What do you do if your AI work receives negative or critical feedback?

AI projects should be focused on visible business value. Solutions should be built focused on user experience and not the ML/AI tech. Getting user feedback into the design and delivery of your solution is essential. Sometimes sponsors wont agree with the focus of your projects, compromise and build quick wins which support their vision, while securing investment for the more impactful projects you can see.

In the world of business, everything comes back to ROI. So, if your AI projects aren't working, ie they are producing negative ROI, then you should expect negative feedback. Then, you should think about how you can make them better.

When I receive negative or critical feedback on my AI work, I take a step back and remind myself that feedback is a gateway to growth, not a setback. First, I carefully review the comments to understand the concerns and identify any recurring themes. I then reach out for a detailed discussion, if possible, to clarify points and gain deeper insights. This often reveals areas for improvement I might have overlooked. Next, I prioritize the feedback, focusing on changes that align with the project's goals and my personal development. Finally, I apply the feedback, documenting the process and outcomes to track my growth and share learnings with my team. Embracing feedback with an open mind has led me to refine my work and enhance my skills.

Stay objective when receiving negative or critical feedback on your AI work. Avoid taking it personally and recognize it as an opportunity for improvement. Assess the feedback impartially and consider its validity in addressing potential weaknesses in your AI solution.

Avalie objetivamente o feedback, identifique áreas de melhoria e adapte seu trabalho conforme necessário. Utilize-o como uma oportunidade de aprendizado e crescimento, buscando sempre aprimorar suas habilidades e contribuições no campo da IA.

I meticulously dissected the feedback, identifying specific areas needing improvement—be it the data set, the algorithm's effectiveness, or the model's performance. This detailed analysis helped me understand the root causes and devise a strategic plan for addressing them.

Take the time to carefully examine the negative or critical feedback. Understand the specific concerns or issues raised by the feedback provider. Look for patterns or common themes to identify areas that require improvement.

Carefully analyze the feedback to understand its core concerns. Is it addressing issues with data bias, model accuracy, ethical considerations, or perhaps the practical applicability of your AI solution? Identifying the specific areas of concern can help you determine the most effective course of action for addressing them.

Negative feedback is maybe the most valuable data for your project. If you can capture this information and work with whoever it originated from to understand the exact situation that resulted in the negative experience, you can then implement changes that will 100% improve your system.

I reached out to the feedback providers for a deeper conversation. This engagement helped clarify their concerns and provided me with additional insights. It was a step towards understanding the gap between my project's intentions and the stakeholders' expectations.

Reach out to the feedback provider, if possible, to gain a deeper understanding of their perspective. Ask for clarification or additional details to ensure you fully grasp their concerns. Engaging in a constructive dialogue can provide valuable insights and help build a better relationship with the feedback provider.

If possible, engage in a constructive dialogue with the feedback providers. Ask for clarifications to ensure you fully understand their perspectives and concerns. This conversation can provide deeper insights into the feedback and help uncover additional context that wasn’t initially apparent. Furthermore, it demonstrates your commitment to excellence and openness to growth.

Based on the analysis of the feedback, develop a plan to address the identified issues. Prioritize the improvements based on their impact and feasibility. Consider involving relevant stakeholders or experts to ensure a comprehensive approach.

Based on the feedback and any additional insights gained through dialogue, develop a plan for improving your AI work. This might involve: Revisiting and refining the data set to eliminate biases. Adjusting model parameters or experimenting with different algorithms to enhance performance. Implementing additional ethical safeguards or transparency measures. Collaborating with stakeholders or experts to address specific concerns. Document your plan and share it with relevant parties to show that their feedback has been taken seriously and to outline the steps you're taking to improve.

Execute the planned improvements by making the necessary modifications to the AI work. Ensure proper testing and validation to verify that the changes effectively address the feedback. Transparently communicate the changes to relevant parties, such as users or clients.

After implementing the changes, take the time to reflect on the entire feedback process. Assess the effectiveness of the improvements and evaluate how well the feedback was addressed. Identify lessons learned and incorporate them into future AI work to avoid similar issues.

Currently, people often see AI as the solution for anything. However, in a digital transformation it’s key to figure out the origin of the faced challenges and find the best solution from a people, process and technology perspective. Sometimes the solution is some improvement in ERP data quality, some connectors or tailoring of a bot… and not AI. Besides, it’s about expectation management. When it comes to technology we often see that that people expect perfection. As AI is a probability based technology, perfection (ie 100% probability) is tough to reach 😉 So, expectation management is a key element

Evaluate and debate the reasons for that kind of feedback and then move forward with a plan to accommodate the outcome if necessary.

It is crucial to establish a feedback loop with users or clients to proactively gather feedback on an ongoing basis. Implementing mechanisms, such as surveys or user testing, can help identify potential issues before they become significant problems. Additionally, staying up to date with the latest AI research and industry best practices can enhance the quality of your work and minimize negative feedback.