Executives' guide to investing in artificial intelligence: Understanding the value proposition in 2025

As AI investment trends continue to evolve in 2025, executives face growing pressure to make strategic decisions about implementing artificial intelligence. With companies rapidly adopting AI tools - 22% implementing them extensively and 33% using them in a limited way - understanding how to evaluate and implement AI solutions has become essential for maintaining competitive advantage. In the book "The Executive Guide to Artificial Intelligence" by Andrew Burgess, the author provided a comprehensive guide for business executives who want to understand and implement artificial intelligence solutions in their organizations.

This book was published in 2017 by Springer International Publishing and provides a practical overview of how companies can leverage artificial intelligence. What has changed today?

Current investment trends in AI 2025

The AI landscape is experiencing unprecedented growth, with organizations making more significant investments to remain competitive.

The basics:

Burgess emphasized the importance of starting by defining clear goals aligned with business strategy, a principle that remains valid today. In the book, he identified eight core AI capabilities:

1. Image recognition
2. Speech recognition
3. Search and information extraction
4. Clustering
5. Natural Language Understanding
6. Optimization
7. Prediction
8. Understanding (today)

Evolution from 2018 to 2025:

Since the book was written, AI has gone from an emerging technology to a mainstream technology. The "Understanding" capability that Burgess considered futuristic has seen significant advances with the advent of Large Language Models (LLM) and generative AI technologies, which had not yet emerged in 2018.

strategic framework for AI investment decisions

The four essential questions

When evaluating investments in AI, it is critical to focus on these critical questions:

1. Definition of the business problem
2. Success metrics
3. Implementation requirements
4. Risk assessment

Note: This four-question framework comes from current knowledge and is not explicitly presented in Burgess's book.

Building an effective AI strategy

The adoption framework:

Burgess proposes a detailed framework for creating an AI strategy that includes:

1. Alignment with business strategy - Understanding how AI can support existing business objectives
2. Understanding AI ambitions - Defining whether you want to:
3. • Improve existing processes
   • Transform business functions
   • Create new services/products
3. AI maturity assessment - Determining the organization's current maturity level on a scale of 0 to 5:
4. • Manual processing (Level 0)
   • Traditional IT automation (Level 1)
   • Isolated basic automation (Level 2)
   • Tactical implementation of automation tools (Level 3)
   • Tactical implementation of various automation technologies (Level 4)
   • Strategic end-to-end automation (Level 5)
4. Creating an AI heat map - Identifying areas with the greatest opportunities
5. Developing the business case - Evaluating "hard" and "soft" benefits
6. Change management - Planning how the organization will adapt
7. Developing an AI roadmap - Creating a medium-to-long-term plan

Evolution from 2018 to 2025:

Burgess' framework remains surprisingly relevant today, but needs to be supplemented with considerations of:

• AI ethics and regulations (such as the EU AI Act)
• Environmental sustainability of AI
• Responsible AI strategies
• Integration with emerging technologies such as quantum computing

Measuring ROI in AI investments

The determinants of return on investment:

Burgess identifies different types of AI benefits, categorized as "hard" and "soft."

Hard benefits:

• Cost reduction
• Cost avoidance
• Customer satisfaction
• Compliance
• Risk mitigation
• Loss mitigation
• Revenue loss mitigation
• Revenue generation

Soft benefits:

• Cultural change
• Competitive advantage
• Halo effect
• Enabling other benefits
• Enabling digital transformation

The measurement of AI ROI has become more sophisticated, with specific frameworks for assessing the impact of generative AI, which did not exist when Burgess wrote the book.

Technical approaches to AI implementation

Types of solutions:

Burgess presented three main approaches to implementing AI:

1. Off-the-shelf AI software - Pre-packaged solutions
2. AI platforms - Provided by large tech companies
3. Custom AI development - Tailor-made solutions

For the first steps, he suggested considering:

• Proof of Concept (PoC)
• Prototypes
• Minimum Viable Product (MVP)
• Riskiest Assumption Test (RAT)
• Pilot

What has changed:

Since 2018, we have witnessed:

• democratization of AI tools with no-code/low-code solutions
• Dramatic improvement of AI cloud platforms
• Growth of generative AI and models such as GPT, DALL-E, etc.
• Rise of AutoML solutions that automate parts of the data science process

Risk considerations and challenges

The risks of artificial intelligence:

Burgess devoted an entire chapter to the risks of AI, highlighting:

1. Data quality
2. Lack of transparency - The "black box" nature of algorithms
3. Unintentional bias
4. AI naiveté - Limits of contextual understanding
5. Excessive dependence on AI
6. Choosing the wrong technology
7. Malicious acts

Evolution from 2018 to 2025:

Since the book was written:

• Concerns about algorithmic bias have become a critical issue (pending further exploration)
• AI security has become essential as threats increase
• AI regulation has emerged as a key factor
• The risks of deepfakes and generative AI disinformation have become significant
• Privacy concerns have grown with the more pervasive use of AI

Creating an effective IA organization

From the book by Burgess (2018):

Burgess proposed:

• Build an AI ecosystem with vendors and partners
• Establish a Center of Excellence (CoE) with dedicated teams
• Consider roles such as Chief Data Officer (CDO) or Chief Automation Officer (CAO)

Evolution from 2018 to 2025:

Since then:

• The role of Chief AI Officer (CAIO) has become common
• AI is now often embedded across the whole organization instead of being confined to a CoE
• The democratization of AI has led to more distributed operating models
• The importance of AI literacy for all employees has emerged

Conclusion

From the book by Burgess (2018):

Burgess concluded with the importance of:

• Don't buy into the hype but focus on real business problems
• Start the AI journey as soon as possible
• Future-proof the company through an understanding of AI
• Adopt a balanced approach between optimism and realism

Evolution from 2018 to 2025:

Burgess' call to "don't believe the hype" remains incredibly relevant in 2025, especially with the excessive hype surrounding generative AI. However, the speed of AI adoption has become even more critical, and companies that have not yet begun their AI journey now find themselves at a significant disadvantage compared to those that followed Burgess' advice to start early (in 2018!).

The AI landscape in 2025 is more complex, more mature, and more integrated into business strategy than could have been predicted in 2018, but the core principles of strategic alignment, value creation, and risk management that Burgess outlined remain surprisingly valid.