publications

Or a hypermath moment while chatting with my newest AI BFF, #Perplexity.  

My 6 AM mornings have adopted a new ritual of cold-pressed coffee, Perplexity, and ChatGPT. A Bose speaker rendering AI voiceovers ideally informs current subject matter for my daily writing. And because, IMHO, all hypermath writers need ChatGPT to organize thoughtful copy, while Perplexity helps with web scraping. Because even well-intentioned random subject matter requires domain knowledge.  

This morning, I hit a wall.

Owning domain knowledge defines your expertise or understanding of specific areas or fields, such as healthcare, finance, law, engineering, or creative fields, such as art and music. Picture yourself as an oracle who knows the ins and outs, the nitty-gritty details, and the big-picture concepts of a particular subject. Just like a doctor knows all about medicine or a financial analyst understands the stock market, having deep domain knowledge means you’re an expert in your chosen field.

My wall came when asking about a subject matter that was not yet published on the web. Or something non-existent as far as AI is concerned.

Now, when it comes to working with all things AI generative. Chatting with fancy machines that can create art and music or even write stories independently requires domain knowledge.  This is the time for your knowledge and understanding, with the ability to ask clear, concise prompts and queries.  

Why do you ask?

1. Subject matter expertise: Imagine you’re trying to teach an AI about medicine. Well, if you’re a doctor or Biologist, you’ll know exactly what medical knowledge needs to be learned. This deep expertise helps you inform the AI in the right direction, resulting in accurate information. [1] [3]
2. Creative thinking: Even though AI can be creative, it still needs a human touch to ensure its creations are top-notch. If you’re good at thinking outside the box and coming up with cool ideas, you can help the AI make some truly awesome stuff. Generative AI models like Midjourney are becoming incredible at creating art with one word. [2]
3. Understanding data requirements: You know how they say garbage in, garbage out? Well, it’s true for AI, too. If you want it to learn correctly, you’ve got to feed it the correct data. As a domain expert, you know exactly what kind of information the AI needs to get smarter. [3]
4. Prompt engineering: This one’s like giving the AI a roadmap. If you’re an expert in your field, you can create clear and specific instructions for the AI to follow. That way, it knows precisely what you want it to do, which saves time. [3]  
5. Interpreting AI outputs: Sure, the AI might spit out some cool stuff, but does it make sense in the real world? That’s where your domain expertise comes in handy. You can look at what the AI creates and tell if it’s useful or just a bunch of gibberish. [1] [3]
6. Ethical considerations: AI isn’t perfect, and sometimes it does things that aren’t quite right. As a domain expert, you can spot these ethical issues and ensure the AI behaves. [1]
7. Specialized tools/software: If you already use fancy tools and software in your field, why not teach AI to use them too? With your expertise, you can help integrate AI into your workflow and make everything run smoother. It’s a needful thing when you’re a hypermath. [4]

AI is cool on its own, but it becomes a fantastic co-agent. It’s even better when you mix in some human domain knowledge. Together, they can do some beautiful things while ensuring they’re done safely, ethically, and with a real impact.

Citations:

[1] https://www.linkedin.com/advice/3/what-most-important-generative-ai-skills-z2a3c

[2] https://hbr.org/2023/11/how-generative-ai-will-transform-knowledge-work

[3] https://www.nature.com/articles/s41746-022-00634-5

[4] https://www.linkedin.com/pulse/generative-ai-adapation-specialized-domains-srinivas-kuppa

[5] https://addepto.com/blog/domain-specific-generative-ai-top-5-examples-of-how-to-use-it-for-your-business/