The GenAI Hype Cycle
The initial euphoria surrounding generative AI led to a flurry of projects in 2023, many of which failed to deliver expected returns. This has prompted a reevaluation of how to measure ROI for genAI deployments and determine where this technology is most effective.
Key Challenges:
- Top-down pressure: Boards and CEOs pushed for rapid AI adoption without proper planning
- Inflated expectations: Early successes in small pilots didn’t translate to large-scale deployments
- Underestimated costs: Data preparation, infrastructure, and ongoing expenses were often overlooked
- Hallucination issues: The need for human verification reduced productivity gains
Rethinking AI ROI
Experts suggest a new approach to evaluating genAI projects:
1. Focus on experimentation and learning rather than immediate financial returns
2. Create AI committees to vet and approve projects across the organization
3. Prioritize smaller, more controllable objectives over flashy large-scale deployments
4. Use genAI as a tool for educated guesses rather than a source of absolute truth
The Path Forward
As the hype settles, IT leaders must protect their organizations by focusing on genAI deployments that bring true value. This involves better training of AI systems, setting realistic expectations, and understanding the technology’s limitations. By adopting a more measured approach, enterprises can harness the potential of generative AI while avoiding costly missteps.
