How I approach AI strategically in product

ChatGPT and other LLMs are the biggest hype and everyone in your company wants to have feature with it. Now.

But as a product leader, you approach it strategically. Here's how I do it:

1. Identify Existing Problem Areas: Identify the current problem areas within the organization where generative models, particularly LLMs (Language Models), can be applied.
2. Brainstorm with Leadership: Collaborate with the leadership team to brainstorm and explore how generative models can be integrated into existing product priorities to address the identified problem areas.
3. Exclude Common Use Cases: Exclude use cases that are likely to be provided by other companies or existing tools. Focus on unique applications that align with the organization's core competencies.
4. Filter Based on Improvement: Filter out and prioritize only those ideas that have the potential to enhance the organization's core competencies and add value to its products.
5. Feasibility and Complexity Assessment: Have Data Scientists assess the feasibility and complexity of the selected ideas. Assign feasibility/complexity estimates to each idea to understand resource requirements.
6. Run Experiments: Implement a couple of experiments to test the selected ideas and evaluate their effectiveness.
7. Evaluate High-Value Opportunities: Measure the success and impact of the experiments. Determine how much of a low hanging fruit the high-value ideas are, based on the results obtained.
8. Mitigate Hype-driven Decisions: Ensure that decisions are not solely based on hype or trend. Focus on practicality and value creation for the company.
9. Consider Sales and Value Creation: Recognize that integrating hyped technologies into products can positively impact sales. Prioritize projects that align with business goals and create value for the organization.
10. Report and Communicate Results: Regularly report and communicate the findings, progress, and outcomes of the experiments to stakeholders within the organization.