top of page
Writer's pictureKate vanderVoort
Aug 84 min read

Dr. Evan Shellshear: Why 80% of AI Projects Fail and How Not To



In this week's episode of The AI Grapple, I sat down with Dr. Evan Shellshear, a leading expert in AI. His journey, starting from a childhood love for maths to a global career in AI, offers valuable insights for businesses of all sizes.


From Childhood Curiosity to Global Expertise

Evan's passion for maths began in grade school, sparked by a simple maths sheet. That early interest led him through an impressive academic journey, including time at the University of Queensland and the Nobel Prize-winning Institute for Mathematical Economics in Germany. His extensive experience across multiple industries and countries highlights the diverse applications and transformative power of AI in business.


Understanding AI, Automation, and Robotics

Evan emphasised the importance of distinguishing between AI, automation, and robotics. They're often conflated, but serve different purposes. AI is about using algorithms to predict outcomes based on data. Automation involves executing pre-defined tasks without human intervention, while robotics focuses on building machines to perform physical tasks. Understanding these differences is crucial for businesses to leverage each technology effectively and align them with their strategic goals.


Overcoming Misconceptions in AI Projects

A lot of our discussion focused on common misconceptions and challenges businesses face when implementing AI projects. Evan mentioned that over 80% of AI projects fail, often due to reasons unrelated to technology itself. Poor use cases, lack of leadership buy-in, inadequate data quality, and misalignment with business strategy are common culprits. His co-authored book, "Why Data Science Projects Fail," delves deeper into these issues, providing a roadmap for avoiding common pitfalls.


Building AI Capabilities: Start Small and Scale

For businesses new to AI, Evan recommends starting small and building capabilities incrementally. This approach involves initially focusing on becoming data-driven, creating dashboards, and visualising data to understand the business better. By starting with manageable projects and scaling gradually, organisations can build a solid foundation and avoid costly failures. This incremental approach ensures that AI initiatives are aligned with business strategy and supported by robust data infrastructure.


Success Stories and Practical Applications

Evan shared inspiring success stories, including his work on a platform that automates programming for robotic welding systems used by major automotive manufacturers. This project significantly reduced production time, showcasing AI's potential to revolutionise industries. For smaller businesses, Evan recommends leveraging off-the-shelf AI tools to gain efficiencies without the high costs associated with custom-built solutions.


The Future of AI: Opportunities and Caution

Looking ahead, Evan expressed cautious optimism about AI's future. AI technologies will become more integrated into everyday tools and processes. However, there's also a risk of backlash and over-investment leading to a potential bubble. Businesses should focus on practical, value-driven applications of AI rather than getting caught up in the hype. Setting clear guidelines and policies for AI use will allow organisations to harness its power responsibly and effectively.


Leveraging AI Tools: Practical Recommendations

One of the most valuable parts of our conversation was Evan's practical recommendations for AI tools. For beginners, he surprisingly suggested Excel or Google Sheets. These tools can serve as an accessible sandbox for playing with data, testing ideas, and developing initial models. For more advanced users, Jupyter Notebooks and Python offer powerful capabilities for building and visualising more complex AI models. This step-by-step approach allows businesses to grow their AI capabilities organically and sustainably.


The Role of Education and Culture in AI Adoption

Evan highlighted the importance of education and cultural change in successful AI adoption. It's not just about having the right tools or data; it's about fostering a culture of curiosity and continuous learning. By encouraging employees to experiment with AI and providing the necessary training and support, businesses can demystify AI and integrate it more seamlessly into their operations. This cultural shift is essential for overcoming resistance and ensuring that AI initiatives are embraced at all levels of the organisation.


Ethical Considerations and Legal Implications

As AI becomes more prevalent, ethical considerations and legal implications are increasingly important. We discussed the need for robust AI policies that address data privacy, bias, and accountability. Organisations must ensure that their use of AI aligns with ethical standards and regulatory requirements. This includes setting guidelines for data usage, implementing human-in-the-loop processes to maintain oversight, and staying informed about emerging legal frameworks related to AI.


Looking Ahead: The Next Frontier of AI

As we look to the future, Evan's insights provide a roadmap for navigating the changing AI landscape. While there are challenges and uncertainties, there are also tremendous opportunities for businesses that approach AI thoughtfully and strategically. By focusing on practical applications, fostering a culture of learning, and addressing ethical considerations, organisations can harness AI to drive growth, innovation, and competitive advantage.


Our conversation with Evan highlights the transformative potential of AI when implemented with care and foresight. As businesses embark on their AI journeys, they should remember that success is not about quick wins but about building a solid foundation and continuously evolving. Whether you're a small business just starting out or a large enterprise looking to scale, the principles of strategic alignment, data-driven decision-making, and ethical responsibility are key to leveraging AI effectively.


 

Links:

bottom of page