Ramesh is the CEO and President of ExpertusONE, a leading learning management system that helps companies create engaging training tools.
New technologies are often so brimming with potential that they’re difficult to define. In turn, that makes them harder to implement as part of an overarching digital transformation strategy. Many tech-based implementations fail because they aren’t correctly aligned with business objectives. A practical framework will not only define a technology in terms of its capabilities but map those capabilities to a specific business purpose or use case.
Artificial intelligence (AI) is becoming a baseline technology for businesses. Global AI adoption was 2.5 times higher in 2022 than it was in 2017, but has plateaued in 2023. One possible reason for the boom and subsequent slowdown is the arrival of generative AI and the need to define it and build frameworks for practical implementation. Businesses are used to AI playing a supporting role in their operations, automating tasks to lighten workloads. But when correctly defined and implemented, generative AI has the potential to deliver much greater value, particularly when it comes to all-encompassing business objectives.
According to Gartner, which named generative AI one of the most impactful technologies for 2022, it will account for more than 10% of all data production by 2025. Many businesses are already attempting to come up with practical use cases and implementation frameworks.
In this article, we’ll look at the three fundamental concepts in AI implementation: AI-supported, AI-augmented and AI-powered. First, let’s outline generative AI in more detail.
What is generative AI?
Generative AI is a branch of artificial intelligence that can operate autonomously, utilizing machine learning algorithms, neural networks and large language models (LLMs) to create original creative content. It involves training models to generate new data that resembles human-like output, enabling the production of unique and innovative content without direct human input.
There are countless practical business use cases to be explored that demonstrate the potential of generative AI. For instance, in customer service, generative AI could provide more advanced consumer-facing chatbots to facilitate self-service and respond directly and personally to customer queries.
In healthcare, researchers are using generative AI models to read a protein’s amino acid sequence and accurately predict its target structure, helping to detect the early onset of diseases. In education and training, LLMs can create personalized learning materials and respond to learner queries in a highly contextualized and useful way.
However, while generative AI is paving the way for a variety of revolutionary benefits, there are still challenges associated with the technology that companies must understand before rushing to implementation. In this instance, organizations “running before they can walk” could be a perilous gamble.
This is true whether a business chooses AI-supported, AI-augmented or AI-powered applications, and that decision will largely depend on what it wants to achieve with this technology. Let’s outline each approach and its associated benefits and challenges.
AI-Supported
This refers to the use of AI tools such as natural language processing (NLP) that provide users with information and support to enhance their experience. For instance, voice assistants or chatbots that can listen and respond to questions in a natural, conversational way would play a “supportive” role, alleviating the burden on sales or support teams in a retail business.
The integration of AI-supported capabilities into existing business systems and workflows, however, can be challenging. The compatibility of AI with existing infrastructure, for example, may require substantial modification or even a complete overhaul to accommodate the new technology. Additionally, the process of integrating AI can disrupt established workflows, necessitating a reevaluation and potential redesign of business processes.
AI systems are only as good as the data they operate on. The relevance of the data is paramount, as irrelevant or outdated data can lead to inaccurate predictions and insights. Businesses must ensure they have robust data management practices in place, including regular data updates, data cleansing and data validation processes.
Businesses should also consider employee buy-in. A lack of understanding of generative AI’s benefits can lead to hesitation, low morale, fear or even ethical concerns around its use, adding friction to the implementation process. Business leaders must communicate their intentions with AI through education, training and open “town hall” meetings to demystify it as a technology.
AI-Augmented
AI-augmented, on the other hand, would take the capabilities of AI one step further. Rather than merely playing a supportive role, an augmented approach aims to enhance employees’ capabilities. An example of this is AI in healthcare for diagnostic purposes. AI algorithms can analyze millions of medical images to detect anomalies, but the final diagnosis is still made by a human doctor, thereby “augmenting” their capabilities.
As with AI-supported, there are challenges here too. These AI systems need a large amount of high-quality data to function effectively. In healthcare, this data often comes from medical images, patient records and other sources. Ensuring the quality and consistency of this data can be a significant challenge.
Interoperability is also a concern here, particularly as the stakes in making key decisions are often so high. The healthcare sector tends to be quite disparate in most countries, with siloed resources and disconnected patient journeys. Any AI system would therefore need to be interoperable with multiple systems, including systems for digital health records and prescriptions.
The goal here is to augment and speed up workflows, therefore creating efficiencies, but human intervention is still needed to fact-check and ensure complete accuracy.
AI-Powered
AI-powered systems are those in which AI takes a leading role, such as AI in autonomous vehicles. These vehicles use AI to analyze data from various sensors and make decisions about steering, acceleration and braking. In the management and orchestration of global supply chains, AI predicts demand and places automatic orders to fulfill it. In this instance, the AI is ostensibly in control of the decision-making process.
The speed and scale at which AI can process data is unparalleled, enabling it to make rapid, data-driven, largely accurate decisions. It can also help reduce human bias, as it relies on data and algorithms rather than subjective judgment. Furthermore, AI-powered systems can learn and adapt over time through machine learning, continually improving performance and accuracy.
However, ethical concerns are paramount, particularly in the case of autonomous vehicles where lives are at stake. Determining responsibility for decisions made by AI-powered systems can be complex, particularly when these decisions have significant, potentially costly consequences.
The “black box” nature of AI can also make it difficult to explain how a particular result was achieved, which can hinder trust and acceptance of AI systems. If a system, for instance, develops biases based on the inherent bias of the data it was fed, how can that bias be detected and routed out if the algorithm’s inner workings are hidden or too difficult for a layperson to understand?
Amazon had to abandon an AI recruiting tool because it was found to be biased against women. The tool used machine learning to rate job applicants, but it was trained on resumes submitted over 10 years, most of which came from men, leading it to favor male candidates. Despite attempts to correct the bias, Amazon ultimately disbanded the project.
Conclusion
AI will undoubtedly impact the workforce and job market. Some basic or repetitive jobs may become obsolete as AI technology advances. However, the key lies in reskilling and adapting to the changing landscape. Where some jobs may become obsolete, newer roles will be created, just as with any technological development.
Ideally, the new breed of workers will embrace new technologies and acquire the skills necessary to work alongside AI systems. Just as those who were once writing Java code are now writing AI code to enhance its functionality, individuals who reskill can find new opportunities and contribute to the development of AI technology.
With the right planning, organizational education and employee training, generative AI can play a significant role in supporting, augmenting and powering businesses.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Read the full article here
