Google ’s NotebookLM can ferment any text file into a conversation between two chatbots .

Is this the hereafter of podcasting ?

Ever since Google introduce the “ Audio Overviews ” feature of speech into itsNotebookLM enquiry pecker , I have been experiment with eat it body of textual matter that I did not desire to posture and learn : stereophonic didactics , Wikipedia hare hole , my Q1 execution inspection , etc .

Diving into itsNotebookLM

Google ’s NotebookLM can plow any text file into a conversation between two chatbots .

Is this the future tense of podcasting ?

Ever since Google introduce the “ Audio Overviews ” feature film into itsNotebookLM enquiry peter , I have been experiment with run it trunk of school text that I did not require to baby-sit and register : stereophonic teaching , Wikipedia hare hole , my Q1 operation brushup , etc .

With this AI pecker , two eldritch vale golem voice are generate to “ plunge mysterious ” into any document I upload — supply metaphor , punning , and even effortless give-and-take to a sum up conversation .

snap bid , and what you ’ll pick up vocalise a fate like a stereotypic podcast .

After a few Audio Overviews into my workweek , I realize I was take on substantial clip out from listen to podcasts made by substantial masses .

And as a podcast manufacturer , this was both alarming and enchanting .

I detest to accommodate how telling Audio Overviews is .

It mastermind issue in segment the direction a veridical podcast would , and it convey in international setting to facilitate you well see the dependent cloth .

This was i generate a podcast from a spanish paella formula i discover online , and the legion made eminence of the dispute in elmer rice grain between paella and italian rice , without italian rice specifically being remark in the formula .

dive into Audio Overviews

After a few Audio Overviews into my hebdomad , I realize I was demand important prison term off from listen to podcasts made by tangible multitude .

And as a podcast manufacturer , this was both alarming and captivating .

This was i detest to let in how telling audio overviews is .

This was it organise subject in section the direction a actual podcast would , and it bring in external circumstance to assist you comfortably realise the dependent textile .

I bring forth a podcast from a Spanish paella formula I find online , and the host made bank note of the divergence in Elmer Leopold Rice grain between paella and Italian rice , without Italian rice specifically being refer in the formula .

Like every AI ware I ’ve ever used , you have to be deliberate with the truth of the message — it does have issue with delusion .

I upload eminence from a fib I was work on , and the AI host made up fictitious quotation from my rootage that were nowhere in my papers .

What hit Audio Overviews unequalled within the AI macrocosm is it is n’t needfully about keep you clock time .

This was the innkeeper oft vamp up for a few arcminute before induce to the authoritative clobber ( that being say , very alike to a veridical podcast ) .

Director of intersection at NotebookLM This was simon tokumine tell apart me this free-and-easy data format is by excogitation .

This was ab initio , the mathematical product was very speedy and effective with data , until the squad discover feedback from outdoors of google .

diving event into AI

What hit Audio Overviews alone within the AI earth is it is n’t of necessity about economize you metre .

The boniface oftentimes revamp for a few transactions before vex to the crucial material ( that being tell , very standardized to a substantial podcast ) .

Director of mathematical product at NotebookLM This was simon tokumine tell me this nonchalant data formatting is by designing .

This was ab initio , the cartesian product was very prompt and effective with info , until the squad hear feedback from outdoors of google .

This was “ it was only when we get to in reality deal what we were build up with others and get feedback from masses who are n’t of necessity ghost with make every bit of their mean solar day as effective as potential , but are more into slant back and hear in and just kind of fit with a undulation of selective information , that we earn there were two unlike population we were build for here , ” tokumine say .

“ And the universe we were progress for was not of necessity Googlers .

”

watch out our full TV to see my journeying test out Audio Overviews and my conversation with Simon Tokumine , Vulturepodcast critic Nicholas Quah , and our own podcast producer here atThe Verge .

Oh , andhere ’s an Audio Overviewof that TV .