For those just joining us, every day during this challenge, I’m going to try and do something different or better, using currently available Artificial Intelligence tools.
Conceptualizing a product isn’t for the faint of heart. It involves an intricate journey that starts with a creative spark, an idea that has the potential to solve a problem or fulfill a need. It is a delicate dance of brainstorming, research, and collaboration, where (often) multiple minds come together to create something new and exciting. To gauge consumer response, it’s crucial to test the waters by gathering feedback through surveys, focus groups, or even social media platforms. This vital step helps refine the concept and ensures the final product resonates with the target audience. Bringing a vision to reality can be both exhilarating and challenging; it involves coordinating design, engineering, and manufacturing efforts to deliver a product that not only meets expectations but also stands the test of time.
Of course, any hacks that allow you to cut down the the time, effort and/or money involved in the process are most welcome. Using generative AI tools can be an interesting way to speed up the ideation and design visualization process, so that the concept can be tested with the target audience (either through traditional research or even via a crowdfunding effort to gauge interest).
Today’s Generative AI challenge is going to focus on specific steps in the process, i.e. ideation and ‘photo shoots’ (or product showcases) for conceptual products. This is a little different to the Day 8 challenge, where I put imagination to the test by creating potential brand/product collabs using tools like Midjourney.
Step 1 – The Product Idea
Let’s say we have a rough sense that we would like to create a unique new style of bluetooth speakers. The first step would be to see what’s out there in terms of competition and popularity, to figure out what works and how to create something really differentiated. For this, the best place to start is popular eCommerce and review sites, for example Amazon:
The challenge is that popularity ratings on mainstream sites may not always serve audience-specific needs. For example, if we want our speakers to appeal to tech geeks, the list may look quite different. In the screenshot above, you’ll see I used the Edge browser and asked Bing Chat (powered by GPT4) about which speakers were most popular with tech geeks, and why. This resulted in a few great items, which start providing clues about what’s most important in terms of tech specs, form factor and other features.
In the above step, I go further by asking Bing Chat about what the minimum specs would need to be to appeal to my audience…I also call out that my niche is really into robots. Looking up the robot-themed recommendations, I get an idea about the designs that appeal to my niche:
I then went straight to ChatGPT to continue my dialogue. Feeding the bot examples of what’s popular with my audience (including basic design premise), I asked for a few more design ideas that may appeal to them:
I settled on the second idea for a “Robotic Arm Speaker”
Step 2 – Visualizing the Product
The next step was to actually visualize our Robotic Arm Speaker. For this, I used ChatGPT to create various sample prompts to feed into Midjourney, with some tinkering and tweaking till I had a few interesting outputs…
…and one chosen design.
Step 3 – Creating A Product Showcase
As this is meant to be a product someone would use in everyday settings, it would be helpful to have a few different photos showing it in use. The challenge with tools like Midjourney, Stable Diffusion and Dall-E is that unfortunately they don’t really allow you to stick with specific objects and change only their environment. New products like Adobe Firefly will probably make that a lot easier…or if you’re a seasoned designer, you can take the photo above and transplant it into various settings yourself.
Given this challenge is all about using simple AI tools, I did a quick search and found a great new site called Mokker, that allows you to swap out image backgrounds quite easily using AI.
Here’s the tool in action – first, I uploaded the speaker image into the tool:
Mokker isolates the desired object (in this case our speaker) and places it in a few random settings, with the ability to then select custom backgrounds. These can be common backgrounds already available on the tool or you can specify your own generative prompts. Given this is a speaker, I chose ‘office desk’ and ‘wooden table’ to generate a few new options:
If you like any of the options, you can then download standard or high definition images for use as needed. In this case, I downloaded a couple of the shots above for use in my product portfolio.
With a bit more time and tweaking, you can get some pretty great results.
Here are a few more BONUS robot-inspired speakers I created (separate to the robotic arm idea:)
There you have it…a conceptual product showcase without the traditional investment of time, high cost and hassle. A great way to test new ideas with your target audience, before you embark on the more intensive and expensive journey to prototype, refine, manufacture and eventually get the product to market.
0 comments on “Day 28 of 30 Day AI Challenge – Showcasing Conceptual Products Without Physical Prototypes or Expensive Designers”