to change?
I would like to decrease the fear and uncertainty that some companies have when it comes to implementing AI in their business. There are many examples of companies across the segments where the implementation of AI was successful and helpful. It is also essential to think about the transformation itself, which is why at Blindspot, we focus on solution delivery and a design of a sustainable framework of AI adoption that company can follow.
What can a successful implementation mean?
For example, in logistics and supply chain planning, artificial intelligence can cause a real revolution. Manually filled-out spreadsheets in Excel and lengthy human-centered processes can be replaced by a centralized semi-autonomous planning system that can reduce days-long activities to minutes. Another significant help is possible assistance and notification about a potential problem in advance. For example, workers will know in time that the capacity of trucks is already reaching its maximum, and the assistant will recommend optimization in time.
AI is explained in many different ways, from clever algorithm-based solutions to humanoid robots that could rule the world. What can we imagine under the one that can help companies with their businesses?
Artificial intelligence includes, for example, many areas of machine learning, including statistical, relational, deep learning, or neural networks. In this case, an algorithm stands above the data and learns a certain model. We all know these models. Thanks to continuous learning, the algorithms can personalize a retail shopping platform and help with product recommendations. Artificial intelligence – specifically machine learning – simply finds new relationships and rules within customer's data.
On the other hand, AI is limited by the data provided. Of course, by the quality of data and the subset of the problem the data describe. For example, if customers' decisions depend on current weather and weather data are missing in my dataset, AI cannot recommend me to extend my dataset.
How many such rules are there in the model?
There can be tens of thousands of complicated (but also unnecessarily duplicated) rules with slight nuances. The learning algorithm creates a structure defined by these rules. But even here Pareto principle applies, i.e., typically, 80% of algorithm's decisions are based on 20% of the rules. The essence of our mentioned shopping models is basically compact and explainable.
However, even compact models are almost impossible to create and maintain by people. In the case of retail shopping platforms and banking or insurance products, it is an increase in the customer experience through personalization that one would never be able to handle manually in such quantities.
Another interesting topic is B2B personalization, where we model the behavior of suppliers in the supply chain: what and when to order so that the company has everything that needs on time.
What other area of AI makes life easier for companies and their customers?
Intelligent assistants, for example, in the form of smartphones and smart devices around us. Assistance as such makes sense in decision-making in everyday life and business. For example, assistance in logistics can point to a constant increase in the capacity of trucks. The implementation of such technologies needs to be considered comprehensively both in B2C and B2B.
What is the right approach when considering the implementation of AI?
There is a well-proven innovation framework where a company develops innovative solutions from ideas through a well-controlled and well-defined process. At Blindspot, we also help companies during AI open days. Together, we generate a large number of use cases based on discussions with decision-makers and company employees. We deliver a pre-validated list of use cases to the management, which typically directly supports the implementation of the top ideas. So, a company can start implementing AI quickly.