Podcast: What’s AI? We made this radio play to assist.

0
4


Defining what’s, or isn’t synthetic intelligence may be difficult (or powerful). A lot so, even the specialists get it flawed generally. That’s why MIT Know-how Overview’s Senior AI Editor Karen Hao created a flowchart to clarify all of it. On this bonus content material our host and her staff reimagined Hao’s authentic reporting, gamifying it right into a radio play. 

Credit: 

This episode was reported by Karen Hao. It was tailored for audio and produced by Jennifer Sturdy and Emma Cillekens. The voices you hear are Emma Cillekens, in addition to Eric Mongeon and Kyle Thomas Hemingway from our artwork staff. We’re edited by Michael Reilly and Niall Firth. 

Full transcript:

[:15 pre-roll]

[TR ID]

Jennifer: Hello there. I’m Jennifer Sturdy… host of In Machines We Belief

Defining what’s, or isn’t synthetic intelligence could be a little powerful. A lot so, that even the specialists get it flawed generally. That’s why Tech Overview’s senior AI editor Karen Hao created a flowchart to clarify it… and collectively, we become this subsequent episode… It’s foolish. It’s enjoyable. And we hope it helps.

I additionally need to let you know about one thing actually particular we’ve been engaged on for greater than a 12 months. It’s known as The Extortion Economic system. It’s a brief podcast collection concerning the ransomware epidemic produced in collaboration with ProPublica. And It’s obtainable now wherever you wish to hear.

[Show ID]

Emma Cilikens: Women and gents… Welcome to ‘That is AI’…  

Gamers will ask questions that resolve what it’s… or isn’t… AI … And… I’ve introduced alongside an “assistant” to assist out with the solutions…  

Voice assistant: Whats up. 

Emma Cilikens: Whats up, Alexa. 

Emma Cilikens: And simply so we’re all on the identical web page… Synthetic Intelligence… in its broadest sense refers to machines that may study, cause, and act for themselves. They will make their very own selections when confronted with new conditions, very like people and animals do. 

Emma Cilikens: Now this bell… [SOT: ding] …means appropriately recognized AI… and this buzzer… [SOT: buzzer, crowd sigh] Effectively… not a lot. 

Emma Cilikens: Okay. So, let’s check your data.. Prepared… set… participant one, go! ..

Eric Mongeon: Can ‘it’ see…

Voice assistant: Sure.

Eric Mongeon: Can it establish what it sees…

Voice assistant: No …[SOT: buzzer]

 Emma Cilikens: Okay, in order that’s only a digital camera…

Eric Mongeon: okay okay… however what if it can establish what it sees? 

[SOT: ding, ding, ding]

Emma Cilikens: Yep – that’s pc imaginative and prescient and picture processing. Participant two!

Kyle Thomas Hemingway: Can it hear…

Voice assistant: Sure

Kyle Thomas Hemingway: Does it reply in a helpful, wise approach to what it hears?

Voice assistant: Sure 

[SOT: DING DING DING]

Emma Cilikens: So, that’s NLP—pure language processing.

The purpose of this type of AI is to assist computer systems make sense of human languages in a means that’s helpful. 

However what if it doesn’t reply in a helpful, wise approach to what it hears. May that even be AI?

Kyle Thomas Hemingway: If it is transcribing what you say… 

[SOT: bell ding, ding, ding]

Emma Cilikens: Sure! That’s additionally AI—it’s speech recognition, which is analogous however working from the spoken phrase as a substitute of textual content. New spherical of questions! Participant 1.

Eric Mongeon: Can it learn?

Voice assistant: Sure

Eric Mongeon: Is it studying what you kind?

Voice assistant: No

Eric Mongeon: Is it studying passages of textual content? 

Voice assistant: Sure

Eric Mongeon: Is it analyzing the textual content for patterns?

Voice assistant: Sure

[SOT: ding, ding, ding]

Emma Cilikens: Sure, as soon as once more that’s NLP—pure language processing. Effectively executed! 

Kyle Thomas Hemingway: I’ll take that very same query once more – Can it learn?

Voice assistant: Sure

Kyle Thomas Hemingway: Is it studying what you kind?

Voice assistant:: Sure

Kyle Thomas Hemingway: Does it reply in a wise, helpful means? 

Voice assistant: Sure

[SOT: ding, ding, ding]

Emma Cilikens: That’s additionally NLP—pure language processing. New query please participant 1.

Eric Mongeon: Can it cause?

Voice assistant: Sure

Eric Mongeon: Is it in search of patterns in large quantities of information? 

Voice assistant: Sure

Eric Mongeon: Is it utilizing these patterns to make selections?

Emma Cilikens: Effectively, if not, that seems like math…. 

Eric Mongeon: However whether it is utilizing patterns to make selections?

Voice assistant: Sure

[SOT: ding, ding, ding]

Emma Cilikens: Then that’s machine studying—which is when a machine learns by expertise. Okay. Remaining spherical!

Kyle Thomas Hemingway: Can it transfer?

Voice assistant: Sure.

[SOT: ding, ding, ding]

Kyle Thomas Hemingway: By itself, with out assist?

Voice assistant: Sure.

[SOT: ding, ding, ding]

Kyle Thomas Hemingway: Does it transfer primarily based on what it sees and hears? 

Voice assistant: Sure.

 [SOT: ding, ding, ding]

Kyle Thomas Hemingway: Are you positive it’s not simply transferring alongside a pre-programmed path?

Voice assistant: [Alexa] Hmmm. I’m undecided.

Emma Cilikens: Very humorous… but when so, that’s only a bot.

[SOT: buzzer, crowd sigh]

Kyle Thomas Hemingway: Okay, let’s strive that once more. Is it transferring alongside a pre-programmed path?

Voice assistant: No.

[SOT: ding, ding, ding]

Emma Cilikens: Okay, in order that’s a wise robotic, which means one which’s utilizing AI to make a few of its personal selections.

Nice….

And that’s the sport.

Thanks for enjoying!

[Music up full]

Jennifer: We’ll be again – proper after this.

[MIDROLL]

[MUSIC]

Jennifer: Many because of the proficient voices on this episode—together with our producer, Emma Cillekens, with Eric Mongeon and Kyle Thomas Hemingway. The editors are Michael Reilly and Niall Firth.

Thanks for listening… I’m Jennifer Sturdy.

[Post Roll: TR ID]

LEAVE A REPLY

Please enter your comment!
Please enter your name here