Is JARVIS possible with current tech?

[slightlyjaded]

So, I found myself rewatching some marvel films the other day, and was trying to figure a way to make the jarvis interface work....

 

The hologram technology doesn't work with current tech, But we do have any number of VR headsets hitting the market in various forms now and in the next year. How about instead of a hologram we use an augmented reality for projection....

 

As to reading gestures, kinect did some great leaps forward but wasn't developed enough to work consistently. However intel demonstrated some 3d modelling software in the last year that could with appropriate hardware be scaled up to real time applications, and a larger scanning base. with the appropriate software  it probably wouldn't even need to be as hardware intensive as it doesn't need to model the entire scanning space, merely those things that change.

 

 

So my question is how nuts is this concept? I'm aware it would take significant software investment. and at the moment it's probably more in the realm of darpa than any single company. I think the tech is there at the moment, is it practical though, and would the creative advances be worth the outlay?

 

Could you imagine ken done working with a virtual canvas?

 

EDIT: Just highlighting the fact that while the AI part might, or might not be possible, I don't feel it's commercially viable at this point in time. What I'm far more interested in is the implications of the interface. I feel that is viable at this time, and might be worthwhile.

[EposVox]

Generally speaking, if something has already been thought of, and the technology exists, then we'd have it.

 

Hologram projection to any solid degree is pretty far off still. We have AR stuff, but in VERY early stages, and from what I've seen controlling it with IRL actions isn't top-notch just yet.

 

So if you crammed Siri/Amazon Alexa with an Oculus and some AR software, you could have a ghetto-rigged JARVIS - but it'd be nothing worth using for more than a novelty toy.

[ThatOneRussian]

Nope, and isnt it even theoretically impossible an AI would have to be able to have a reaction for every single action or thing that can happen in the universe. Program it for every single skin colour of a human or race. Racists robots lol. Who would have time to write all that code.

[Virtual01]

Nope, and isnt it even theoretically impossible an AI would have to be able to have a reaction for every single action or thing that can happen in the universe. Program it for every single skin colour of a human or race. Racists robots lol. Who would have time to write all that code.

 

The idea is that the AI will be able to write it's own code, becoming exponentially smarter and self-improving. This is exactly the reason people are terrified of what it could do if left unchecked.

[slightlyjaded]

Nope, and isnt it even theoretically impossible an AI would have to be able to have a reaction for every single action or thing that can happen in the universe. Program it for every single skin colour of a human or race. Racists robots lol. Who would have time to write all that code.

was more interested in the interface than the actual AI part of the program.

 

My point is that it may not be financially viable outside the enterprise, and military market, But with current tech it "might" be possible

[ShadowCaptain]

A lot of it is kinda possible with todays tech yes, though it would be little more than a VR interface with an advanced SIRI

[Maleko]

Generally speaking, if something has already been thought of, and the technology exists, then we'd have it.

 

Not really, so much tech is being used that 'we' the general population have no idea about. 

[slightlyjaded]

A lot of it is kinda possible with todays tech yes, though it would be little more than a VR interface with an advanced SIRI

That was actually the point of this post. I feel that with the current tech, the really cool part (to me) of the program is possible. Apple, and microsoft are really pushing the touchscreen concept ( and yes that's nice in appropriate applications) But I think with current tech we could remove the screen.

 

The problem i feel isn't so much the hardware that i think does exist and could do the job.... It's the software to integrate various parts and make it all work together properly.

[ShadowCaptain]

That was actually the point of this post. I feel that with the current tech, the really cool part (to me) of the program is possible. Apple, and microsoft are really pushing the touchscreen concept ( and yes that's nice in appropriate applications) But I think with current tech we could remove the screen.

 

Well that stuff already exists, ie manipulating virtual objects in a 3d space, we have augmented reality glasses like hololens, and we have voice assistants

 

And these are all pretty much commercial projects - the prototype and research stuff is even more advanced

[Bl00dgod]

Nope, and isnt it even theoretically impossible an AI would have to be able to have a reaction for every single action or thing that can happen in the universe. Program it for every single skin colour of a human or race. Racists robots lol. Who would have time to write all that code.

the ai itself wouldve time to do that and the internet its connected to would deliver the data

 

 

remember GPS? once upon a time it was a military "toy" to hit the correct target with a rocket

we dont know what awesome stuff the military keeps as a secret and uses    ( invisible bond car? )

[ThatOneRussian]

the ai itself wouldve time to do that and the internet its connected to would deliver the data

 

 

remember GPS? once upon a time it was a military "toy" to hit the correct target with a rocket

we dont know what awesome stuff the military keeps as a secret and uses    ( invisible bond car? )

yeah because having a robot listening to ISIS is a good idea.....

[slightlyjaded]

yeah because having a robot listening to ISIS is a good idea.....

I thought that was called echelon.....

[EposVox]

The problem is that technology has gone away from being user-friendly and only brand-friendly.

 

Touch screens tech, GPS tech, VR tech, hololens tech, especially smart home tech - all of it is proprietary and you need to only use 1 brand ecosystem at a time; nothing works together.

[AlTech]

YES it is possible. Cortana could be made to be as good as Jarvis. You just have to code the extra bits so that it can respond like a human being would.

[DebatED Nothing]

Gesture reading is already pretty good with the leap motion controller (see this for example).

[Zyre]

For us to build JARVIS, we first have to develop quantum computing. The intelligence of JARVIS can't be done with current silicon imo. 

[RedWulf]

Well Jarvis is just the personality created to give life to the AI and interface...so, a virtual assistant that can manage multiple systems and give you some various stats/news is very possibly. Basically a very complex and hyped up Siri. The read out/gui that Tony sees from the helmet is basically google glass on steroids. So, if you put time and alot of funding into the development of the google glass technology then you could have the helmet. Jarvis himself would require a massive amount of programming and trust. Most of the Iron Man scanners(thermal, UV, object/distance sensing, ect) are already widely used. The "self driving" cars and security systems employ them, you would just have to set them up and create a program to manage/interpret them. As for the holographic and interactive interface that he uses to give form to Jarvis, design his toys, and uses as screens...good luck, WebGl has the closest thing to that but its a small chamber with very limited interactivity.