3DExperience AURA

They have been posting SolidPractices documents in the SOLIDWORKS News & Info community for the past few years. I suspect that is what they are referring to.

Those are a collection of voodoo practices and generic “you should do this and that, but also the opposite” depending on the situation.

We have some very heavy use of SW (like more than 1000 bodies cutlists) and the drawings just freeze completely.
Export the mess in parasolid with zero rebuild time and try to draw the 2d again: it is acceptable.

My take is that SW is physically limited by the parametric engine it uses and when you reach one of those limitations it is game over. best practices or not, everything you do is going to be window dressing at best.

That’s not just SW.
We need multithread 3D model engine/kernel.
We should ask AI to design one.

If we ask Aura it will say “Done, 3dexperience is ready for you to use”.

I figure that’s the next step in AI, throw in subtle marketing bs to trying and sell you something. Like ads, but worse.

Sales and Marketing have been “Artificial” for a decade or two, especially when it comes to intelligence.

I resisted temptation to continue this topic, jacking the thread, but since Fred did…

The Parasolid kernel shouldn’t have much to do with drawings. While I agree that a good asynchronous kernel would be great, that’s not much to do with drawings. All the SW users go on about drawing speed while Solid Edge users are wondering what the fuss is about. Solid Edge projections work very much different than Solidworks. I don’t know how to explain it, but when SW drawings select model features, SE drawings select the 2D edge projected onto the sheet, not the feature. I’m convinced SW drawing problems are their own, and not from the licensed Parasolid kernel.

Maybe AURA can explain the difference between SW and SE drawing behaviors.

lol I’d ask Grok before Aura to ensure a more accurate result

Don’t insult AI like that!

Yeap, AI is not the problem.

Its just everyones insistance of raming it down oru throats when its:

  1. Not needed everywhere
  2. Is not actually AI
  3. Is not needed everywhere
  4. Will not solve everything like everyone claims
  5. Is not needed everwhere
  6. Is not the solution to every problem you’ve just created to sell the solution to.
  7. See point 1).

I got some AI generated oil that can fix any problem.

One month later the Beta has shown up in my sidebar again and, well…

I kinda prefer that then it spewing bullshit like a lot of other AI’s would of done personally.

Yes, admitting it doesn’t know anything is weirdly an improvement on many others.

Edit;

Though it does fail the strawberry test.

They haven’t even hooked it up to their own Knowledge Base, which is (arguably) a useful data source.

It also says this:

This thing is hot garbage.

Still wrong, but now with added CONFIDENCE.

Great, we all need another arrogant coworker.

I asked it what the longest word in the English language was. It gave a bunch of good replies and then I dropped a Dad joke on it:

Still wrong, but now with added confident ambiguity.

2 + 2 = 5