Natural Language Generation at Google Research

[@ropro9817]

More of these videos please! Love this series!

[@junyank3]

It's kind of surreal watching this in 2024. Especially considering the part at the end when they joke about writing a research paper using one of the models. 2 years later ChatGPT would be released, and today, it's basically commonplace to write a research paper with the help of these very models described in the video.

[@SogMosee]

When the two smartest kids work together on a group project and present skynet to the class.

[@zakuro8532]

which skynet though

[@oonid]

the talk looks very.. natural :D I mean naturally entertaining with lots of information about NLP and RNN :) great episode!

[@kimberlywellington4198]

This was awesome! I love the way you two interacted and shared the information. I understood the points in this video and I'm totally looking forward to watching the other videos in this series.

[@beignet1682]

I'm a total newbie in this, but that's the world I'm living in so I'm trying to learn and I absolutely love your videos - recommended by a friend. thank you so much for the videos !

[@Raaddller]

I still don't know how you guys do it. AI is beyond disabilities and languages. My 12 year old son Kevin has autism and severe learning disabilities, physically he is absolutely fine. Yet he can breeze pass through his favourite songs on KZbin on any new device that he comes across. At first we noticed he types "asdfgh" or any random array of letters, the AI recognises him as a kid and shows him an "abc kids song". He takes the pathway to quickly scan through the many options of songs "baby shark", "wiggles", "bounce patrol" and the next thing you know he's on Meghan Trainor, Bruno Mars, Pharrell Williams etc. He can switch languages to Hindi songs and English back and forth, though he can't understand any of the languages. We have given him completely new devices and he's managed to get his playlist. He has his own tablet where he's restricted and isn't allowed youtube till he completes 13. He has found a way to get around that too. He clicks an advert which takes him to google play. On Google Play he surfs a video option among the sample images, that takes him to youtube on the browser. The rest is easy peasy. Isn't that convoluted enough for any standards. Hats off to what you guys do, amazing neural network. I'm thoroughly impressed, it has taught me that intelligence is beyond what we comprehend from our traditional assessment and the possibilities are limitless even for those without conventional intelligence. Bravo you guys.

[@abhishekfnu7455]

I am feeling lucky to find this video on the internet. Super Cool techniques. Very interesting topic.

[@forrestcarbonell509]

Justin seems like such a humble guy

[@amitxerxes8465]

This was awesome! Great video and great research!

[@abhishek_0010]

It was a great episode...! Loved it. Most importantly i understood what is going on in here. Thank you.

[@郭明君-o7j]

thanks for your conversation, begin know a little about the natural language generation from structured data.

[@Garentei]

Great interviewer

[@JoeCodeswell]

What fun. Thanks for making this accessible.

[@marcusaurielius7237]

love this format

[@jenn8220]

Very helpful video!!! Love it

[@jakeroosenbloom]

This was so amazing! It would be nice to have when their research in these videos is actually being applied to Google's Product like a timeline maybe?

[@TeganBurns]

Do you think it's not? lol

[@RobertoCarlos-zu2ll]

I agree with Jake, they talk in the video about following the conversation and for example Google Assistant doesn't follow you, it's hard to get the help you need instead of just web searching.

[@micheleleggieri109]

Great!!! But... where are the tons of links? 🧐

[@GlobalMaximumDegreesOfFreedom]

About RNN. As you said, if u know the word u said, and the word u said before, then u can figure out the next word. (LIKE A MARKOW CHAIN). Insted of RNN, why not let the loop connect itself externally.... the 2 words u say can come as input to the neural network from the ear.... because u said the words....and u can hear yourself...so the words come as input using an ear. This must be an even more basic approach then having internal loops in the NN.

[@jakeroosenbloom]

andenandenia I suppose the RNN at the end of the day would be more 'Natural'

[@bhailog_gaming_2000]

More videos like these please ! loved it thanks a lot

[@santoshvaidya3752]

this is amazing and very informative.................

[@roryridge4242]

Great overview and approachable. Do you give talks at schools?

[@krispan182]

1:12 Let's talk about Google's latest research! *Pulls up a Macbook

[@roryridge4242]

😂

[@teeperry547]

Threadcount

[@BlakeEdwards333]

Great video!!! Thank you!

[@Otonium]

Awesome video as usual!

[@luis96xd]

This is an excellent video!

[@jojo21861]

very helpful video !

[@kerron68]

very good at explaining

[@Spacebaby21]

Informative, thank you.

[@stuffenjoyer]

Thank you !

[@tonini617]

More on NLP please!!!

[@JoeDoeOutdoors]

Hi I really enjoy your page ☺ Keep it up!

[@zhangxiang18]

The video is just awesome. May I know when Google Cloud Platform will launch a NLG API sot that average people like me can leverage it to create texts from structured data?

[@gandikotanagarjuna212]

Natural language generation is focused on ? Please answer my question.

[@PoosNweez]

thanks for sharing

[@muruganandanarumugam1617]

really superb

[@mobiusinversion]

Next time how about a demo?

[@adarsh_hatwar]

awesome video :)

[@tallurinagapoornima2356]

Suppose if I have a directory of files and images and all my data on my own pc, by passing a text message of data I want I have to access that file . We have to train the model such that it understands the text message I pass. Please say me how can I do that using nlp

Fan of that humor!!!

[@motoracingxp]

please explain RNN in details.

[@malingatembo8850]

Please make a followup with transformers

[@mohits94]

Gave you a like for the Hindi example!

[@eddisonlewis8099]

Awesome

[@amalaugustine2183]

how to do seo with NLP

[@elbouziadyabderrahim8086]

awesome and rich video

[@TheeSurferJim]

Dude, I think I'm ahead of Google --- Like the video :-)

[@aravindemails]

How to use image data with natural processing language ?

[@TeganBurns]

They have an API for recognizing text in images and video. The just feed it to their Natural Language API.

[@aravindemails]

Tegan Burns thanks

[@moumitadhar190]

I like it

[@achyutanandasahoo4775]

cool. good for beginners.

[@user-or7ji5hv8y]

Why use character based approach instead of word based approach? An alphabet doesn’t have any semantic like a word would.

[@samarabob]

Character level allows a much smaller vocabulary size (imagine 24 letters vs. tens of thousands of words) and actually often handles semantic/syntactic relations better. If a word-based model only knows the word "sing", for it "sings", "singer", etc. are just unknown words, whereas a character-based model can capture these similarities.

[@mikexue5104]

@@samarabob no, current NLP has a very basic function of lemmatization, so machine can tell the diifference as you listed. so, i still don't understand the goodness of character-based method.

[@ahlamabdulghani7056]

@@mikexue5104, I think character-based approach can solve the problem of unknown words, thus it can generate more choices of words.

[@matthews4159]

M-5, a computer that has synapse's like those of the human brain... - stop the attack, kzbin.info/www/bejne/r4u6hpRrhLdmftE - tie in, kzbin.info/www/bejne/oJusm2BorNxoadE 1968 star trek en.wikipedia.org/wiki/The_Ultimate_Computer

[@011azr]

Why did that guy on the right talk like a Valley girl? :/

[@kratosjordan8438]

原来中美之间的AI大战，是华人内战！

[@arar2012ify]

If any someone come from doctor شاهد ☺

[@kavipritha9669]

Awesome

[@mithsaradasanayake3211]

Awesome