Fantastic! Have you tried using generative design iteratively? e.g. put the sensors onto the new generativly designed car, drive it and gather data from sensors, then use data to do generative design again to make it even better! Then repeat a few more times until it settles into some kind of fully optimized design. Thanks for the mind blowing video.
Exactly @Jason Voss Collect data, have the Generative AI produce a million virtual iterations of the design, pick the best, build the car, collect data, then the AI needs to analyse the data to check if it's model was in fact optimal for the real world, then do another million iterations. They don't communicate very well with the public...lack of details shared. Also how hard is it to respond to 28 comments on youtube ? this is the future...so frakking talk about it.
The only question I have is how generative design stands the test of time. I mean, let's say there's a thin support of steel, supporting a structure ; it might be optimal when the structure is in perfect shape. But what happens when it starts to rust? Can we make it intuitive enough to predict something like that?
no. it makes the most horrendous ugly bridges. look at the golden gate replacement generative design!!! there is no comparison to human engineering... yet
@Eucalypticus Imagine the complexity of an algorithm that can generate optimal designs given a complex set of constraints and tons of important variables, its just crazy to me that their capable of that
@@viktorvlasov483 AI computers are supposed to be flawless. When was the last time a computer made an error in mathematical calculations? People make mistakes, and that's a part of the reason why calculators replaced the abacus. The computers that are flawed are the fiction from Hollywood.
It's fine, as long as you've considered and used ALL of the load cases in the optimisation.
Fantastic! Have you tried using generative design iteratively? e.g. put the sensors onto the new generativly designed car, drive it and gather data from sensors, then use data to do generative design again to make it even better! Then repeat a few more times until it settles into some kind of fully optimized design. Thanks for the mind blowing video.
That's basically what they did with AlphaGo.
Exactly @Jason Voss Collect data, have the Generative AI produce a million virtual iterations of the design, pick the best, build the car, collect data, then the AI needs to analyse the data to check if it's model was in fact optimal for the real world, then do another million iterations. They don't communicate very well with the public...lack of details shared. Also how hard is it to respond to 28 comments on youtube ? this is the future...so frakking talk about it.
The only question I have is how generative design stands the test of time. I mean, let's say there's a thin support of steel, supporting a structure ; it might be optimal when the structure is in perfect shape. But what happens when it starts to rust? Can we make it intuitive enough to predict something like that?
uuuhhh crack propagation/progression?
Amazing!!! It's best options to create cable-stayed bridges or skyscrapers
*space elevator*
no. it makes the most horrendous ugly bridges. look at the golden gate replacement generative design!!! there is no comparison to human engineering... yet
Using a tool to make a better tool which will make a better tool
That is fucking amazing. I want to cry.
Incredible, love to work on it.
Can I use this concept for the Final Year Project (bachelors in Automotive) to design a part like the suspension and such?
When you will launch Dreamcatcher?
this is the ultimate question
I’m screaming
How about an inertia engine that can levitate itself under load?
I can not listen to professional people talking with the high rise at the end of their sentences.
"Upspeak", it drives me up a tree.
ARE YOU ASKING A QUESTION??
Amazing
Very cool!
Which sensor are used?
I need to see something that shows the benefits because ugly but functional can tolerated better than ugly, non-functional and very expensive
But why this software is not available for free as opensource for all the humanity?
agreed
@Eucalypticus Imagine the complexity of an algorithm that can generate optimal designs given a complex set of constraints and tons of important variables, its just crazy to me that their capable of that
Because no one will improve it for free because people have to eat.
@@wolfofautumnnight Non-sense. Why should coders be allowed to make a living?
What happens when the cameras of the AI cars are covered in mud, debris, semen and other environmental factors?
What happens with a human driving a car.... Depends on a human, the same goes for every individual AI build.
@@viktorvlasov483 AI computers are supposed to be flawless. When was the last time a computer made an error in mathematical calculations? People make mistakes, and that's a part of the reason why calculators replaced the abacus. The computers that are flawed are the fiction from Hollywood.
They also come with LIDAR sensors, which are not effected by the above elements.
Semen! LOL
Lol did you even watch the video ?
whata pathetic CHAIR