Artificial Intelligence (AI) is here to stay and is growing. And, we believe it needs to be complemented with creativity and design thinking. We don't think of AI algorithms as the last word in every situation. They may provide better and faster outputs than humans at certain tasks, but we need to ensure algorithms are designed to augment and assist humans, not to rely on them as they upend our experiential knowledge.
Today, the global AI market is estimated to be worth about $28 billion and is expected to grow at a CAGR of 33% to $200 billion by 2026, according to Fortune Business Insights. However, this does not necessarily mean replacement of human intelligence by AI, as it has been framed in many arguments globally.
It means more and more collaboration between humans and machines to bring together their intelligence to solve complex business problems. This is already happening across industries with new business use cases to be added going forward, as we are in the ‘Age of With’ as we call it at Deloitte -- humans working with machines to continue fostering innovation for businesses and individuals.
Businesses have already started leveraging this to reap benefits of combining human and machine intelligence.
As per a Deloitte global AI study, more than 8 in 10 leaders see AI as being very important to their business already, and being critical to their business within the next two years. Globally AI spend and adoption is surging, with close to 37% of organisations having adopted AI (a whopping 270 percent increase over 4 years!). Even in developing markets like India, it is estimated as per IDC that by 2025 at least 50% of enterprise application releases will have embedded AI functionality.
Through significant advances in research cross-leveraging AI and neuroscience, businesses have been able to benefit from AI applications mimicking certain functions of the human brain. Already we see that the most successful AI algorithms are inspired by the human brain.
For instance, computational models of AI are actually inspired by the human neural network. Other notable research in AI includes attempts to recreate human cognitive skills such as attention, inference, imagination and planning. The same holds true in AI models and robots that mimic the sensory and motor skills of the brain for business applications.
For example, computer vision as a technology being used with other AI Models to implement AI-powered robotic surgeons or have robots to man hazardous parts of chemical factories. Also we have home devices such as Alexa, Google Assistant, etc. which leverage speech analytics (NLG) to answer questions, or to have speech powered chatbots which can give conversational reports. Latest advances in this area have even helped incorporate Emotional Quotient (EQ) to AI’s Intelligence Quotient (IQ) to help correlate events to human emotions to get better responses from AI solutions to human needs, as per Deloitte Tech Trends report.
Technological advancements have further helped the increase in AI adoption.
For example, the challenge AI automation had in being inherently input data intensive and not very relevant in situations of data paucity. This has led to development of technology tools which address this issue. For instance, data scientists have used Matrix Completion techniques which are extremely effective in building AI models with sparse data.
Also, another main catalyst for this increased adoption of AI by businesses has been leveraging the power of human brain, that is, business domain knowledge, intuitive analysis, etc. in AI feature design, model variables selection, etc., to deliver effective AI solutions which deliver actually meaningful outcomes for business users taking into consideration domain specific aspects and incorporating necessary aspects of human intuition.
The future for this trend of combining human and artificial intelligence also looks promising. As per Deloitte Tech Trends report, CIOs across the globe have consistently rated AI / Cognitive technologies among the top applications they plan to invest in for their business. We believe that this trend will in the near future lead to most businesses becoming AI-fuelled organisations, wherein humans and machines work together to build sophisticated AI/ machine learning models incorporating such a level of human decision making ability that they will provide autonomous intelligence, that is, the AI models can even implement business decisions on their own basis the model outcomes, as good as a human could do (a corollary being moving from assisted intelligence, that is, a car warning a driver of changing lanes, to autonomous intelligence, that is, a driverless car which can drive itself on its own). Given further are some examples of AI applications helping businesses to leverage a mix of the powers of the human brain and AI.
To give an example of how a mix of ‘experiential’ human knowledge of our and client’s team members when combined with the power of AI helped one of our clients, the client was faced with an issue of fluctuating power prices making the production cost to be uncertain. The business objective was to plan production in such a way so that power consumption cost is minimised. The domain knowledge of team members helped identify the list of factors that could affect power demand and past patterns based on experience which impacted one another. For example, weather patterns, changes in the economy, demand from other users, etc., and the same were used to build AI models to predict power consumption costs6. This helped the client to better plan production work and save costs.
Considering another case of combining design thinking, augmented with machine intelligence -- the synergy between human centred design thinking and AI leads to swifter movement from ‘empathizing to prototyping’ and accelerates AI adoption.
One of our banking clients was unable to build an AI solution because of lack of data and more so, usable data. The client wanted to develop a customer segmentation framework but with less demographic information and limited transaction history the segmentation was ineffective. Design thinking framework allowed us to think end to end and realise that through a simple change in the application form we can ensure more relevant data is available for analysis. The clients was also encouraged to use alternate data sources and enrich internal data ensuring right quality and quantity of data is available for AI to build segmentation models. This in turn helped create ‘segment of one’ as the most appropriate product/service offering for an end customer, arrived at using a specialized recommender system. Thus the role of AI is that of a force multiplier in implementing business strategy as arrived through design thinking.
In summary, we have arrived in the ‘Age of With’, where humans and AI are working together seamlessly - centred around human creativity and experience, but driven by the power of artificial intelligence to synergistically achieve more than was possible by either separately.
Prashanth Kaddi is partner at Deloitte India. The views in this article are his own.