AI as a Sparring Partner: Lessons from Tara Bosenick’s Visit

We recently had a really insightful visit to Bangalore from Tara Bosenick. As a Managing Director at our parent company, ReSight Global, and a Managing Partner at UIntent in Germany, Tara brought a valuable perspective. It was a great week of exchanging ideas, especially seeing different cultural approaches and ways of thinking come together. This synergy really highlighted the core of ReSight Global – our commitment to learning from each other and growing both individually and as a collective of employee-run businesses.

Following Tara’s visit, we also held a very interesting workshop focused on the practical applications of Artificial Intelligence in our daily work. Thinking about how seamlessly calculators became part of our routines years ago, we’re now exploring how AI can become a similar everyday tool. Tara, who has significant experience experimenting with AI, shared her insights, the tools she’s tested, and some objective ways we can evaluate new ones. 

Here are three key takeaways I have from understanding the potential of these AI tools for our work:

• The AI is only as good as the prompter behind it (Garbage In, Garbage Out): Generative AI is a great technology only if you know what you’re doing with it. It could mean, and quite literally so, “garbage in, garbage out”. As we go beyond the Ghibli trend, and the number of r’s in strawberry, we learn that the real value is in pointing it in the right direction. We know that Generative AI technology in layman terms, looks at ‘’ that are similar and puts them together, which also means it hallucinates easily, creates new information, or throws out things which are not in the input. The only way you can know accuracy is if you know your craft well, so you know what to expect and what’s a good outcome. AI is a great sparring partner and a side-kick to have to help, but it is not the expert by itself, you are.
  
• Test-test-test, and keep track of the evolution: AI tools and their responses are constantly changing. What works today might yield different results tomorrow. It’s crucial to rigorously test different tools and prompts, keeping a “dated” log of your experiments and observations to understand how each tool evolves over time. 
• Become better at your craft, and ‘craft’ a way to explain it (Articulation is Key): The deeper your understanding of your work, the better you can leverage AI. Now, the skill isn’t just in doing but in clearly articulating your processes and knowledge to the AI. Turning that tacit “muscle memory” into explicit frameworks for the AI to understand is the critical skill to cultivate.