"OK Google, activate gun"

[HEXUS]

Apple assassinating art installation raises questions about ethics, AI and legality.

Read more.

[Nelviticus]

What a crock of nonsense (to put it politely).

What if you trained a dog to activate the gun on a spoken command? Would we then be worrying about 'the rise of the dogs'?

[Ttaskmaster]

Would we then be worrying about 'the rise of the dogs'?

Nope - Most dogs can be halted in their tracks and turned to your control simply by presenting them with a Dentastix... or in extreme cases, a 4" strip of biltong!

[Phage]

Our dogs love those rawhide treats. they don't seem to last long

[TeePee]

Using Google to literally destroy Apple.

Raises no questions to me... This is just a voice activated solenoid. Sure, if it were a real gun rather than an airgun it would be illegal, but that's not the point. It's the Google vs. Apple that is the point of the 'Art'.

[Ttaskmaster]

Sure, if it were a real gun rather than an airgun it would be illegal, but that's not the point.

Neither is it the point that an actual firearm would probably half-destroy the actual 'art installation', too... but still.

[Dashers]

Suggesting that AI is shooting the gun is tenuous at best. AI might be used in the voice recognition, but there is plenty of non-AI based voice recognition stuff out there. It also not discovering and learning about the weapon and how to use it. It's merely running through the steps to activate the solenoid.

I know this is art and supposed to be representative, but I feel it does a rubbish job of it and I'm not sure why it's news worthy.

[Corky34]

What if you trained a dog to activate the gun on a spoken command? Would we then be worrying about 'the rise of the dogs'?

No, instead we'd be probably be questioning the ethics of training a dog to activate a gun on a spoken command.

It's not so much a question about the rise of the machines, or dogs for that matter, but a question of ethics, in either case who's responsible the machine, or dog, who activated a gun and hypothetically killed or injured someone or would it be the person who made it possible for the machine/dog to carry out a murder.

It's just another way of framing the trolly problem.

[Saracen]

I'm struggling to see where "AI" comes in.

I mean, years ago (10-15-ish) I was using Dragon voice software to write articles. It aas pretty good at analysing my spoken voice, and turning it into text. Which it then put on-screen.

But it wouldn't have taken much by way of digital to analog conversion to have something physical happen in response to a computer command, like clicking "light on" and have an electronic switch thrown.

So .... connect the two. You then have voice command resulting in real-world event.

After all, we've had robot-aided manufacturing and assembly for years, and this just adds a voice-control element to a simple motor.

That's not my interpretation of AI, at all.


Had the voice said "kill that apple" and the AI had determined what needed to be done, ordered the parts, built a control mechanism, selected the best weapon and killed the lone apple in a basket of mixed fruit and veg, I'd be impressed .... assuming of course that it had worked out what to do by itself and not explicitly been programmed with "If told "kill apple" then follow this sequence of actions".




That said, I did laugh(silently, for about 0.25 seconds) at the choice of an Apple to be shot by Google.

A better joke would have been the other way round. If a SkyNet robot ever decides the best action is to shoot Google, I want to shake it's .... erm ..... robotic external manipulation device.

I might even give it a kiss.

[Ttaskmaster]

would it be the person who made it possible for the machine/dog to carry out a murder.

This obviously, for without human intervention and training, neither machine nor dog would have known how to handle the firearm... and both would have just been responding with whatever basic and limited understanding they had of the scenario. Neither can fully comprehend the complexities of human society and interaction, which is why the human is responsible for training them well and keeping them under control.

The day a computer can fully understand all this and reliably make sound, safe judgement calls... I'm installing it in a black Pontiac Trans-Am Firebird with a red Larson scanner in the nose, and going out to fight crime in a leather jacket and bouffy haircut!!!!

[CAT-THE-FIFTH]

What a crock of nonsense (to put it politely).

What if you trained a dog to activate the gun on a spoken command? Would we then be worrying about 'the rise of the dogs'?

Good thing Moose are untrainable!!

[scaryjim]

Suggesting that AI is shooting the gun is tenuous at best. ....

I'm struggling to see where "AI" comes in. ...

Well, this is art, so its link to actual science was always likely to be tenuous.

AFAICT the assertion is that the smart speaker system could, via ML, 'learn' under what circumstances you normally fire the gun, and then 'decide' to fire the gun without a direct command should those circumstances arise.

I have no interest in smart speakers so I have no idea if that degree of ML occurs when you make a request, although I could easily see, for example, Amazon identifying patterns in your product orders for consumables and pre-emptively ordering you stuff when it thinks you're going to run out. I can see where the artist is coming from, even if it's a bit of stretch.

[Saracen]

Well, this is art, .....

I suspect there's a whole and complex thread to be had just in refuting or supporting that four-word assertion.

Is it art just because the 'artist' claims it is? Well, personally, it's about as much 'art' as a certain famous .... or infamous .... unmade bed.

So if I collect a large pile of highly fresh, and 'fragrant' cow dung, is it art? Maybe if I call it "Reflections of society's future", call myself an artist, and claim it's a social commentary, it'd win a Turner prize.

To some it'd be insightful. To others, it's a load of s.... erm, is that the doorbell? Gotta go.

[Dashers]

Making statements about AI and what not, without being functional seems like a perfectly good pursuit for artists commentating on technology and society. This though I feel is bad art. Getting the message across would be better done without a functional Google box and have it hard-wired to illustrate AI decision making firing weapons. This is just a poor attempt to associate AI with something dangerous, because outrage.

[Ttaskmaster]

Is it art just because the 'artist' claims it is?

Nope.

A piece of art should be "expressing the author's imaginative or technical skill, intended to be appreciated for their beauty or emotional power".

While this is a statement piece and on some levels is quite clever (IMO), it's still a statement, not actual artwork... even if you stretch it to be all about his 'imaginative skill'.

[Pob255]

This is just a poor attempt to associate AI with something dangerous, because outrage.

This.
Sorry but the whole thing is not insightful it's just trying and failing to be controversial.
It's mediocre art at best, however it is controversial enough, esp to those without a basic grasp of what current AI systems are, to get news media coverage.
Which elevates it above mediocre.
It's a subject art should look at and challange, shame this boring flat piece is what's turned up

While I'm not sure Tom Scott started out to make an art installation/performance piece, this is still an amazing tech/art piece