“The rise of powerful AI will be either the best or the worst thing ever to happen to humanity. We do not yet know which.” – Stephen Hawking.
Yes, indeed. Whether it is better or worse will ultimately be decided by the goals we choose for using AI. This aspect has an ethical, practical, and strategic connotation as well. So, the first choice we need to make is whether we use AI for rightful outcomes or for something grey that can be harmful to humanity. From here on, we move to other parts of this question. And the answer would depend on who is asking that question. Who is the stakeholder here?
Artificial intelligence (AI) benefits and stakeholders
Let us consider what the research firm ESI ThoughtLab recently published as a benchmarking study (cosponsored by Deloitte). It unravelled how companies approach their AI implementations, what value they seek, and what they achieve. Companies are seeing a positive ROI from their AI implementations. The following were the top areas for returns.
- Customer service and experience (74 percent)
- IT operations and infrastructure (69 percent)
- Planning and decision-making (66 percent)
So, if the stakeholder is a CEO, the goal would be different. But for a regulator or a customer, the goal would change a bit. Similarly, even in an enterprise, the goal would change if we talked about a developer, functional manager, or operations person.
Companies are gaining value from their AI implementations. But it is spread across five major areas—higher productivity, increased customer satisfaction and retention, improved employee engagement, improved profitability, and new products and services. Finally, some expect value from new AI-powered products and services, and their number has changed from 19 percent to 42 percent. So there is also a marked trend from efficiency-focused benefits to strategic ones.
AI goals would change or dilute or accentuate as per the stakeholder, their AI maturity, expectations from AI, and their constraints. That brings us to the next important question.
What are artificial intelligence (AI) goals?
AI is a subset of computer science that helps in designing intelligent computer systems. These systems simulate human traits like language, learning, reasoning, solving problems. They cover knowledge representation, learning, rule systems, search, etc. The AI goals are different for psychologists, philosophers, and cognitive scientists. They are called scientific goals. But for companies, users and developers – they are put in a separate bucket – here they are called Applied goals. So AI can also be divided into Strong AI and Applied AI.
Strong AI entails the ability to reason, solve puzzles, make judgments, plan, learn, and communicate. In comparison, Applied AI covers advanced information processing and creates applications that an enterprise can use to augment its outcomes. This can be done using Natural Language Generation, Chatbots, Speech or Image Recognition, Sentiment Analysis.
The challenge with defining artificial intelligence (AI) goals
AI is complex. Whether we see its use, techniques, inner working, or its intended applications. The goals are challenging to define and measure, and that is why an organization needs to have a good understanding of what it is looking for and how it will achieve that. It is better to express AI goals as well-posed questions and hypotheses around a specific and intended benefit or outcome. Of course, this would change as per a particular stakeholder. In the ‘State of AI in the Enterprise’ report by Deloitte, it has been recommended that businesses should move beyond efficiency and use AI technologies to differentiate themselves.
Why and how to set good artificial intelligence (AI) goals?
Deloitte has advised that the good idea is to find out the enterprise’s AI maturity and then decide the roadmap. After that, even frameworks like SMART – with a solid process of discovery, exploration, and experimentation – can be put to use.
In the Deloitte report – Seasoned (26 percent) enterprises were seen setting the pace of AI adoption maturity. They have undertaken many AI production deployments and have developed a high level of AI expertise across the board—this has manifested in selecting AI technologies and suppliers, identifying use cases, building and managing AI solutions, integrating AI into their IT environment, and business processes, and hiring and managing AI technical staff. Compare that to the other category, and we see that Skilled (47 percent) have launched multiple AI production systems but are not yet as AI-mature as the Seasoned organizations. As a result, they lag on their number of AI implementations, their level of AI expertise, or both. So interestingly, seasoned AI adopters are investing more with 68 percent having spent more than US$20 million over the past year. But what is worth noting is that they also typically achieve payback on their investments in a shorter time. Almost 81 percent were seen reporting their payback period is less than two years. In addition, about 26 percent of all respondents said that AI technologies enable them to establish a significant lead over their competitors. But for Seasoned adopters, this level had risen to 45 percent.
For some enterprises, the goal is that AI will substantially transform both their organization and industry within the next three years. And for some, the outcome is about an early mover advantage. For example, about 57 percent expect AI would transform their organization in the next three years, and 38 percent feel that the industry would change in the same time frame.
As we can see in the report, all adopters are using AI to improve efficiency; mature adopters harness the technologies to boost differentiation. Some are using AI for automation and optimization, which can provide significant benefits.
But the overall AI’s “early adopter” phase is ending. As per IDC forecasts, spending on AI technologies will grow to US$97.9 billion in 2023. So whether an enterprise is an early adopter or a mature one, a seasoned AI user or a skilled one, the outcome paradigm would vary. This would dovetail with where the AI is going – for the final dartboard. Efficiency, augmentation, automation, revenues, or a disruption altogether.
Ask some tough questions. See whether you are using AI as a business goal or from some other stakeholder’s perspective.
Define before you step forth. That will make a massive difference to your AI outcome.