Put That There [video] (1983) | Hacker News Reader

Put That There [video] (1983)

youtube.com

78 pointstosh16 comments

The understanding of context here is maybe the most "hidden" feature, but the most impressive to me, because it's something that's still not commonplace in today's technology. For example.

I have an Ecobee thermostat with Alexa in my kitchen. The Ecobee has the location "kitchen". My kitchen has IoT-enabled lightbulbs.

I have a Google Nest Mini in my living room. That Nest Mini has the location "living room". I have a lamp in my living room with an IoT-enabled lightbulb.

I have a Google Nest Mini in my master bedroom. That Nest Mini has the location "master bedroom". The nightstands in the master bedroom each have a lamp with an IoT-enabled lightbulb.

If I ask any of these devices to "turn on the lights", they turn on every single light in the entire house. Me, my wife, and my children are all trained now to tell the devices exactly which room should have the light bulbs turned on, because even though I've told each device what room it's in, they're still completely context-unaware.

I've thought about trying out whatever home-mini-pot-hub thing Apple came out with most recently, to see if they're more context-aware. But this is still a far cry from the demo in the OP. I wouldn't dream of telling these devices to "turn on all light lights on the second floor", or "turn on the lights in the big room next to the kitchen", or "turn off the lights after we leave the house".

Forget security, competing standards, walled gardens, lack of visibility, etc. Until IoT devices are more aware of this kind of context, it will always be easier and more reliable to flip a lightswitch yourself than to ask another device or robot to do it for you.

That's weird. My Google Home knows what room it is and "turn on the lights" acts accordingly. Although I did do the extra step to assign room locations to all the devices in the Home app.

HomePods are in fact context-aware like you want. They will only turn on lights on their room if you say “turn on the lights.”

And yet, that Eames chair he's sitting in looks as timeless and cool today as it did in 1983 (and as it did in 1963).

I find that notion an interesting sub-commentary on the quaintness/coolness of the tech.

And yeah, all three times, totally out of reach of mere mortals on mere mortal salary. :(

I keep looking at Eames replicas, wondering if that's an acceptable trade-off in quality. I also REALLY want an Eames. And I can afford a replica.

(Not all knockoffs are of dubious quality/legitimacy. I hear the people at Iconic Chair bought the rights to use the same molds from Herman Miller.)

I bought a new knock off Eames Lounge replica with Ottoman, made of faux leather for 600 EUR. Now, 5 years later, the buttons are gone, the leather in the seat is torn. Wood and metal are great though.

Nice to see Chris' work here. He's an amazing indivdual, hacker, and geek.

Check also some of his later works. Early Audio Augmented Reality work: https://www.media.mit.edu/speech/papers/1997/sawhney_ISWC97_...

And skin interfaces: https://dl.acm.org/doi/10.1145/2971763.2971777

Hey Siri. Tell @dang to put [1983] in brackets heree.

I love the "pay attention" command :)

This makes me wonder if Siri's ever got told "Hey Siri. I want to talk to your manager".

“I’m not sure I understand.”

I would bet if you made a modern version of this tech demo today with modern graphics, people would be excited about it.

To me this is evidence that for some reason no one wants these kinds of interfaces, and they are largely just cool tech demos. We've had the capability for almost 40 years, and there's effectively been little to no progress from a pure functionality standpoint. Sure, your voice commands can (sometimes) be recognized by a device you carry in your pocket, but fundamentally there hasn't been a WOW moment for a very long time.

I worked in a 3d graphics research lab. It seems some of the issues the video is indirectly addressing (pointing at the screen without touching for "cursor") are still tricky. The accuracy is still not there close up, is it? We have eye trackers in helicopters, but without a laser/styli, me pointing at a screen is not that accurate without a lot of (probably still today) expensive gear.

I do agree, show people this today with modern screens and they might get excited. I can see some real use cases. A classroom or presentation setting for example.

In VR, the controller can be a laser pointer/cursor with very high precision and reliability (if your tracking setup is optimal), many VR menus use this mechanic. But then you're wearing VR controllers and a headset, and only you can see the screen unless it's mirrored to a real screen or you're with others in VR.

Where?

There

Reply on news.ycombinator.com