Not too long ago, I set myself on a journey to learn how to create expert profiles, similar to personas, to analyze real-world scenarios: using prebuilt AI products.
In addition to that, I wanted to learn how to analyze simulated real-world scenarios through consolidated data models to forecast risks, identify market gaps, and solve challenges in new ways. I suppose I should add that when I started, I didn't know how to accomplish any of those things.
Diving into the world of data with no prior expertise was no easy task. I'd say it's been the most challenging domain I've learned. I recall spending, roughly, three weeks learning and experimenting: day and night with little to no sleep. Fueled by Subway sandwiches, pizza, and coffee I finally cracked the code on those challenges.
Utilizing prebuilt AI products, here are the five things I've learned and applied in actual business scenarios:
Text Data Analysis: Gather, process, and extract meaningful insights from unstructured text data, such as documents, articles, Zoom chats, PDFs, or social media content.
Profile/Persona Creation: Develop detailed and structured representations (profiles or personas) of individuals or groups based on data and insights, to understand target audiences.
Scenario Analysis: Imagine and assess different situations or scenarios to understand their potential impact on specific individuals, groups, or strategies.
Human-Centric Data Processing: Ensure that data processing methods consider human nuances, behaviors, and thinking patterns to enhance the quality and relevance of analysis results.
AI and Human Thinking Integration: Integrate artificial intelligence with human expertise to analyze complex data, combining the strengths of automation and human understanding in data processing and analysis.
My reasoning for choosing to work with prebuilt AI products is that they are packaged up and ready to use. It’s like walking into a store and buying a product right off the shelf. Not to mention, prebuilt AI products are more accessible which means you don’t have to know coding or hire Data Scientists to help you figure it all out.
I'm sure a lot of people working with data would say that off-the-shelf AI tools are really useful. But they'd also tell you those tools have limitations. They're made for general tasks, and sometimes, you need something more specific. There's truth to that, however, it's not the whole story.
In my opinion, just from the research I’ve done, a lot of people scale to custom-built AI products too soon. In other words, they haven’t maximized the use of prebuilt AI products. Or, they are under the impression they need custom solutions but can’t afford the costs. So they remain stuck in an analog way of doing business. This is where better education on prebuilt AI products is a huge need.
For instance, by using a simple spreadsheet and a basic generative AI tool, I was able to predict market trends and simulate day-to-day business scenarios for a small business. This could not have been done without considerable time spent learning or investing in custom software or scaling to custom-built solutions. This way of doing things is tied to my love of learning and figuring out how to do a lot with a little. It's like being the underdog but still managing to keep up with the competition.
A catalyst in my learning to make products and processes accessible has been my communications fellowship work at Internet Archive’s Open Library. During this fellowship, I learned the importance of starting simple when it comes to using a new tool—especially when it involves collaborative teams. This way, everyone can decide together if something more complex is needed later on.
Now picture this: you're in a meeting with your team or maybe it's just you and another person. You throw out a question about which collaborative project management tool they'd recommend. Some will likely say, Miro, Notion, Asana, Monday.com or maybe Trello. But have you ever wondered why these particular tools often come up?
Most often, people recommend products they know work for themselves. Which is to be expected, after all, would you recommend a tool or solution for your business that doesn’t work? However, the problem with personal recommendations is they often don't consider those who may not work well with those suggestions.
Just by suggesting a fancy tool or app, you might slow down your business if you actually start using it. Can you imagine the amount of money and time that's wasted on tools that don’t represent the needs of everyone? But if you start simple, you can build a strong foundation for your business without unnecessary complexity or expense. That's what I focused on when I was figuring out how to make sense of complex data problems.
My approach to learning to work with AI products and complex processes is about making things easier. I avoided complicated steps such as custom AI setups and complex tools like Python, My Apps Script, and TensorFlow. Though I’m familiar with these components, from an accessibility standpoint, they can be tough for beginners to learn.
Before going any further, I’d like to point out that prebuilt AI solutions aren't all difficult to use; in fact, many are very easy to use. The real challenge lies in discovering innovative ways to leverage these tools to their full potential without the need to start a completely new project or scale up significantly. This of course is where you get into data literacy and accessibility.
From my perspective, there are two general categories of accessibility:
The ability to comprehend how a tech tool or product functions.
The ability to employ tech in multidisciplinary ways to address challenges without scaling until absolutely necessary.
From what I’ve seen, discussions tend to revolve around the first category – comprehension-related accessibility, which remains a significant global need. However, concerning the second category, discussions about the multidisciplinary use of AI tools appear to be rare. I'm sure such discussions are happening somewhere, however, they are scarce at the least.
I can assure you that tech creators aren't ignoring these categories of accessibility on purpose. They're just busy with new ideas. And it would take considerable amounts of time for them to experiment and innovate new ways to utilize their own creations in the simplest ways. And this has opened new opportunities for people like me who enjoy learning and explaining complex information to make it accessible to everyone.
As I'm sure you know, the need for innovators and educators has never been greater. Striking this balance is one way many others can benefit from technology... regardless of their background, location or resources.
Image Credit: Image by rawpixel.com