To better understand when to use generative AI, let’s use another baking analogy created by Maha Bali.
If you need a cake for an event, you have some options. You might:
None of these options for cake are inherently good or bad. Instead, each option comes with its own pros and cons and will be appropriate in different situations.
If you want a cake for a close friend’s birthday, spending hours of extra preparation making a cake from scratch would add a personal touch and give you full command of the cake’s taste and texture.
If you’re bringing a cake to a casual social gathering, however, since the cake merely serves to bring people together, buying a decent quality cake from the bakery would save you hours of unnecessary work.
Similarly, you should adapt your use of generative AI and LLMs when writing to the specific needs and context or your project.
For instance, relying on generative AI to write a complete research paper based off a single prompt would be as effective as bringing Twinkies to a baking competition.
Working with generative AI to craft a working outline for a research paper, however, could save time and provide a solid foundation, much like baking a cake using a boxed cake mix.