Google Assistant Fired A Gun?
Yes. And no. But mostly no.
Or at least, if you have ascribe any weird fears with the situation, you're perception of reality is quite skewed.
Let's talk about what happened. An individual built a device which allowed a gun to be fired remotely. He then wrote a skill for Google Assistant which could trigger this device. He then deployed this skill in his own personal test environment, and triggered it.
What I think scares most people is that the person wrote the skill to respond with language which sounds like the AI had a hand in it, and the implication is that it would be possible to propagate this and that it would be novel.
My beef here is... all this is, is a remote firing mechanism. It could have been triggered by a cell phone, an email, or a string on a door. None of those things, at the current point in time, would cause anyone to go berserk and believe the machines are uprising. The only thing different here is the thing which triggered it... talked. It spoke a pre-determined sentence.
And just to get ahead of it... this could OBVIOUSLY be used in conjunction with ANYTHING triggered remotely by appropriately skilled individuals. Like bombs. It still wouldn't make it anything more malicious than a classical, remotely triggered bomb.
Or at least, if you have ascribe any weird fears with the situation, you're perception of reality is quite skewed.
Let's talk about what happened. An individual built a device which allowed a gun to be fired remotely. He then wrote a skill for Google Assistant which could trigger this device. He then deployed this skill in his own personal test environment, and triggered it.
What I think scares most people is that the person wrote the skill to respond with language which sounds like the AI had a hand in it, and the implication is that it would be possible to propagate this and that it would be novel.
My beef here is... all this is, is a remote firing mechanism. It could have been triggered by a cell phone, an email, or a string on a door. None of those things, at the current point in time, would cause anyone to go berserk and believe the machines are uprising. The only thing different here is the thing which triggered it... talked. It spoke a pre-determined sentence.
And just to get ahead of it... this could OBVIOUSLY be used in conjunction with ANYTHING triggered remotely by appropriately skilled individuals. Like bombs. It still wouldn't make it anything more malicious than a classical, remotely triggered bomb.
Comments
Post a Comment