When setting expectations for AI pilot projects, how do you balance the potential capabilities of AI technologies with the limitations and constraints that may exist?
Sort By:
Oldest
IT Manager in Constructiona month ago
The adoption of AI requires a new mindset. Unlike traditional file location and search methods, AI requires precise prompt engineering. Users must know how to ask the right questions and be able to assess the quality of the results they receive. Moreover, the nature of AI services, often provided as Software as a Service means that updates and changes can occur frequently. This requires ongoing learning and adaptation by users to keep up with the evolving technology.Director of IT in Educationa month ago
It's crucial to choose the right business partners who understand the capabilities and limitations of the AI being implemented. We work with partners who are as informed and at the cutting edge as we are, which eases the process. However, AI is advancing rapidly, and we often find ourselves trying to catch up with new capabilities that suddenly become available, which can be both exciting and challenging.Director of Systems Operations in Healthcare and Biotecha month ago
The rollout of pilot projects involves setting clear expectations that the technology will not be perfect and encouraging feedback. However, a major challenge arises when users compare the new AI tools to more established products like ChatGPT. They see what other technologies can do and expect the same from the new tools, which creates a continuous battle to manage expectations and explain the limitations and reasons behind certain technological decisions.An important aspect we had to address was training users on how to properly interact with the AI. This involves teaching them how to effectively prompt the model to ensure they can fully leverage its capabilities.
Global Digital Workplace Sr. Director, Global Technology & Security in Healthcare and Biotecha month ago
In rolling out AI technologies, such as the copilot example, we faced numerous limitations. Initially, a significant challenge was the lack of skilled resources. Everyone was on a learning path, which meant that my team had to manage their daily jobs alongside learning new technologies. Furthermore, compliance issues played a major role. Before we could even begin using the AI tools, we spent months dealing with legal assessments, data privacy, and cybersecurity risk assessments. These are substantial limitations, not to mention the cost, which is another critical factor.Another significant learning from our pilot was the importance of data quality. If the input data is of poor quality, the results will be affected regardless of the AI technology used. This was a major limitation for us, especially in terms of information classification, which impacted the effectiveness of the AI solutions.