The end of Rhino?

The interesting thing is that not a single AI slop acolyte can ever show a basic step-by-step example, how, taking some user research, mechanical requirements, and production parameters on board, even a reasonably innovative product like this or that is prompted into existence.

In engineering, the hard part was never only producing geometry. It’s tolerances, process limits, standards, and someone taking responsibility when a part fails. In moulds and tooling, a mistake costs steel and machine time, not another prompt.

This post is an April Fools’ joke, right? Or is someone already planning to put people on rockets designed by bots?

tesla-roadster

yes there is far too much to be considered and details for these automated tools to do it. the only way forward for these tools is for everyone to give up any expectation of specificity. as long as you dont have any specific opinions about details or combinations of details, and you are happy that there is no individual or organisation that will take responsibility for the correct choices having been made then a.i. will be the choice but thats already the same choice that ‘takes’ low effort / low value work from seasoned proefssionals and delivers it to younger less trained workers. a.I. is goign to make starting in any idustry harder and cause a massive backlog in skills in aroudn 5-10 years time when nobody has been recruited to do entry level anything, and nobody has been afforded the time and projects to learn and grow on.

I’ve read a while ago that they expect around 40 to 60% of these jobs to not come back because of the increase in produtivity.

Yeah, they are improving the prices, do you think that is coincidence that all these companies made their token usagen limits and stuff higher after DeepSeek released their inference model?

As I said before, I’ve been using agentic solutions myself to do all kinds of work. Agentic use allowed me to work on side projects while I am at my regular work, I just setupd a taks lists and prompt guidance before going to work everyday and leave the agents doing it so I can take a look when I am at home, fix some mistakes and deliever the job.

The thing is, they are not and will not be the solution to everything like its being said here, “There will be no Rhino 10” is a inflated as hell expectation. Unless they remove all the workers from all the factories there will be need for detailing and production and workshopt drawings and this kind of stuff.

Another thing is, can an AI bot have a Professional Engineer License? Can it be sued by a mistake that causes an accident and takes people lives?

Another question is about energy, the powergrids for the datacenters are accompaning the consumption? I’ve see a news article about the US closing a deal with Paraguay to install datacenters over in SA because the energy is cheaper here. Why is that any different from moving the factories over to China a few years ago? Will there be job loses on that side of the equation as well?

I said it before and I will say it again, Grok Bot will not end Rhino, it will end workflows and replace designers, architect and engineers to some extent, but Rhino itself for now is not going anywere, and if it goes, we create a persona of Bob McNeel using AI and create a clone to keep it going using AI itself for it :stuck_out_tongue:

Not to start a war or anything, but fairly certain the OP is an AI bot.

It’s message doesn’t evolve with the convo, no real responses. Just the same challenge. Multiple posts in a row, etc…

Things will go boring after the IPO, they worked hard for almost 11 years and people need cash. after that AI will be just normalized with less cool drama involved.

your example begs the question - if you have to spend 3 months checking output, then are you actually saving any time vs a “traditional” building design process?

sorry to derail but out of curiosity..

Any of those cars use 3d printed metal parts?

We need only look at the typical quality that AI generates, and the average level of accountability. It’s actually very similar to bad BIM work where its the people next in line that fix everything.

If I get AI to spit out a model and as I go through that model, I have to fix every little detail… it would have just been better to model it myself.

“Design Development” is a process and it takes time. Let’s say a product takes about 8-10 weeks of design development followed by a couple weeks of testing. AI advocates might claim they can knock the project out in mere hours… but unless they spend AT LEAST the same amount of time testing it, the likelihood is that it’s not going to be good.

FWIW, If I hire someone to do something technical and find out they’re relying on AI to the extent claimed here, I would fire them and either hire someone else or just do the work myself.

I read half of it, but the rest I might tell claude to copy it verbatim then read it back to me, then decipher it and read it to me in english version, and maybe make it make sense. :joy:

I’m busy building Rhino plugins, though so might be a minute.

No, but the car in your photo uses plenty of 3d printed metal parts that are prone to breaking due to bad optimization of the geometry.
That does not mean that the car will disassemble itself while driving it, but I see a lot of room for improvement in many of these parts, both for reducing the mass and increasing the strength.
This is what happens when people rely on automatic solutions made by programs.

An easy example. They use the same diameter for the “rods” of parts where one side is pulled out and the other is pushed against. What’s worse, some of their parts have noticeably thinner pulling rods than pushing ones.
People with real experience know that the pulling rod must be thicker due to the effective tensile strength (ability to take pulling stress) being far lower than the compressive strength.

On a side note, when it comes to cars, robots, etc, manually designed mechanical parts look far better than the “organic” shapes.

Thats just a breath of fresh air

I’m starting a new course with metal AM this is what we are supposed to learn

I have had my doubts about it knowing from past experience with cnc that metal sometimes is looked under a camera for defects, a slight crack and the whole part is thrown away

Thanks for the input i learn a lot from your posts here, very educational.

Another thing that those who are obsessed with AI-generated structural and mechanical parts often forget is that billet aluminum alloy is far stronger than its equivalent 3d printed alloy. When they brag about the “superiority” of the 3d printed parts that were compared to the CNC-milled counterparts, they mostly use exactly the same material properties in the FEA analysis. That makes the CNC-machined parts look “weak” in comparison.

On top of that, 3d printing is still not 100% safe due to possible cracks or simply incomplete melting and bonding of the metal powder.

Also, many of the 3d printed parts are machined improperly where bolts, stainless steel sleeves and other fasteners will be connected. After the machining, straight edges are left that cause concentrated stress and weak point of breaking due to the lack of a gradual removal of the material. It’s crazy how many of those who manufacture 3d printed metal parts lack basic knowledge about proven mechanical principles.

Through my journey so far even tho I like the tech i am skeptical alot of the times about 3d printing usage.

There is a drive to paint green fields with it when reality hits it’s not all that great

I’ve witnessed all levels of management fail and put the wrong people in places they should not be in this.

It’s infuriating sometimes but there is nothing i can do when even the CEOs don’t care.

In the end they just put those mistakes and cost of failures on the customers.

Sorry for the off-topic, but could you substantiate that claim with scientific sources?
We did exhaustive testing with 3D printed Inconel 718 about 10 years ago and came to very different conclusions than that. And, no, I can’t publish those results.
-wim

give us the dirty details when you got time.

I’m interested in metal printing but as far as I know the (shrinkage) of 30% +/-5%ish doesn’t do it for me. but hopefully lasers get cheaper and stuff.

I would promote 3D printing molds out of metal, but it’s still in the stone ages imo.

better off printing out of plastic and using electrolysis, electroless, magnetrons etc.

from the AM side metall 3d printing does not make sense when you try and scale it up it is just to slow and expensive

fits for prototypes and companies like NASA that dont have to produce in series but make complex geometry that would be a nightmare cnc machine it

seeing how things move in light speed these days in this field this could change very rapidly but for now the general consensus is just that.

seeing how to prepare a job with these machines is also really strange aswell as how sought after it is.