From Algorithms to LLMs: Changing focus in Design and Manufacturing
A year ago, my days were consumed by computational design and additive manufacturing. Optimizing algorithms, crafting intricate geometries, and exploring the endless possibilities of 3D printing was my professional playground. It was a world of technical challenge, pushing the boundaries of what's possible and driving innovation. It culminated in the Siemens Inventor of the year award in December 2022
However, the rise of large language models (LLMs) and their potential impact on design and manufacturing sparked some deep interest from me at this time. Could language be the key to unlocking even greater creativity and efficiency in our field?
In June, I presented at the CDFAM in NYC where LLM discussions really got me going. Many of the startups there were connecting their design tools to AI. Plus, our research in representing geometry as a computer language was a natural fit for this technology. Here is my presentation. See if you can hear where LLM's hit me over the head. (BTW, if you are looking for high level thinkers to talk to, connect with Duann Scott for the 2024 CDFAM events in Europe and the US.)
I spent the next few months evaluating the top projects around Siemens Xcelerator for the Werner von Siemens awards in Munich.
While this was hard work, it left my “research” side free to explore. I spent countless hours delving into the intricacies of LLMs like ChatGPT, Bard, Claude, and Llama, exploring their capabilities and limitations through prompts and API interactions. I devoured research papers and engaged in discussions with leading experts, all in an effort to fully understand the potential of these powerful tools.
In October, I spoke with Siemens’ experts in creativity Tobias Eismann and Bianca Höflinger where we discussed how to take over the world🗺️😄. We discussed how creative AI can truly be. How might it help product designers develop better products? Siemens colleagues and customers should talk to these folks. Here is the pod cast.
As I delved deeper, the parallels with the software development world became increasingly clear. Just as AI co-pilots are transforming the way programmers work, allowing them to become "10x" more efficient and innovative, LLMs promised to revolutionize design and manufacturing. Imagine the possibilities: generating optimized designs from simple natural language descriptions, streamlining manufacturing workflows with AI-powered instructions, and collaborating seamlessly with virtual counterparts to achieve unprecedented levels of creativity and efficiency.
I didn’t feel this is just a technological shift, but a paradigm change with the potential to democratize design and manufacturing and unlock new avenues for innovation across countless industries. However, the transition presented its own challenges. The sheer volume of information, the rapidly evolving nature of the technology, and the inherent uncertainties associated with venturing into uncharted territory all demanded a willingness to learn, adapt, and embrace the unknown.
But the possibilities far outweighed the challenges. The journey has connected me to a remarkable community of researchers, developers, designers, and manufactures who share my passion and vision for the future. Their collaborative spirit, dedication, and unwavering belief in the potential of LLMs have been a constant source of inspiration.
Today, I’m excited to be a part of this transformative journey. I love change. I am confident that LLMs will revolutionize the design and manufacturing landscape, enabling us to create products faster, more efficiently, and with greater creativity than ever before. My commitment remains the same: to contribute my technical expertise, share my experiences and insights, and collaborate with others to accelerate the pace of progress.
I think this is just the start. I plan to share more useful (and less personal) content in 2024 as I try to draw others over to the dark side🌑. Together, we can leverage the full power of LLMs to reimagine the future of design and manufacturing, pushing the boundaries of what's possible and shaping a world where creativity and technical prowess go hand-in-hand.
#LLMs #AI #design #manufacturing #computationaldesign #3DPrinting #additivemanufacturing #futureofdesign #futureofmanufacturing #technicaljourney #softwaredevelopment #copilot #10xprogrammer
I sell money to small business owners | High intent Data & AI automations
1yLove the insightful perspective! Can't wait to read your upcoming technical posts. 📚
Principal Designer & Entrepreneur
1yA great post-thanks for sharing your insights.
CDFAM Computational Design Symposium Series - Bits to Atoms
1yI agree, we are on the precipice of a major shift, and as we see the first hint of what might be possible, we will need to rethink how we approach these tools, and how we 'explore the solution space' to engineering problems, allowing for errors, glitches, hallucinations to take us into directions that may not be 100% intuitive at first, but open up ways of thinking that we may not have otherwise considered. There will be a fair few sessions at CDFAM - Computational Design Symposium in Berlin that explore AI + ML in solving defined engineering problems, but also researchers using VLM (VISION-LANGUAGE MODELS) to explore design, engineering and manufacturing processes, maybe to solve difficult problems, maybe just to show us how much work we have ahead of us to prepare the data to train the systems we hope might help us in the future. #DeusexMachina https://cdfam.com/
Managing Partner at Sterling Marine Group | Marine Asset Strategy & Lifestyle Operations
1yCongratulations on your shift to exploring AI and LLMs in design and manufacturing for 2024! Looking forward to your technical posts.