Dr. Maor Farid, Co-Founder & CEO at Leo AI
Computer-aided design (CAD) has been a central part of mechanical engineers' lives for the last 60 years. It was introduced by Patrick J. Hanratty, designed to serve as an extension to the engineer’s mind, which until then was forced to use pen and paper, relying only on their (often lacking) sketching talent. Some love it, and some hate it.
In the ideation stage, when we brainstorm and think about different concepts with our peers or ourselves, it’s a handy tool that serves us to elaborate on our thoughts, projecting them from our minds to the external world and sharing them with our peers, employees, or superiors. In the preliminary and critical design stages, it is our main tool, allowing us to perform the magic of engineering creation.
However, it hasn't really changed during the last decade. The same monotonous user interface and feature buttons make it feel like the (traditional) CAD world has reached a certain plateau where changes are merely incremental.
Many talk about how AI is going to revolutionize the traditional CAD world, especially after it changed so many professional fields from medical diagnostics, legal, art, creative, and more. Many players have started to emerge, but recent studies show that only a negligible part of the market has indeed shifted to adopt these tools.
Yet, the AI revolution has yet to come to the mechanical world for a few reasons, which will be discussed in another blog post. Our world as mechanical engineers is much more complex than that of lawyers and marketing people. It's not because the problems we solve are harder, but because the dimensionality of the problems is higher. Lawyers’ problems are built from words. Ours are built from geometry, material, and physical properties. While paper can absorb almost anything, our domain of problems is limited by numerous constraints: the laws of physics, physical and engineering constraints, international standards, time, budget limitations, and more.
In other terms, only highly complex, domain-specific, multi-modal learning models could address the problems we face in our daily professional lives. And when it comes, it will be massive. The change will be inevitable. Or as we call it at Leo AI, the first LMM - Large Mechanical Model (more about it in another blog post).
Will the change come in one day? Yes and no. The appearance of the LMM on the stage of history will be like thunder on a sunny day, but the adoption will be extremely gradual according to Geoffrey Moore’s law in Crossing the Chasm [1] - 62.5% of any market are mainstream adopters who wouldn’t harness any change unless it’s already a best practice. But if you ask me, in our market, it will be even slower - a world dictated by traditional forces, restrictions, bureaucracy, standardization, and accustomed to extremely slow technological changes due to the governance of old-school, low-creativity, closed-minded giant companies that educated us MEs that software should look gray, feel gray, and taste gray - from the UI’s background to the toolbar - gray.
That if software is slow, it’s because our designs are not elegant enough. If you can’t find the right button, it’s because you are not trained enough. If an action demands 100 clicks, it’s because you are not using the software correctly. But the winds of change are coming, and they are closer than ever. And when they are here with the LMM, CAD will leave its place to Artificial Intelligence Aided Design - AIAD. An AI-centric design experience that will allow us to unleash our human creativity and focus on the things that matter - innovation - better products, faster and better machines, instead of hours of tedious work of part searches, 3D sketching, and writing documents.
When exactly is it going to be here? In 1998, Moravec predicted that human-level AI would be achieved by 2040, with intelligence far beyond human capabilities, i.e., AGI, by 2050. Today, after the LLM was introduced, the predictions have become significantly shorter. Elon Musk and Sam Altman predict that it will be here by 2026 or 2029. And the same for the LMM - it could be 10, 5, 3 years, or even one month away.
References
[1] Geoffrey Moore, Crossing the Chasm, link
[2] Elon Musk - Forbes Article
[3] Sam Altman - Reddit
[4] Hans P. Moravec, Robot: Mere Machine to Transcendent Mind, 1998
Leo AI: 160 Alewife Brook Pkwy #1095, Cambridge, MA 02138, USA