Three ways AI enables the entrepreneurship process

Now, more than ever, we have tools at our disposal that help us write, code, and ideate. From an entrepreneurship perspective, large language models (LLMs) like ChatGPT, Claude, and Gemini offer a faster and cheaper route to prototyping. Costly experimentations with product design and application development may be well behind us. For pharmacists, getting an idea to market may require less time and effort than we previously thought. But there is still much to be learned on how to best leverage the new tools at our disposal. Building a pitch deck using ClaudeLLMs are not just search engines. They have a wide range of capabilities that can help entrepreneurs test ideas. Firstly, Claude has demonstrated versatility in building slides. When prompting Claude, you can ask for the LLM to develop a slide with your requirements. The output will be a slide that you can adjust through further prompts or you can download as a PowerPoint file to adjust manually. Think about one-pagers and presentations where you communicate your value proposition. This technique can save you lots of your valuable time. Experimenting with Claude’s “skill” capability is also worthwhile. Users can create and share pre-set instructions that can automate a specific function upon demand. For example, one pharmacist colleague at work recently shared a skill with me that would ask Claude to revamp my slides into the format and style that matches my company brand palette. This can help expedite processes with pre-set instructions. Creating a website using Github CopilotIf you’re a student, you’ll have the option to apply for a free Github Copilot subscription where you can better integrate the power of LLMs with writing code. I personally enjoy using Visual Studio Code with the Github Copilot integration. For me, writing and debugging code, especially with front-end work like website design, is time consuming. This is where LLMs can really shine. The way I like to think about it is that we used to have to write in the computer’s language to ask it to do the things we want. Now, we can use our own language to ask the computer to perform tasks. With my integration, I have the option to prompt copilot to develop my website, building on top of the ideas and materials I have. However, you still need to set up the infrastructure, such as buying a domain name and hosting the website on an appropriate platform. But the work to learn how to code and design may be no longer necessary if you want to quickly bring an idea to the public domain without hiring a developer. Check out my website here.Prototyping software applications using Google GeminiI recently thought about building a meal tracking app on my own to test this idea of prototyping. It turns out that there are a number of integrations with Google Gemini and the Apps Script platform on Google Sheets. Essentially, Gemini told me how to build the app step by step after I prompted it with my idea: “list out the meals I eat when I upload a photo to Google Drive and then promptly delete the photo afterwards.”I was a little skeptical at first, not wanting the LLM to lead me down the wrong path, but after setting up the integration by following Gemini’s instructions and also using it to develop the code to automate the translation of photo to text, it worked like a charm. It wasn’t perfect, there were a few times where it didn’t get the meal exactly right. Sometimes it was more specific than I would have thought, having seen it interpret a chicken dinner as a meal from Medieval Times. It also added functionalities I didn’t exactly ask for like a calories counter. But it was functional, and a way for me to test ideas without hiring a technical team.ConclusionI think now is an incredible time to experiment and play with ideas. Try to write down your fleeting ideas when they come, you won’t always remember and given our efficiencies in execution, valuable insights are still worth their weight in gold. The most important part of this journey right now is to get into the weeds of building things with AI. I gave a talk recently to a pharmaceutical company emphasizing that the technology is moving so quickly that the only way to find out what works for you is diving into things and experimenting. I can’t wait to hear what you will build. Share your ideas and prototypes below. Peter Zhang, PharmD, MBA is a technology professional, hospital pharmacist, and a PhD candidate at the University of Toronto studying entrepreneurship education.