Embracing the Future: An Introduction to AI for Today’s Business

Index

AI revolution has started (and we all noticed that)

In recent years, the Artificial Intelligence (AI) market has shown exponential growth. According to a report by Precedence Researchthe global artificial intelligence (AI) market size was valued at USD 454.12 billion in 2022 and is expected to hit around USD 2,575.16 billion by 2032,progressing with a compound annual growth rate (CAGR) of 19% from 2023 to 2032.

Moreover, Generative AI has demonstrated significant potential in enhancing productivity across various domains, including software development, customer operations, product R&D, sales, and marketing [1]. This aspect of AI is particularly transformative as it can generate new content, automate repetitive tasks, and provide innovative solutions to complex problems. Recognizing its vast potential, many big tech companies are actively competing in the development of Large Language Models (LLMs), which are at the forefront of this technological wave. These models, characterized by their ability to understand and generate human language, are becoming a pivotal factor in the AI race, leading to groundbreaking applications in numerous industries.

In our previous article “Artificial Intelligence & Business: Is your company ready for AI?” we introduced a framework to assess the AI-readiness of a company. Assuming your organization has met these readiness criteria, the next crucial step is to address key questions that will guide your AI implementation journey.

First of all: what do you really need AI for?

Understanding AI: More Than Just a Buzzword

Artificial Intelligence often perceived as a monolithic technology, actually encompasses a variety of concepts and applications. At one end of the spectrum, we have Artificial General Intelligence (AGI), a theoretical form of AI that can understand, learn, and apply its intelligence broadly and universally, akin to human cognitive abilities. On the other hand, narrow AI focuses on specific tasks and operates under a limited pre-defined range. Narrow AI is prevalent today but the rocketing development of multimodality models has revamped the debate on AGI.

But how to understand if your company really needs AI?

The foremost question to address is whether your requirement calls for a deterministic answer, which is precise and definitive, or a probabilistic one, where outcomes are based on likelihoods and predictions. Minimizing a route when planning delivery to your clients or optimizing the scheduling in order to reduce unvailabilities, for instance, are tasks that can be completed with optimization algorithms, especially when variations are limited and can follow pre-defined rules/constraints.

The second question is whether a statistical approach could bring to satisfying results. When the amount of variables is limited and relations are mostly linear, statistical analysis could be easy but still effective approaches. This may happen for time series analysis in demand forecasting or quality controls. Conversely, when faced with an expanding array of variables and their progressively non-linear interactions, the indispensability of AI becomes apparent. AI surpasses the limitations of both the human mind and traditional computational methods in discerning intricate correlations and insights from complex, intertwined data sets, offering unparalleled efficiency and depth in data analysis.

The latter question is related to the type of task to perform. This question is pivotal in discerning the appropriate AI application. AI can be categorized into two functional types: predictive and generative. Predictive AI analyzes historical data to make predictions about future events, widely seen in applications like demand forecasting or clients’ retention prediction in retail. Generative AI, however, takes this a step further by creating new content, like text, video, images, exemplified by tools like Chat-GPT or DALL-E that generate images from textual descriptions.

Ensure Human In The Loop

When we come to decision-making processes, it is apparent that problems vary significantly in their level of criticality. Machine-driven or enhanced decision-making can be mapped onto 3 levels [2]:

  • Higher-level decision support: Decisions are primarily made by humans, “based on principles and ethics, experience and bias, logic and reasoning, emotion, skills and style”
  • Augmented machine support: Machines an AI “generate recommendations, provide diagnostic analytics for human validation and exploration”
  • Highly automated settings: There is still a need for “guard rails or a human-in-the-loop for exceptional cases

Explainable AI or XAI is a branch within artificial intelligence that aims to bridge the gap between the opacity of complex machine learning models and the need for transparency and interpretability especially for higher level decision support algorithms. Traditional AI models, such as deep neural networks, have consistent performances but they are not transparent enough to let the decision makers understand the factors that influenced the outputs (“black box”). . .

According to a survey run by IBM in 2022, 4 out of 5 companies declared they could not apply AI on a large scale as they were not able to explain how their AI arrived at a decision. Among the barriers to AI adoption, we find the need to avoid unintentional biases and to control the deterioration of model performance (drifting), the need to better understand the importance of input data and to explain why the model returns a prediction.

As Intellico, we are exploring techniques in Explainable AI, such as Graph Neural Networks (GNNs) in predictive AI and Knowledge Graphs in generative AI, that play a pivotal role in enhancing decision-makers’ understanding. GNNs are a type of neural network designed to directly operate on the graph structure, making them particularly adept at capturing the relationships and interconnections within data, which is invaluable for predictive analysis. On the other hand, Knowledge Graphs represent a network of interconnected data and concepts, providing a dynamic and holistic framework for AI systems to generate new content and ideas. These approaches not only clarify the reasoning behind AI-driven predictions and content generation but also significantly bolster trust in the reliability and accuracy of these models.

Conclusions

AI transcends the realm of futuristic speculation; it has become a pragmatic tool that businesses are actively harnessing for a competitive edge. Simultaneously, comprehending its multifaceted nature is crucial for implementing a systematic approach towards its integration. At Intellico, we are supporting our clients in grasping the underlying principles of AI in specific envisioning sessions. This deep understanding is essential to masterfully harnessing the technology’s full potential, concentrating at the same time efforts towards the highest value use cases.

[1] The economic potential of generative AI: The next productivity frontier, McKinsey, 2023

[2] AI Isn’t Ready to Make Unsupervised Decisions, HBR 2022

[3] Global AI Adoption Index 2022, IBM 2022

Do you need more information?

Fill out the dedicated form to be contacted by one of our experts.