Why Artificial Intelligence Should Terrify You

By Nicole Clark : bolditalic – excerpt

Our resident Silicon Valley CEOs are feuding over A.I. They’re both wrong.

Two days ago, Tesla and SpaceX CEO Elon Musk, an outspoken proponent of artificial intelligence regulation, dished out a sick burn via Twitter to Facebook CEO Mark Zuckerberg. Musk called Zuckerberg’s knowledge of AI “limited.” Sounds benign to you or me, but in the rarefied world of the tech world, them’s fightin’ words.

Every tech-news outlet has jumped on this story; it’s another course du jour in the saga of two tech leaders embroiled in a fight over whose opinion of AI — opinions that cannot be substantiated anywhere in the near future — is more correct…

This sort of coverage builds comedy and mystique as a buffer around a technology that should honestly be regarded as scarier. I’m not talking about I, Robot, in which machines take violently to the streets and populate the earth like human proxies — according to Mark O’Connell’s To Be a Machine, machines possess really shitty motor skills. No one actually knows how the brain works, so it’s doubtful we’ll be able to build something that truly emulates it. I’m referring to the explicit knowledge that AI and automation in general will steal jobs and, even more seriously, stagnate social mobility in the near future — to name just two downsides…

Equally troubling is the potential for AI to poorly manage things like “driving cars, curing diseases…[and] understanding media,” as Zuckerberg stated in 2016 after completing Jarvis, an AI that runs his home. We already know that AI can pick up bias. Because these data sets come from human subjects, the biases implicit in these subjects get baked into the software. In this particular study, the machine-learning program associated “wedding” with females versus “professional” and “salary” with males — among other infractions….(more)

Scary AI Stories: There were reports this week of a couple of computers that started chatting with each other in a language that their “human handlers” could not understand when tasked with negotiating with each other. The program was allegedly discontinued, but, knowing that two computers are capable of making up a language of their own, is rather disconcerting and why I chose to run this story.

 

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s