"Be curious, not judgmental."
– Walt Whitman
1. ChemGPT
Well, who would have thought? LLMs are useful for predicting the properties of crystalline materials and more so than conventional simulations, even! At this point, is it even a surprise? Did we really need Princeton to tell us this? No, but they carried out the scientific method, and we didn’t, so there you go. They trained a version of Google's T5 model on descriptions of over 140,000 crystals and demonstrated its predictive capabilities on known substances. So, when will we see LLM-enabled CAM already?
2. Terminator T-1000 IRL
We all remember that scene from “Terminator 2: Judgment Day” when Arnold says, “Mimetic poly-alloy.” You probably read this aloud in his voice just now: “Liquid metal!” I miss the ‘90s sometimes, and if you do too, shapeless computing is bringing a piece of the decade of grunge back! “Researchers from Tsinghua University in Beijing have developed a fully flexible resistive random-access-memory device, known as FlexRAM, using a gallium-based liquid metal (GLM) to write and read data.”
3. 3D Printed Pneumatic Logic Modules
Deciding between two different technologies is often answered with a single question: Which one is faster? Comparing a CMM to optical inspection technologies is not so simple. The driving factors are accuracy, speed, precision, surface conditions, and portability. Well, I guess it can be answered in five questions.
4. A New Occupational Category
“Bill Bonvillian and I lay out the case for a new occupation to help firms move into advanced manufacturing: the ‘technologist.’ They would bridge the gap between technicians and engineers. We would create this new workforce by equipping people with timeless, industry-agnostic competencies. Our plan would also address gaps in education, wages, and opportunity.”
5. Manual vs. Automatic CNC
Movie time – and this one is a banger! Watch Brandon white-knuckle race his grandfather’s 30-year-old manual machines against a modern high-speed CNC! OK, so I’m not going to spoil anything, but the results did catch me a bit off guard, and, well, let me put it this way: This race was surprisingly neck and neck, and, uh, not for nothing, but it’s because the current state of CAM software needs work. That’s me being nice. Come for the shop talk, stay for the innuendos. Enjoy!
To get the latest tech developments delivered directly to your inbox, subscribe to the weekly Tech Report here.
To access Tech Trends, log in to or register for an MTInsight account at https://www.mtinsight.org/
