What is a Design-thinking based workshop for AI Creation?
Identifying and solving business problems with AI
As part of the DX promotion in companies and local governments, there are more opportunities to consider AI in improving the operational efficiency of routine work. However, while many hope to improve business efficiency with AI, there are cases where users are implementing them in the workplace without having a clear idea of what AI can do or how to utilize it in our work.
To utilize AI in business in the best way, we need to understand what AI can do and organize our business issues first.
At AI inside, we are conducting the “Creators Program,” a workshop using the “Learning Center” for creating high-accurate AI models with no-code. This workshop promotes the implementation and operation of AI utilization. In this article, we will share details of this workshop.
The “Creators Program,” a workshop for creating AI
The “Creators Program” is a workshop that uses the “Learning Center” to conduct the process of searching, creating, and materializing ideas for AI applications based on design thinking. It’s conducted for 10–15 participants, in groups of about five each.
The foundation of the workshop is the information shared by someone who understands the business and workflow, and conducted based on a cycle of idea generation and testing. By doing this, participants can eliminate the challenges in AI utilization and plan the most suitable AI that the frontline workers can implement and operate.
The workshop is divided into four sessions; (1) Discover (brainstorming ideas), (2) Define (narrowing down), (3) Develop (creating training data), and (4) Deliver (training and evaluating AI). Over a period of one to two months, one session per day is held for four days, and each session takes about two to three hours.
Why Design Thinking?
Design thinking is a framework used to find solutions to problems. This is done by considering your approach and techniques for thinking or organizing thoughts.
When thinking of how to utilize AI, it’s essential to identify areas for improvement based on the actual user’s perspective. Design thinking is an efficient approach to plan the use of AI. This is because design thinking is based on the user’s insight (potential needs) to determine the essential value for the user and connect this value to the ideas.
For this reason, workshops are more effective when members from various backgrounds gather together. The repeated input and output of thoughts will generate ideas and organize information.
The flow of the Workshop
We will introduce the actual process based on a workshop we held with our partner for the “Learning Center,” Dai Nippon Printing Co., Ltd.
We used “miro,” an online whiteboard tool that allows you to place stickers, shapes, and text freely in the workshop. To start the workshop, we had a lecture to understand deeper about AI, followed by four sessions from (1) to (4).
- Discover (Brainstorming ideas)
The first step is to identify the business issues and come up with solutions for them.
This step will start by writing down the current workflow and connecting the related tasks with lines. We will also identify the issues we are experiencing in work here. In addition, to organize the problems identified, we will group similar or overlapping issues into one. Then, we will bring out ideas for AI-based solutions, and summarize each idea of the AI that we think would be good to have.
Furthermore, we will go through ideas to check the impact on the work or business and how easily we can collect the data. Since AI requires data for AI-model training, we need to consider how much data we can collect.
(2) Define (Narrowing down)
While considering the background information and data types required for the AI utilization plan examined in (1) and how to incorporate AI into business operations, we will sort out what should be detected by image recognition.
If there are multiple ideas for using AI at stage (1), we will also sort out the importance of AI to the business. From the perspective of priorities and effects expected from these plans, we will narrow them down to the plan creating AI.
By narrowing down ideas and building AI utilization plans, we start collecting data for the AI to train.
Since the “Learning Center’s” core technology is image recognition, the necessary information includes the type of image, shooting method of the image, and its resolution.
(3) Develop (Creating training data)
To create AI models, we collect images as data and create training data.
Next, upload the pre-collected images to the “Learning Center.” Using the Learning Center’s function for training data creation, you can label (annotate) uploaded images with simple inputs and clicks. You can see the actual annotation in this video.
(4) Deliver (Training and Evaluating AI)
By using the function for training and verification of the “Learning Center,” we proceed with the training of the AI model.
After labeling all images, simply click on the “Start training” button to start. When the training is complete, it will show the verification results, such as the performance, and this is where you can evaluate the accuracy of the AI model.
Comments from the participants
We received various comments from the participants on how much they understand AI further and how the workshop helped to ideate for solving problems.
It was my first time attending a workshop like this and I really enjoyed it. It was a good experience to discuss with many members. It was really good to learn and think about more details on how to use AI in our lives.
Discussing issues with members from other departments broadened my perspective and led to new discoveries. It was a really good experience.
It was my first time attending a workshop where I can create AI from scratch. Design thinking was also new for me, but now I have more understanding on how to use it, so I want to apply it to my actual work.
The next step after the workshop is to improve the performance of AI and implement it in work systems. Through the workshop, participants can identify issues using design thinking, share and organize issues, and find clues that will improve business efficiency using AI. We look forward to more people participating in the design thinking workshop as a way to generate ideas to solve problems and work towards DX.
For more information on the workshop “Creators program,” or AI utilization using AI inside’s “Learning Center,” please send your inquiries here.
Contact for “Creators Program” or “Learning Center”:
AI inside Inc.(https://inside.ai/en/contact/contact-form/)