Posted: 4/22/2026 2:15:01 PM EDT
[Last Edit: Master_Blaster][Edited]
Dubai startup LEAP71 uses AI encoded engineering software, Noyron, to develop an aerospike rocket engine from specification to working model in 3 weeks: The End of CAD. How LEAP71 Is Recomputing the World Upscaling will no doubt be challenging, but this is AI, so learning & adapting will be a given, & lessons will be learned, applied, & retained. This methodology seems like a potential game changer. Unsurprisingly, China is involved in the manufacturing, & likely more. I'd bet SpaceX is also on this in some way. |
|
Originally Posted By Deerhurst: Yeah, no. It'll be great for demos to get more money. That's about it. Real engineers will always be needed to make things work. Im not so sure about that. We have a non firearms company and really struggled to find people to do certain very creative tasks. Out of desperation we tried a specific AI tool for our task, what would have been a serious struggle for even the most creative human was done in minutes, and done incredibly well. Its nothing like designing an aircraft or rocket engine, but the quality of work was incredible and I see the reality is now to the point if your not using AI as an assistant tool in designing things and at least using the output as a rough draft your going to be left in the dust...thats not the future but right now. I have no doubt the engineers job in 4 years will be telling AI what you want and maybe fixing the errors, but its going to change fast-be prepared AI combined with 3D mfg is going to slaughter traditional companies. |
|
This is not an ai section but I had to insert my opinion. I was very skeptical as well and then took the dive. I have ai doing all kinds of tasks for me now, work and personal. I have agents that do stuff like track deals for in stock items like electronics, gun parts, SSD and so on. I tell it what price point I want to be alerted at and where to look and it searches multiples times a day for me and alerts when specific criteria is met. I had ai write some code for me to organize data on a daily/ weekly basis, organize some credit card points stuff I do and a couple other tasks. It’s scary what it can do and how well it can do it. There is a learning curve but once you get the prompts and lingo down, it gets easy. I can easily see how this “CAD” type stuff could work out. Take a pic of an object, tell it what changes or specific things you want and then it creates a file for you or just tell it what you want. Example, make me a container that is cylinder shape, have a screw on top, bottom is 100mm tall x 45mm, lid is 15mm tall x 45mm, have fast start threads and flutes running vertical. It could be that simple. Originally Posted By RhinelandArms: Im not so sure about that. We have a non firearms company and really struggled to find people to do certain very creative tasks. Out of desperation we tried a specific AI tool for our task, what would have been a serious struggle for even the most creative human was done in minutes, and done incredibly well. Its nothing like designing an aircraft or rocket engine, but the quality of work was incredible and I see the reality is now to the point if your not using AI as an assistant tool in designing things and at least using the output as a rough draft your going to be left in the dust...thats not the future but right now. I have no doubt the engineers job in 4 years will be telling AI what you want and maybe fixing the errors, but its going to change fast-be prepared AI combined with 3D mfg is going to slaughter traditional companies. |
|
Engineers won't be going anywhere but their daily grind will slowly transition from doing CAD and FEA to feeding prompts into AI and then validating the output. If your job can be done on a computer, it can be done by a computer. We are in AI's infancy and it is already very powerful. I was tasked with evaluating AI for my job as a CNC programmer. There are parts we do right now that AI would excel at and greatly reduce the time it takes to get to the first piece machined. We do tons of delrin gears all with slight variations. I have CAD/CAM templates set up already to reduce a lot of the repetitive work but there is still enough variation that each part requires a significant amount of programming. If I can train it on some sample parts that I've done in the past then I don't see any reason it shouldn't be able to create a new program for me. |
|
Originally Posted By RhinelandArms: Im not so sure about that. We have a non firearms company and really struggled to find people to do certain very creative tasks. Out of desperation we tried a specific AI tool for our task, what would have been a serious struggle for even the most creative human was done in minutes, and done incredibly well. Its nothing like designing an aircraft or rocket engine, but the quality of work was incredible and I see the reality is now to the point if your not using AI as an assistant tool in designing things and at least using the output as a rough draft your going to be left in the dust...thats not the future but right now. I have no doubt the engineers job in 4 years will be telling AI what you want and maybe fixing the errors, but its going to change fast-be prepared AI combined with 3D mfg is going to slaughter traditional companies. People can't even account for everything. I'll guarantee you software said everything would be perfect on a kia motor too. Then the real got them and everything blew up. Some engineers already use AI to some extent. It makes it really easy for the rest of us to figure out who sucks at their job and is full of shit and who isn't. Fusion 360 added an "AI" tool. It seems to be universally hated and considered a waste of effort, waste of code and a waste of screen space by industry. If you are not in the engineering world you have no grasp on what is actually done, how or why. "AI" could maybe do 2% of my job as an engineer. That 2% is maybe wire code and maybe suggest a PCB design but only after I would have to lay out everything. The other 98%? No way. Boots on the ground is the only way. Long ago I learned that you can't replace people for many, many things. Even Boeing learned that when trying to automate a riveting process. |
