Episode #198: Speed Up Your Toy or Game Invention Process With AI - Part 1
Listen Now
Tap Play Below or Listen On Apple Podcasts | Spotify | Google Podcasts
Have you ever wondered how Artificial Intelligence (AI) is revolutionizing the toy industry? From speeding up the toy invention process to comparing traditional and AI-assisted methods, AI is transforming how we create and innovate. Dive into this engaging exploration of AI’s role in the toy industry.
In this episode, join The Toy Coach as she shares insights from her talk at the People of Play Innovation Conference. Discover how AI can enhance market research, ideation, product visualization, and even prepare you for your next big pitch. Get ready to be amazed by practical AI applications that are reshaping the toy development landscape.
EPISODE CLIFF NOTES
Introduction to AI in Toy Industry [00:01:00]
AI’s Impact on Market Research [00:07:00]
Ideation and AI: Generating New Toy Concepts [00:15:32]
Visualizing Toy Designs with AI [00:22:27]
Preparing for Pitches Using AI [00:26:21]
AI in Email Communication [00:28:24]
The Future of AI in Toy Development [00:29:00]
-
This episode is brought to you by Publisher Services, Inc. (https://www.pubserviceinc.com)
Listen and Learn more about AI in the toy industry with The Toy Coach (https://www.thetoycoach.com/ai)
-
EP 198
[00:00:00] Azhelle Wade: You are listening to Making It in the Toy Industry, episode number 198.
[00:00:17] Azhelle Wade: Hey there, toy people, Ajele Wade here, and welcome back to another episode of the Toy Coach podcast, Making It In The Toy Industry. This week's podcast episode is sponsored by Publisher Services, Inc. PSI is a game and toy sales operations company helping manufacturers place their products in the hobby, specialty, and mass market channels.
[00:00:39] Send a message to info at pub service, inc. com or visit pub service, inc. com slash contact dash us to get your market ready game or toy placed in hobby specialty and mass market channels. Be sure to let them know you heard about PSI from this podcast.
[00:00:57] Recently, I was asked to speak about [00:01:00] AI's impact on the toy industry at the People of Play Innovation Conference.
[00:01:05] And it was a phenomenal experience. I put together a piece that gave a wide overview of a select few AI tools out there that can improve your toy or game development process. There were moments of my talk where people laughed, gasped, and were blown away by the power of Now, if you've been following this podcast for some time, you already know that this is not my first time talking about AI here, but it does seem that the industry is more ready to listen than they were when I first released my four podcast episodes about AI back starting in January of 2023, you can hear all of those episodes by visiting the toy coach.
[00:01:44] com forward slash AI. By the end of this episode, you will learn how AI can be used to speed up your toy invention process, all the while comparing the traditional methods of toy creation to aI assisted methods of toy creation. Here is the order of how we'll do that. First, we're going to go [00:02:00] over the basics of AI, my warning for using it, as well as the definitions you need to know to be able to fully understand this episode. Then we're going to get into how AI can be used to improve your toy development process, starting with market and research trends, then moving on to how to use AI
[00:02:16] to assist in ideation, then how to use it to visualize ideas. And finally, we will talk about how you can use AI to help you prepare for your pitch, whether you're pitching your idea for licensing or your product for retail placement. Then I'm going to highlight other options for AI tools that have more legal protection and blow your mind when I make it clear how you're likely already using AI in your everyday life.
[00:02:41] Don't worry. I'll tell you where and how, so you can stop if you want to. At the very end, we'll have a mini masterclass on prompting because knowing how to write a good AI prompt is paramount to making the best use of the tool.
[00:02:55] I know that's a lot, so if you haven't already guessed, this will be a two part episode.[00:03:00]
[00:03:00] But we can do this. Let's dive in. Now, I can't do an AI episode without first providing a disclaimer. The laws around using AI generated content to make a profit are unclear. So most AI tools run on an open system, which means that they learn from the information you put in, in order to generate the output that it gives.
[00:03:19] Everybody else. So any confidential or sensitive information should be kept out of your AI prompt generation. However, there is one AI tool, Adobe Firefly, that was intentionally built with a closed system. This was done so you can own what you generate, but we'll get into that in more detail in the end of this episode because own is kind of a loosely used term.
[00:03:43] Next, let's talk definitions really quick. I want to make sure that you're understanding what I've said so far and what I'm going to be saying moving forward. So first. Inputs. What are inputs? Inputs are the text or the command that you will give an AI tool. Prompts. [00:04:00] Prompts are the type of input that you give an AI tool that's typically text. It can also be images. AI tools are the branded programs like ChatGPT and also refers to the plugins that might be added to things you already use like Gmail subject line suggestion or sentence completion tools. AI models are the AI platforms with different source code that makes it all happen.
[00:04:26] The term training refers to giving an AI model information or examples to train it on how to output information for you. OpenAI is the company that is responsible for ChatGPT. And the last term I want you to know is OpenAI.
[00:04:41] It's the company that is responsible for ChachiPT, which is likely the AI you hear the most about. Okay, you've got all the basics. So let's dive into our first area of the toy development process that we could put together. potentially enhance with AI and that is market research. So the beginning of the [00:05:00] toy development process starts with market research.
[00:05:02] Typically we start that research because we've been inspired by what we saw at a toy trade show. Maybe we saw a cool trend, or we've been inspired by what we've heard or learned at industry events. Maybe someone forwarded us a viral TikTok or there was a sales conversation with our retail partner or our boss came up to us in our office and said, Hey, There is a new category we need to hit.
[00:05:24] Let's do it. What triggers the market research portion of the toy development process varies, but once triggered, we all dive into trade magazines, research reports, surveys, forums, and it and it all leads us back to data that's stored online. The process of market research can be short you're rushing to market and have a small budget or a small team, or extend to six months of searching, finding supported articles, data reports, interviews, and general insights to support and guide the direction of whatever you're developing in a particular toy [00:06:00] category.
[00:06:01] And guess what? Finding that data is one of the easier steps in the process. It gets harder when you've got to compile it, analyze it, pull out key phrases or data points, and pull all of that information into a presentation deck that summarizes the findings while still keeping your findings organized enough to cite your original sources.
[00:06:21] But it doesn't stop there, because if you're doing the market research right, once you've compiled all that data, you'll probably have new physical places that you want to go and visit to further your research into the category or trend.
[00:06:35] You might identify new toy stores to visit, trade shows to attend, or realize that you need to identify new retail partners to really assure that this new toy or game product will be a success. This process takes months and months to complete, but with AI, you could shorten it to days or even minutes. So there are a lot of AI tools [00:07:00] out there, but not all AI tools are created equal.
[00:07:03] ChatGPT is an AI tool that has a ton of data, but that data is, well, dated. The information that ChatGPT is giving you is coming from an old library So, in the case where you need current and fact checkable data, I recommend Google BARD. Here's why out of all the other AIs, Google BARD should be your go to for AI supported market research. This tool can instantly scour the web for qualitative and quantitative data. Data. Analyze it and contrast that data.
[00:07:37] Summarize consumer reviews, build a comparison chart of competitor products and fact check the accuracy of it all. So let's imagine I was tasked to come up with the next big plush toy line. How would I start that research? Well, I can ask Google BARD to analyze the size of the global plush market to see [00:08:00] if this is a viable endeavor.
[00:08:02] And when I did that as a test, BARD responded this, quote, according to persistent market research, the global plush toy market size was valued at 3,500 million in 2022 and is expected to expand at a CAGR of 4.5% during the forecast period, reaching 4,557 million by 2028,
[00:08:26] And reading that I thought, well, that's impressive, but is it even accurate? But because BARD is linked to Google, there is an integrated fact checker. Cool. To activate that, you've got to click on this G Google logo to check your response.
[00:08:42] And so after some time thinking, the AI tool highlighted everything that it could source from online articles in green and everything that couldn't be sourced in yellow. And clicking on anything highlighted in green would instantly take me to that source so I could review it and deem if it was credible or [00:09:00] not.
[00:09:00] Now that I could dive into that article and know, quote unquote, that a plush category is a viable category, I need to know who the players are. So now I'm going to ask Google Bard this. Can you analyze sales data of plush products in the U. S. to identify the best selling and worst selling products?
[00:09:20] And it gave me a list, with number one being Squishmallows. And it continued down to five various plush products. Now at this point you might be thinking, well, Ajelle, even with all this incredible research, I still have to compile that data, and it's going to take me hours. Well, no, you don't. Because instead of spending my time creating a table and compiling that data, I just asked Bard to do it.
[00:09:44] So here's what you would prompt Bard.
[00:09:46] Create a chart of the top five best selling plush, including their price, as well as the summarized customer reviews about those products. And. It did it. It gave a summary of customer [00:10:00] reviews, the price, a link to buy, and the product for all of the five competitor products. Now the exercises that I did with BARD are just the tip of the iceberg.
[00:10:09] Again, since BARD is an AI tool owned by Google, You can direct it to analyze any of the information that it can find online as well as any Google products. So if you have Google Docs or Excel sheets, you can have BARD analyze that data as well. If you are uncertain about the data that it's pulling from online, which honestly you should be, you have to double check everything, you can use it in a different way.
[00:10:36] Let's say you just paid 5, 000 for a fancy market research report filled with data that you don't have time to go through and summarize and report on. Well, Bard could do that for you. So the benefit here is we can spend more time managing the data that we have and pulling out the important bits rather than spending so much of our time collecting and optimizing it.
[00:10:58] We can get a second [00:11:00] set of eyes on identifying opportunity in complex reports and Bard can suggest different ways that we can present that information. Google Bard is a fantastic resource for market research, and as more and more companies look to rank in Google in the age of AI, I believe it will only become more and more reliable. And here's why. When Google rolls out their new AI optimized search engine, they're going to highlight sites that feature unique statistical data and any other unique research, qualitative and quantitative data.
[00:11:37] The reason is they want to be able to generate trusted answers to questions that people are typing in search and they wanna be able to use AI to do that. Over time, I think we'll see that. The reports that we're getting from AI generated prompts will become more accurate as more reputable sites and companies like NPD try [00:12:00] to rank for the opportunity to be featured first in Google from AI generated results.
[00:12:05] If you are a member of Toy Creators Academy, you know that one of the tools I'd like to use in market research is Amazon. You know the types of reviews that we like to look for. There is a new addition to Amazon reviews that I've got to tell everybody about. If you've been on Amazon recently and you've looked at a product and you've gone down to the review section, you might've seen a section that says this.
[00:12:29] Customers say, and it's an AI generated summary of customer reviews. Amazon is using AI to distill thousands of reviews into a paragraph that tells you what people like. The most and what people like the least. So instead of scrolling through looking at different five star reviews and two star reviews and three star reviews and trying to find the holes and opportunities, Amazon is summarizing that for you.
[00:12:59] Since [00:13:00] we're on the Squishmallows kick, I am on a Squishmallow official holiday. I am, when I'm looking at a Squishmallow official holiday Kelly toy plush, super cute, looks like a mushroom and you can scroll down to the AI generation summary of customer reviews and you'll see it says customers like the softness and appearance of the toy figure.
[00:13:23] They mentioned that it's very cute and very soft. Others, however, are not happy with the size and value of the product. They say that it is very small and overpriced. Underneath that, they even highlight with green check marks, what's a positive point of the review in single words. So we've got two check marks for appearance and softness on this product.
[00:13:45] And then we've got a kind of like a line through orange circle for the negative comments, which is size and value. So this is. Awesome. Amazon is literally showing you where [00:14:00] the market opportunities and gaps are in a summary.
[00:14:03] Now the process of typical idea generation might start with a global trend research trip led by market research findings. It might lead to buying samples of competitor products to open, play with, and experience. You might follow that up with a deep dive into YouTube for unboxing videos, get the customer experience, and ultimately culminate in a series of brainstorm and ideation sessions between a team of designers, product developers, and marketing pros at a toy company.
[00:14:31] That session could result in written concepts of ideas, sketches and marching orders for the design team to move forward with and expand upon. Usually brainstorm sessions start out a bit slow due to the blank page phenomenon. There are randomized tactics that I've used in the past when holding interdepartmental creative brainstorming sessions at Toys R Us, where I would have people randomly select words and categories and [00:15:00] brands to ideate on to be that first spark of inspiration.
[00:15:04] Because we all know once ideas start rolling with creative people in the room, usually they don't stop. The question here is how can AI support your company's idea generation? Well, enter the most well known tool, ChatGPT. The hardest part of starting anything is that blank page, right? Text based AI generators are great for helping generate ideas when you're stuck on that blank page.
[00:15:32] The best text based AI for Idea generation, in my opinion, is ChachiBT. You can train ChachiBT to do what I did when I sat down in a conference room and selected randomized categories, brands, and features from a bowl to have a team of designers ideate around them. But you can do it on a much larger scale and a heck of a lot faster.
[00:15:53] You can tell it to consider your market research data points, include randomized [00:16:00] variables, and generate concepts around that in over a dozen toy and game categories at the same time in a matter of seconds. Let's start small. To practice, you can ask ChachiBT to just come up with 10 new ideas for toys that have never been made before.
[00:16:18] After doing that. Let me tell you what chat GBT came up with. It gave me a full list with names of the products and a one line description of each. So we had the quantum puzzle box, a puzzle box that incorporates elements of quantum physics, teaching kids about super position and entanglement whilst.
[00:16:35] Solving Puzzles to Unlock Hidden Compartments to Augmented Reality Nature Explorer, a handheld device that combines augmented reality with real world exploration, allowing kids to interact with virtual creatures and plants as they explore. Nature. Three. Kinetic Sand Sculpture Studio. A kit that includes a special type of kinetic sand and tools for sculpting [00:17:00] intricate and detailed sand art, providing a satisfying and creative tactile experience.
[00:17:05] Four. Holographic Storytelling Globe. A globe that projects holographic stories and historical events related to different countries and regions, making geography and history lessons engaging and interactive. Five. Music Making Robot building kit, a customizable kit that lets kids build their own music making robots, teaching them coding, engineering, and music composition in a fun and educational way.
[00:17:32] It went on and on.
[00:17:34] Now, imagine starting a brainstorm with these rough and sometimes wild ideas. Instead of starting with a blank paper, imagine how much faster you can get to innovative when you're already starting from something over nothing. Okay, now let's imagine that you've done some research in the game industry and your research actually showed you that physical games and puzzles for six players are what's trending [00:18:00] next.
[00:18:00] You could do something like this. You could put this into ChatGPT. You could say, Come up with five tabletop puzzles. Board game ideas for six players that are puzzles to tease your brain. When I did that, ChachiBT came out with this.
[00:18:16] Circuit Conundrum In this game, players take on the role of electrical engineers, tasked with completing circuits to power various devices. The game board is a grid representing a network of wires and components. Players must strategically place components, resistors, capacitors, switches, etc. on the board to complete circuits and connect devices while overcoming challenges like Voltage restrictions and limited resources.
[00:18:43] There is a brain teasing element in this game where circuit conundrum challenges players, logical thinking and problem solving skills as they optimize circuits to power devices efficiently.
[00:18:55] If you like that idea, but you feel like it needs more developing, or it [00:19:00] needs a toyetic element in it, you could type in an additional prompt to have ChatGPT reference the idea that it just came up with and build on it. So I typed in this. Expand on circuit conundrum. This game needs a physical toy piece built into the game board that players interact with. This could be where the players complete the circuits.
[00:19:21] What are some ideas for the game piece?
[00:19:23] ChachuBT came up with the Circuit Builder Module. The Circuit Builder Module is a physical toy piece integrated into the game board, it says, and it consists of the following components. It says it will consist of the circuit grid. The main part of the module is a grid with connectors that represent nodes in an electrical circuit.
[00:19:43] These connectors are embedded with magnets.
[00:19:45] 2. Component tiles, resistors, capacitors, switches, LEDs, are all included with the game, each featuring magnetic elements on the back that allow them to securely attach to the circuit grid. 3. Indicator lights. [00:20:00] The Circuit Builder module includes LED indicator lights that can be placed on the grid to visualize the flow of electricity.
[00:20:08] Now we know toy people that we are not going to put magnets in this game, too expensive, too dangerous, not worth it. But chat GPT did come up with a great base for a game concept.
[00:20:20] If you want to see the full gameplay that it outputted just go over to thetoycoach. com forward slash AI, and that is where you're going to just be able to see the full from what I generated. If you're wondering what the benefit of this is, I mean, the benefit here is speed.
[00:20:37] We all know, it takes a hundred bad ideas to yield one good one, but that takes time yielding those 100 bad ideas with AI. We can fully think through a hundred ideas. In seconds, we can save our time and energy developing good seeds instead of bad ones. Let's say we're beyond the ideation stage, right?
[00:20:58] You've thought through your [00:21:00] concept, but you are still struggling to visualize what this product will look like, right now in the typical toy development process, you would either illustrate the idea yourself in Adobe Photoshop, Illustrator, Procreate, maybe your hand sketching still, you might even go to modeling the concept in 3d.
[00:21:18] If you're not skilled in any of those programs, you might hire someone or work with another team member from your company in another department. And the goal would be to execute a visualization of the design without being able to visualize it. You really can't move forward. This phase is important because whatever design you land on is going to be the foundation of what is mass produced and brought to market.
[00:21:42] Sometimes, if you're working at a bigger toy company, that first illustration or rendering is either going to get this idea to go to the next stage Or it's going to end the idea right there. Ideally in the visualization phase, you want to try out different looks for the product, but sometimes it's rushed [00:22:00] because we're rushing to get a product out the door.
[00:22:02] Illustrating just can take so much time. So we just execute the design that we know will get approved by our bosses or sell into our retailers. But what about AI supported idea visualization? Typically, I use Mid Journey to generate images for anything, but now ChatGPT, can do that too.
[00:22:27] ChachiBT recently partnered with DALI to make image generation possible within their tool. While Mid Journey will give you four images for every one prompt, ChatGPT's image generator will only give you one. But I'm telling you, it's probably going to be the only one you'll need.
[00:22:46] For this next example, I referenced the game concept that ChatGPT generated, called Circuit Conundrum, and asked it to create an image of this board game, and here's how I prompted it. Create an image of this [00:23:00] board game called Circuit Conundrum on a pristine white background. Here is the game explained.
[00:23:06] And then, I just copied in the game rules that ChatGPT generated itself. And hit enter. The image that it rendered looked like a photograph of an actual game. And a pretty well designed one at that. The game looked to be in mid play, with various decks with intriguing symbols, translucent play pieces, and it matched the game rules.
[00:23:31] Again, to see this rendering, head over to thetoycoach. com forward slash AI. And let me tell you just how you can generate images with chat GBT because it did take me a minute to figure it out. So you might find this valuable. first up, you're going to need a paid account. It's not something you can do in their free account.
[00:23:50] And in the top right, tap chat GPT 4 to reveal a dropdown and choose DALI. Ask chat GPT to create an image and be sure to add as much [00:24:00] description as possible, hit enter, and you will see a note that DALI is generating images. If you don't like the results or you are curious about others, you can hit regenerate.
[00:24:10] Now what's interesting to me is DALI was one of the worst image generators in my opinion, or I just struggled to use it. It was never really part of my rotation, but now with its partnership with ChachiBT, I think it took both platforms to a whole new level. Now, as I said earlier in this podcast, my go to image generator for a long time was Midjourney and I do still love Midjourney.
[00:24:33] The images that it creates seem to default to this like fantastical and maximalist feeling. So for inspiration, for creative imagery, I do find Midjourney will render quite inspiring imagery of your product ideas. It'll add in intricate design details that you just might not have thought of or you just wouldn't have had time to go there with it.
[00:24:57] So if you want to learn how to [00:25:00] generate images with MidJourney, I will walk you through it quickly right now.
[00:25:03] Just head over to midjourney.com, get a free account, join the Discord channel. Get onto one of the newbie chats, type backwards slash imagine. Click on the prompt and then type a description of the Toyer game you want to render. Hit enter and watch your image come to life if you want that video training.
[00:25:22] You should definitely take the Toy and Game Idea Bootcamp. I will teach you how to render your toy ideas with chat GPT in a step-by-step video and how to create the right prompt for it and it's really affordable. Go to learn dot the toy coach.com/bootcamp the benefit here. With idea generation supported by AI is visualization.
[00:25:44] Some people need to see an idea on paper before they buy into it. And instead of spending hours illustrating or prototyping your idea, you can render it in seconds. I mean, you can explore various design styles before settling on the [00:26:00] ideal one for your product.
[00:26:01] I mean, it's, it's, it's really just an incredible innovation that's so useful. Now, what if you're at the toy development stage where your concept is already done? Got a visualization of it and maybe you even have that prototype made, and you're getting ready to go to a sales meeting or you're going to a pitch event or trade show.
[00:26:21] Can AI help you get ready for that? Well, heck yes, it can. Now, normally if you're preparing for a sales meeting, you would open up a PowerPoint template. You'd start filling in your product information, dropping in imagery, taglines, all that wholesale prices you might spend an hour finding the perfect wording to explain the function of the toy for the most exciting headline Or for the most exciting headline for a slide, but not anymore, my friends, enter beautiful AI.
[00:26:50] So beautiful AI allows you to use text to actually jumpstart a presentation. You can literally prompt it with logos of the [00:27:00] top nine toy companies in the world. And then current nine toy companies in order of how much revenue they generated in 2023. And it will list. that information in graph and presentation form.
[00:27:13] It will design slides for you. Wild. It is not the only one out there that does this. It's just one I've played around with. But. You, there are AI generators out there that will create PowerPoint presentations for you. Now, what if you are not pitching to a retailer?
[00:27:31] Maybe you're pitching an idea to an inventor relations rep. If you're an inventor. Well, in that scenario, you are likely preparing a sizzle video, my friend. And again, that blank page phenomenon can leave you frozen for hours, but you can have ChachiPT design a script based on a game that it came up with, or you can train the tool on your own game. You can even train it on the way that you want it to write a sizzle video. Maybe you have a formula that you already know works for you.
[00:27:58] Now, if neither of [00:28:00] those situations are a good fit for you, or you're like, that's not useful for me, Azhelle, well, I've got one for you no matter how you're pitching your toy or game idea.
[00:28:07] No matter what job you have in the toy industry, you are likely using email and you might be using email to network and follow up with new business opportunities. And that is where one of my favorite tools, WordTune, comes in handy.
[00:28:24] WordTune is a tool that will rewrite your copy to make it easier to understand. So if you're struggling with your elevator pitch, or the opening line of an email, or struggling to write a whole email, WordTuning is it. Now a pro tip is to jump back and forth between ChatGPT drafting an email for you and WordTune to help you analyze different ways to say the same thing.
[00:28:48] Visit app. wordtune. com to try it out. And this. is where we're stopping for today. Yep. It's a two part episode. So we're going to pause here. And if you're interested, [00:29:00] please, please, please join me back next week to pick up where we left off. I will be holding a presentation online about AI, teaching some of the things I talked about in this episode, but doing it visually so you can follow along and we can do the process together.
[00:29:18] So if you're interested in that, head over to thetoycoach. com forward slash AI.
[00:29:22] To temporarily close this conversation until next week, I'd like to give a quick shout out to Mary Cousin for inviting me to do the AI talk at ChiTag.
[00:29:32] Really appreciate that. It was a great talk. There's also now an article on the People of Play website that I posted about AI. Thank you for all your support, Mary.
[00:29:41] As always, thank you so much for spending this time with me today. I know your time is valuable and that there are a ton of podcasts out there, so it truly means the world to me that you tune into this one.
[00:29:53] Until next week, I'll see you later, toy people. [00:30:00]
-
🎓Learn more about how you can develop and pitch your toy idea with Toy Creators Academy® by clicking here to visit toycreatorsacademy.com and join the waitlist.
Not ready for the Toy Creators Academy online course? Start by connecting with fellow toy creators inside our online community. Click here to join.