 Poster: A snowHead
|
| ski3 wrote: |
| It further looks remarkably like data acquisition for analysis |
ah, you mean "big data"
as someone pointed out on the other thread (could we merge them maybe?) it is an "expert system" not an AI.
|
|
|
|
|
|
Obviously A snowHead isn't a real person
Obviously A snowHead isn't a real person
|
| Quote: |
| Three thousand metres up on Austria’s Hintertux glacier, the world’s best ski racers and a few noddys trying to stay upright on all the ice are being put through their paces. |
If this was the last week this is how it should have read as I was there too.
|
|
|
|
|
|
Well, the person's real but it's just a made up name, see?
Well, the person's real but it's just a made up name, see?
|
There is probably a neural net in the system which is one current definition of AI.
Applications like this can pick up unwanted patterns in the training data.
An example of this was a system trained to distinguish between wolves and dogs. It had been trained on lots of pictures of both and seemed to be working well until someone realized that it was actually discriminating between trees and curtains, all the pictures of wolves were shot outdoors while the dogs were indoors.
In a similar way, a data set from racers that is then applied to recreational skiers may not identify the real "faults" that they need to work on.
|
|
|
|
|
|
You need to Login to know who's really who.
You need to Login to know who's really who.
|
Clearly, the AI marketing is pompous.
But you can now see where the ski industry is headed for the next 10 to 50 years.
Smart goggles and smart clothing today are being joined by smart skis and smart boots tomorrow.
Not forgetting firstgen exoskeletons (like Mojo).
Skiing and boarding have been slow to surf the 21st century tech wave.
But they're finally catching up.
|
|
|
|
|
|
Anyway, snowHeads is much more fun if you do.
Anyway, snowHeads is much more fun if you do.
|
There was a great piece on Countryfile last night about assisted surfing.
If these 'AI' and 'Mojoesque' technologies allow the less able / skiers who no longer have the physical strength to experience the thrill of the mountains then more power to them.
|
|
|
|
|
|
You'll need to Register first of course.
You'll need to Register first of course.
|
|
Almost certainly this isn't AI but ML.
|
|
|
|
|
|
|
|
@Whitegold, ...catching up...or shifting the activity in an undesirable way?
I climb a lot (when I can)
GPS can be lifesaving. But not when it erodes the underlying nav skills, which MUST be there.
Meteo data is vital - but so is a sense of, and eyeballing, the local conditions.
I like my light pack, with GPS, with water, with compass and map.
I don’t want to be surrounded by tech, or feel that not having it is a barrier to participation
Anatoly Boukreev had the right approach - ‘....don’t find yourself over dependent on things....’ - that approach led to him being able to save many people in ‘96. He was extraordinary, and whilst attacked by Krakhauer for no o2 use, G Weston de Walt’s analysis vindicated Boukreev’s approach.
Now I know this tech might help with LEARNING and is not about safety on the hill, but I think most of the caveats made by people here are well-aimed, and whilst (possibly) the developers have already thought of them, the developers really need to show that the issues are being addressed.
The discussion of AI versus ML is important. Obviously not everyone has ideal instructors, but ROBR is really spot on - goodness me a good instructor can really get beneath the surface of performance to the deeper issues and problems.
Last edited by Then you can post your own questions or snow reports... on Mon 29-10-18 13:13; edited 1 time in total
|
|
|
|
|
|
|
|
The A.I. that could make ski instructors redundant
It can teach beginners then, can it?
|
|
|
|
|
|
You'll get to see more forums and be part of the best ski club on the net.
You'll get to see more forums and be part of the best ski club on the net.
|
|
|
|
|
|
| rjs wrote: |
There is probably a neural net in the system which is one current definition of AI.
Applications like this can pick up unwanted patterns in the training data.
An example of this was a system trained to distinguish between wolves and dogs. It had been trained on lots of pictures of both and seemed to be working well until someone realized that it was actually discriminating between trees and curtains, all the pictures of wolves were shot outdoors while the dogs were indoors.
In a similar way, a data set from racers that is then applied to recreational skiers may not identify the real "faults" that they need to work on. |
Haha.. love that! Hasn’t there been a few cases where AI struggles to distinguish between bicycles and birds?
|
|
|
|
|
|
|
|
FAQ
