Still Thinking About Artificial Intelligence

I have been watching developments around artificial intelligence pretty closely for the past year, as I am sure most of you have also been doing. In the design and simulation space we are already seeing a lot of activity around AI and machine learning in several sectors, although I am not sure who (if anyone) is adopting some of these solutions yet.

I talked about this in our January issue, but there have been more announcements since then. Ansys, which was recently acquired by Synopsys, has announced several AI-based solutions, including SimAI (a physics-agnostic, Software-as-a-Service solution) and a customer support tool that uses AI. At Autodesk University last year, the company announced Autodesk AI for creating 3D models.

Fictiv is giving the technology a slightly different spin with Materials.AI, an AI assistant (based on ChatGPT) that helps mechanical engineers select materials for their products.

There have also been announcements around other AI-based simulation, additive manufacturing quality management and other types of tools.

A new report from GlobalData indicates that generative AI (the type of AI that can learn from existing material and create new content) is poised to go mainstream, although the industry may shift to smaller, custom language models. Just the generative AI space could reach $33 billion by 2027.

But for designers to really leverage this technology, there need to be some guardrails, particularly when it comes to intellectual property. There have already been lawsuits centered on AI algorithms being trained on copyrighted material (like books and songs) to create new content.

Companies that want to use AI to help guide engineers through the design space, or to speed up simulations, will need to be careful to make sure these tools are trained on their own proprietary data. In Europe, the AI Act is poised to put some of these types of restrictions in place when it comes to how AI is used. There are bills being developed in the U.S. to help manage the potential use and misuse of AI, but I am not confident that our current Congress could pass a resolution declaring that water is wet, let alone create thoughtful legislation around an emerging technology.

For designers and manufacturers, AI should be viewed as a tool; one that can help creative designers and engineers arrive at good decisions faster. But still a tool that shouldn’t be making decisions on its own.

The healthcare industry (the focus of our issue this month) is also exploring AI for a variety of applications, including product design. In the following pages, we explore how design for additive is being used for better medical device design, as well as other innovations in data management, design and simulation that are affecting the space.