00:03 - 00:06

[Music]

00:06 - 00:11

hey everybody Welcome to day 11 I'm

00:08 - 00:14

Kevin wheel I lead product at open aai

00:11 - 00:16

and I am definitely outclassed by the

00:14 - 00:19

two gentlemen on my right who I'm told

00:16 - 00:21

did not just get these uh suits 48 hours

00:19 - 00:25

ago on Amazon definitely definitely own

00:21 - 00:26

this already 100% yeah all right so you

00:25 - 00:29

might have noticed we've been putting a

00:26 - 00:31

lot of effort into our desktop apps uh

00:29 - 00:32

so we we launched our Mac desktop app

00:31 - 00:34

about 6 months ago we launched our

00:32 - 00:37

Windows desktop app just a couple months

00:34 - 00:40

ago and as our models get increasingly

00:37 - 00:43

powerful jet GPT will more and more

00:40 - 00:47

become agentic uh and that means we'll

00:43 - 00:49

go beyond just questions and answers

00:47 - 00:51

Chet GPT will begin doing things for you

00:49 - 00:54

we're seeing that already uh with

00:51 - 00:57

products like canvas where uh you're

00:54 - 00:59

collaborating with chat gbt to help

00:57 - 01:01

improve the your writing and your code

00:59 - 01:03

and that shift will continue and we'll

01:01 - 01:05

do chat GPT will do more and more on

01:03 - 01:07

your behalf the desktop apps are a big

01:05 - 01:09

part of that too because being a desktop

01:07 - 01:12

app you can do so much more than you can

01:09 - 01:14

just in a browser tab that includes

01:12 - 01:16

things like with your permission of

01:14 - 01:18

course being able to see what's on your

01:16 - 01:20

screen and being able to automate a lot

01:18 - 01:23

of the work that you're doing on your

01:20 - 01:25

desktop so we'll have a lot more to say

01:23 - 01:26

on that as we go into 2025 but we've

01:25 - 01:29

also got some exciting stuff that we're

01:26 - 01:32

launching today so let's dive in all

01:29 - 01:34

right hi I'm John nastos and I work on

01:32 - 01:36

the chat GPT desktop team my name is

01:34 - 01:39

Justin Rushing and I also work on the

01:36 - 01:40

chat GPT desktop team uh we've got a lot

01:39 - 01:42

to show you today so I'm just going to

01:40 - 01:45

go ahead and get started

01:42 - 01:47

here um so first things first this right

01:45 - 01:50

here is the fully native chat GPT

01:47 - 01:51

desktop app for Mac uh it does all the

01:50 - 01:54

things that you know we've come to

01:51 - 01:55

expect from our clients um but what I

01:54 - 01:57

really love about it is that being

01:55 - 01:59

native it's really lightweight doesn't

01:57 - 02:03

use a lot of resources it uh it lives

01:59 - 02:05

its own window and um it I'm able to use

02:03 - 02:07

it without having to context switch away

02:05 - 02:09

from what I'm already doing right so

02:07 - 02:11

we've got this keyboard shortcut option

02:09 - 02:13

space that makes it really fast to show

02:11 - 02:14

and hi chat GPT so it's it's always

02:13 - 02:16

there when you need

02:14 - 02:18

it this button right here is our entry

02:16 - 02:20

point for working with apps on your

02:18 - 02:22

computer and the way that I like to

02:20 - 02:25

think about this feature is that we all

02:22 - 02:27

copy and paste things into chat GPT

02:25 - 02:29

right all the time all the time all the

02:27 - 02:31

time um this feature makes that way

02:29 - 02:33

Smoother by when we are working with an

02:31 - 02:35

app on your computer we'll automatically

02:33 - 02:37

pull that context in for you so you just

02:35 - 02:40

focus on asking your question and we we

02:37 - 02:42

handle the rest so you might notice that

02:40 - 02:44

I've also got this warp uh console

02:42 - 02:47

window open as well it's currently

02:44 - 02:50

navigated to a repository that that I'm

02:47 - 02:51

getting up to speed on um and it might

02:50 - 02:53

seem kind of silly but I want to figure

02:51 - 02:55

out how many commits per day are

02:53 - 02:57

happening in this repo you know we talk

02:55 - 02:59

about velocity a lot here so I want to

02:57 - 03:00

see that for myself I have no idea how

02:59 - 03:01

to do that that though so I'm going to

03:00 - 03:04

use chat

03:01 - 03:05

GPT so when I click on this button I'm

03:04 - 03:07

going to see all of the apps that are

03:05 - 03:10

currently running on my computer the

03:07 - 03:12

chat GPT can work with uh important note

03:10 - 03:14

until you select one of these we will

03:12 - 03:16

never look at the contents of another

03:14 - 03:19

app so you are always fully in control

03:16 - 03:21

over what you're sharing with chat GPT

03:19 - 03:23

so to get started I'm going to click on

03:21 - 03:25

warp and at this time huge shout out to

03:23 - 03:27

the warp team for all of their help in

03:25 - 03:28

getting this going uh when we first

03:27 - 03:30

announced working with apps we did not

03:28 - 03:33

have support for Warp and it was I think

03:30 - 03:36

literally the first request was adding

03:33 - 03:38

support um so huge shout out to the team

03:36 - 03:41

they worked really hard to help us uh

03:38 - 03:43

get it ready for today so thank you so

03:41 - 03:46

I'm going to get started by saying uh

03:43 - 03:51

write a command to get the number of

03:46 - 03:53

commits per day over the past two months

03:51 - 03:54

and now I don't need to tell chat GPT

03:53 - 03:57

that I use get because it can tell from

03:54 - 03:58

warp that I do um and it's just going to

03:57 - 04:00

give me the command that I need so I'll

03:58 - 04:04

push this button to copy and paste it

04:00 - 04:05

into warp and I I think this looks right

04:04 - 04:07

yeah it's looks like the right

04:05 - 04:08

information but it's also kind of hard

04:07 - 04:11

to tell what we're looking at right yeah

04:08 - 04:12

I'm I'm a visual learner myself so um

04:11 - 04:14

normally what I would do is I'd figure

04:12 - 04:16

out how to get this into a spreadsheet

04:14 - 04:18

make a chart there and then find that

04:16 - 04:19

spreadsheet again in three years but um

04:18 - 04:23

instead I'm just going to uh I'm just

04:19 - 04:24

going to ask for one so make a bar graph

04:23 - 04:28

with all of the

04:24 - 04:30

results why not make it holiday themed

04:28 - 04:32

great idea

04:30 - 04:34

and uh awesome so this is going to show

04:32 - 04:35

off what I think is the coolest part

04:34 - 04:37

about working with apps which is that it

04:35 - 04:40

works with all of the other features and

04:37 - 04:42

all of the other models in chat GPT so

04:40 - 04:44

in this case uh for decided to use

04:42 - 04:46

Advanced Data analysis to crunch some

04:44 - 04:48

numbers and give me back a bar graph and

04:46 - 04:49

what that means if you really think

04:48 - 04:51

about it is that when we build features

04:49 - 04:53

like Advanced Data analysis and bring

04:51 - 04:55

them to chat GPT it's kind of like we're

04:53 - 04:58

bringing them to every app that chat GPT

04:55 - 05:00

works with yeah that's great so while it

04:58 - 05:01

thinks about this um do you want to talk

05:00 - 05:03

a little bit about what the model is

05:01 - 05:05

actually seen is it just what we see on

05:03 - 05:06

the screen or is it something more great

05:05 - 05:08

question so an easy way to do this would

05:06 - 05:11

be to just grab a screenshot and and let

05:08 - 05:13

Vision do the rest um but we actually

05:11 - 05:15

can reach into the application to grab

05:13 - 05:18

ascreen content as well and so uh these

05:15 - 05:21

results will contain everything here not

05:18 - 05:22

just what you see on screen yeah I was

05:21 - 05:24

thinking hard about this it might be the

05:22 - 05:27

the holiday themed

05:24 - 05:29

part okay perfect all right I mean this

05:27 - 05:31

is that looks pretty holiday themed to

05:29 - 05:32

me what do you think John uh it's

05:31 - 05:34

holiday themed I don't know if it's as

05:32 - 05:37

holiday themed as we are in these suits

05:34 - 05:39

but it's not bad literally nothing is as

05:37 - 05:41

holiday themed as you are yeah perfect

05:39 - 05:42

perfect but I'd say it's good enough so

05:41 - 05:44

I'm just going to download this and now

05:42 - 05:46

I can share it with a teammate but with

05:44 - 05:48

that I'm going to hand it back to John

05:46 - 05:50

to talk a bit more about programming

05:48 - 05:53

great so I think that the use case that

05:50 - 05:55

Justin showed is really important and

05:53 - 05:56

useful to be able to interact with a

05:55 - 06:00

terminal but I want to show what it's

05:56 - 06:03

like to interact with code in an IDE so

06:00 - 06:05

I have xcode open here which is my IDE

06:03 - 06:07

of choice and it's running a sample app

06:05 - 06:09

that is actually a little bit of a peak

06:07 - 06:11

behind the scenes into how this work

06:09 - 06:14

with apps feature works the sample app

06:11 - 06:17

uses the Mac OS accessibility apis to

06:14 - 06:19

look at xcode and tell us some things

06:17 - 06:21

about what's on the screen so it's

06:19 - 06:23

telling us that there's a text field

06:21 - 06:26

with these Dimensions it tells us that

06:23 - 06:29

it has 37 lines and we can go down and

06:26 - 06:30

check that 37 lines yep and it shows us

06:29 - 06:32

the content of the text field and we

06:30 - 06:34

actually use this to make the feature

06:32 - 06:36

right that's right yeah this is a useful

06:34 - 06:38

sample app for us for sure so this is

06:36 - 06:41

nice but it doesn't do any live updating

06:38 - 06:43

so I'm going to use chat GPT to help add

06:41 - 06:45

that feature I'm going to bring up the

06:43 - 06:47

chat bar with a very similar shortcut to

06:45 - 06:49

what Justin showed earlier but with a

06:47 - 06:52

slight change I'm going to use option

06:49 - 06:54

shift one and what that does is it

06:52 - 06:56

brings up the chat bar with xcode

06:54 - 06:59

automatically paired to it xcode being

06:56 - 07:01

the topmost app that is open that we

06:59 - 07:02

support with this feature it makes it

07:01 - 07:04

super quick to start working with an app

07:02 - 07:06

yeah it's great and you get this

07:04 - 07:10

immediate feedback that it sees xcode

07:06 - 07:12

here so these accessibility apis are a

07:10 - 07:14

little bit inscrutable definitely hard

07:12 - 07:17

to remember how to use and pretty

07:14 - 07:18

complicated actually um so I'm going to

07:17 - 07:22

use the model selector here and I'm

07:18 - 07:25

going to switch this to 01 01 is one of

07:22 - 07:27

our new newer models here at open Ai and

07:25 - 07:29

it does a great job thinking about these

07:27 - 07:32

difficult coding problems

07:29 - 07:34

um and I should mention as well that uh

07:32 - 07:36

this feature is also available with 01

07:34 - 07:40

Pro if you really want to throw the deep

07:36 - 07:41

end coding problems that model yeah all

07:40 - 07:45

right so let's give it a prompt here I'm

07:41 - 07:49

going to say add an

07:45 - 07:54

observer uh so if selection

07:49 - 07:57

changes load text areas is

07:54 - 07:59

called and we'll kick off this request

07:57 - 08:01

to the model so o1 is one of our Chain

07:59 - 08:03

of Thought models and you can see that

08:01 - 08:04

it's thinking about this issue it's

08:03 - 08:07

going to tell us some of the steps that

08:04 - 08:09

it goes through as it considers and wow

08:07 - 08:10

that was a pretty fast response from it

08:09 - 08:11

didn't have to think too hard on that

08:10 - 08:14

one I guess I guess not you got to give

08:11 - 08:16

it a harder problem next time yeah wow

08:14 - 08:18

all right so it's generating some code

08:16 - 08:22

and you know I have a fair amount of uh

08:18 - 08:23

of trust in 's code here so as soon as

08:22 - 08:26

it's finished generating I'm just going

08:23 - 08:29

to copy this into xcode and we're going

08:26 - 08:30

to run it and see what happens I don't I

08:29 - 08:33

don't see anything that could go wrong

08:30 - 08:36

with that no demos Poss yeah live demos

08:33 - 08:38

work 100% of the time it was one of the

08:36 - 08:40

rules of the universe all right so I've

08:38 - 08:43

copied that code and I'm pasting it

08:40 - 08:45

directly into xcode I'm going to take a

08:43 - 08:47

quick scroll through it to see if it's

08:45 - 08:49

finding any issues right now it's

08:47 - 08:53

looking pretty promising all right so

08:49 - 08:54

let's run this and see what happens you

08:53 - 08:56

know it would be really cool if you

08:54 - 08:58

didn't have to copy and paste that back

08:56 - 09:00

into xcode though that would be cool and

08:58 - 09:01

you know people have been suggesting

09:00 - 09:02

that should I build that you should

09:01 - 09:06

definitely build that all right PM

09:02 - 09:09

approved great all right so uh let's

09:06 - 09:13

it's running again uh let's take a look

09:09 - 09:15

if I select content oh no it didn't it

09:13 - 09:16

didn't work like we thought okay should

09:15 - 09:17

we should we give it one more shot yeah

09:16 - 09:20

why don't we why don't we ask yeah all

09:17 - 09:22

right so I think I'm actually just going

09:20 - 09:25

to go back to our previous state since I

09:22 - 09:30

don't have a specific error here uh

09:25 - 09:33

let's try to discard the changes

09:30 - 09:35

all right let's give it one more shot

09:33 - 09:37

yeah one more shot yeah and while it's

09:35 - 09:41

working we can uh talk about some more

09:37 - 09:45

of the features here okay add an

09:41 - 09:51

observer so if selection

09:45 - 09:53

changes load text areas is called again

09:51 - 09:55

all right maybe it didn't think hard

09:53 - 09:57

enough on that we'll try it again got

09:55 - 09:59

overconfident yeah well it thinks about

09:57 - 10:01

this I should mention that I'm using X

09:59 - 10:04

code uh like I said this is my IDE of

10:01 - 10:06

choice when working with swift but we do

10:04 - 10:09

support a whole bunch of other idees um

10:06 - 10:11

that means vs code uh the jet brains

10:09 - 10:14

ecosystem which includes Android studio

10:11 - 10:17

and Pie charm Ruby mine things like that

10:14 - 10:20

some really uh standby Mac apps like

10:17 - 10:22

textmate and BB edit so we've got a a

10:20 - 10:24

whole lot of different support here yeah

10:22 - 10:26

I'm I'm actually unreasonably excited

10:24 - 10:29

for mat lab support I would have totally

10:26 - 10:30

used that in college yeah mat lab is is

10:29 - 10:32

another exting one I think some students

10:30 - 10:35

are really going to find that useful

10:32 - 10:37

okay it's uh it is still generating some

10:35 - 10:40

code there it goes it's done I'm going

10:37 - 10:42

to use this copy button again and again

10:40 - 10:44

with full trust that everything is going

10:42 - 10:47

to work I'm going to now we know what

10:44 - 10:51

could go wrong though right yeah sure

10:47 - 10:55

all right so let's run this again and uh

10:51 - 10:56

see if we have slightly better luck okay

10:55 - 11:00

it's

10:56 - 11:03

running hey look at that if I things it

11:00 - 11:06

changes wow it's a holiday Miracle you

11:03 - 11:08

got the incantation to the demo Gods

11:06 - 11:09

right the second time yeah exactly

11:08 - 11:11

awesome so we've been talking a lot

11:09 - 11:14

about coding today right uh but another

11:11 - 11:16

thing that I love to use chat GPT for is

11:14 - 11:18

helping me with my writing and I know

11:16 - 11:20

I'm not alone here um and so that's why

11:18 - 11:23

today we're announcing support for three

11:20 - 11:25

new applications Apple notes notion and

11:23 - 11:26

quip uh we think this is going to open

11:25 - 11:28

up a brand new set of use cases for

11:26 - 11:31

working with apps and so we can't wait

11:28 - 11:33

to see what you all do with it uh with

11:31 - 11:35

that uh John Kevin you already know this

11:33 - 11:37

but for the rest of you uh I give

11:35 - 11:40

walking history tours in San Francisco

11:37 - 11:42

outside of outside of work um big

11:40 - 11:44

history buff San Francisco's got a great

11:42 - 11:46

story to tell and I'm actually working

11:44 - 11:47

on a brand new walking tour and so why

11:46 - 11:50

don't we try out this feature and help

11:47 - 11:53

me out with that let's do it great so

11:50 - 11:55

here I have a notion document open on my

11:53 - 11:57

computer I always write my tours in

11:55 - 11:59

notion and this is actually a real

11:57 - 12:02

walking tour that I'm working on so

11:59 - 12:04

I I hope you all find it interesting um

12:02 - 12:06

but I'm actually working on a news stop

12:04 - 12:09

for my favorite character in San

12:06 - 12:10

Francisco History Emperor Norton um I

12:09 - 12:12

know some high level talking points he

12:10 - 12:15

was the self-proclaimed emperor of the

12:12 - 12:18

United States and protector of

12:15 - 12:20

Mexico who lived in San Francisco in the

12:18 - 12:22

1800s and uh he even made his own

12:20 - 12:23

currency that was actually valid in the

12:22 - 12:24

city for a while that's something you

12:23 - 12:26

can just do yeah apparently apparently

12:24 - 12:27

you can and uh I think it's going to

12:26 - 12:30

make for a great tour stop but I'm a

12:27 - 12:32

little bit fuzzy on the details and so

12:30 - 12:34

I'm going to use chat GPT to help me out

12:32 - 12:35

uh one option would be to copy and paste

12:34 - 12:37

these bullets over and I think chat

12:35 - 12:39

would do a pretty good job at at going

12:37 - 12:41

with that but it would be helpful if it

12:39 - 12:43

had context of the entire document right

12:41 - 12:46

and so instead I'm going to have chat

12:43 - 12:47

GPT work directly with notion so I'm

12:46 - 12:51

going to hit option space to bring up

12:47 - 12:53

chat GPT have it work with notion and

12:51 - 12:55

I'm actually going to go ahead and just

12:53 - 12:57

um highlight this stop here so that the

12:55 - 12:59

model knows what to pay attention to and

12:57 - 13:01

so now we can see we're working with

12:59 - 13:03

notion in the walking tour document

13:01 - 13:06

focused on selected lines and I'm just

13:03 - 13:08

going to go ahead and say fill out these

13:06 - 13:11

talking points right don't need to be

13:08 - 13:13

any more specific than that um but one

13:11 - 13:15

thing that is really important is that

13:13 - 13:17

this is a walking tour right this is a

13:15 - 13:19

history tour things need to be factually

13:17 - 13:21

correct and so to help with that I'm

13:19 - 13:24

going to push this button to turn on

13:21 - 13:26

search and now to answer my question

13:24 - 13:27

chat GPT is going to search the web and

13:26 - 13:29

everything it tells me is going to be

13:27 - 13:32

grounded in citation

13:29 - 13:34

right and so I want to find out more

13:32 - 13:36

about I can go ahead and click the links

13:34 - 13:38

and you really start to see this awesome

13:36 - 13:40

interaction Loop pop up where chat GPT

13:38 - 13:43

is helping me with my research in the

13:40 - 13:46

context of the document I'm

13:43 - 13:48

writing so awesome this looks like all

13:46 - 13:50

of the stuff that I'm hoping to cover

13:48 - 13:51

and so it doesn't really sound like me

13:50 - 13:53

though this sounds like you know

13:51 - 13:56

official results and so I'm going to

13:53 - 13:59

turn off search and just say uh make it

13:56 - 14:04

match the style

13:59 - 14:06

of the rest of the stops keep it short

14:04 - 14:08

two paragraphs and now chat GPT is going

14:06 - 14:11

to go out and read the rest of my

14:08 - 14:12

document learn how how I talk and how

14:11 - 14:15

I've written the rest of these and do

14:12 - 14:16

its best to imitate that and so awesome

14:15 - 14:18

this this looks great let me introduce

14:16 - 14:20

you to one of San Francisco's most

14:18 - 14:22

beloved characters you'll have to come

14:20 - 14:23

to the tour to find out the rest sounds

14:22 - 14:26

like

14:23 - 14:28

you so I'm just gonna go ahead and

14:26 - 14:30

highlight this to copy and paste it back

14:28 - 14:31

into notion and of course you know I

14:30 - 14:33

want to iterate from here I'd want to

14:31 - 14:35

refine this um but that's just a quick

14:33 - 14:38

example of using chat GPT to work with

14:35 - 14:39

notion that's awesome I think that it's

14:38 - 14:41

really compelling to work with your

14:39 - 14:44

documents like this um not just code

14:41 - 14:48

like I showed before but your uh written

14:44 - 14:50

Pros um this is excellent but it's just

14:48 - 14:53

one way to work with the model this sort

14:50 - 14:56

of text in text out method and what I'd

14:53 - 14:58

like to introduce today is support to

14:56 - 15:00

use advanced voice mode with this

14:58 - 15:02

feature we're really excited about this

15:00 - 15:05

we think that this Paradigm is a really

15:02 - 15:07

interesting way to talk with chat GPT

15:05 - 15:09

and get insight into your documents or

15:07 - 15:11

your code for that matter all right so

15:09 - 15:13

we're going to show one of mine in a

15:11 - 15:15

second here um when I'm not at open aai

15:13 - 15:17

I spend a lot of time as a professional

15:15 - 15:19

saxophone player and I've got this

15:17 - 15:21

holiday party set list that I'd like

15:19 - 15:24

some input on I think chat GPT would be

15:21 - 15:26

great about this but again I'm I want to

15:24 - 15:27

use the voice to do this is this why

15:26 - 15:30

your holiday parties are all so Good by

15:27 - 15:32

the way no that's a different okay okay

15:30 - 15:36

is it the suits yeah the suits help for

15:32 - 15:38

sure yeah all right so uh yeah I want

15:36 - 15:42

someone special to really help out with

15:38 - 15:43

this um so let's bring up chat GPT again

15:42 - 15:47

with the shortcut that I mentioned

15:43 - 15:48

earlier which is option shift one and

15:47 - 15:51

that brings up our chat bar it's

15:48 - 15:53

automatically paired with notes but this

15:51 - 15:55

time instead of typing to the model I'm

15:53 - 15:57

going to use this new advanced voice

15:55 - 16:00

mode support so I'll use this icon in

15:57 - 16:02

the bottom right to bring in our special

16:00 - 16:02

guest

16:04 - 16:08

great hey Santa are you

16:09 - 16:15

there yes I'm here how can I help you

16:12 - 16:18

today Santa get it is so great to talk

16:15 - 16:20

to you listen I've got this holiday set

16:18 - 16:22

list that I'm working on and I've got a

16:20 - 16:24

couple songs that I'm just not sure

16:22 - 16:26

about this one I've got my love to keep

16:24 - 16:28

me warm I honestly don't really know

16:26 - 16:30

that song very well can you recommend

16:28 - 16:30

something else

16:32 - 16:37

how about replacing I've got my love to

16:34 - 16:41

keep me warm with Frosty the Snowman oh

16:37 - 16:45

that's that's a good idea let's see the

16:41 - 16:45

Snowman did I get that

16:45 - 16:50

right not

16:48 - 16:54

quite oh keep

16:50 - 16:55

going sure thing so is there anything

16:54 - 16:58

else you'd like to know about your

16:55 - 17:00

holiday set list I I think Santa was

16:58 - 17:02

bored with the copy editing but we can

17:00 - 17:04

move on uh let's talk about the last

17:02 - 17:06

song of the second set Al Lang Z I mean

17:04 - 17:08

that's a little bit dreary I'm hoping

17:06 - 17:10

for something more energetic to really

17:08 - 17:13

get you and the elves dancing what do

17:10 - 17:16

you think what should I replace it

17:13 - 17:20

with dick the Halls would be a perfect

17:16 - 17:23

replacement it's Jolly festive and sure

17:20 - 17:23

to get everyone

17:23 - 17:27

dancing Santa I think that's a great

17:25 - 17:30

idea I think that song's really going to

17:27 - 17:32

slay

17:30 - 17:34

I think you know interacting with the

17:32 - 17:36

The Voice model like this is such a

17:34 - 17:38

great way to talk to your documents and

17:36 - 17:41

get live feedback on what you're doing

17:38 - 17:43

totally and and uh good news uh right

17:41 - 17:45

before we stepped on this live stream

17:43 - 17:46

actually we pushed the button to ship

17:45 - 17:48

all of the features we just showed off

17:46 - 17:50

so um they should be available for you

17:48 - 17:52

on your computer just make sure that you

17:50 - 17:56

are updated to the latest chaty BT app

17:52 - 17:59

for Mac OS awesome so that's day 11 it's

17:56 - 18:00

about using the desktop apps to see

17:59 - 18:03

understand and automate the work that

18:00 - 18:05

you're doing with chat GPT uh this is

18:03 - 18:07

coming this is available today as of

18:05 - 18:09

what 20 minutes ago on our Mac desktop

18:07 - 18:12

apps coming soon to

18:09 - 18:13

Windows I love the desktop apps that

18:12 - 18:15

they're they're in the background when

18:13 - 18:17

you're doing work and then it's a

18:15 - 18:18

keyboard shortcut and they're right

18:17 - 18:21

there and able to help you do whatever

18:18 - 18:23

you're working on so that's day 11 we

18:21 - 18:25

have one day left we'll be coming to you

18:23 - 18:28

tomorrow morning day 12 we got something

18:25 - 18:30

super exciting so don't miss it yeah we

18:28 - 18:31

can't wait to get these features out to

18:30 - 18:33

you we're really excited but in the

18:31 - 18:36

meantime I've got to start practicing

18:33 - 18:39

this yeah that Santa recommended all

18:36 - 18:58

right let's see

18:39 - 19:01

[Music]

18:58 - 19:01

w

Enhancing Productivity with Chat GPT Desktop Apps

In this insightful discussion, Kevin Wheel, the lead product at openAI, alongside John Nastos and Justin Rushing, introduces the new features and capabilities of the Chat GPT desktop apps. The focus is on how these desktop apps, now available on Mac and soon on Windows, are revolutionizing the way individuals interact with technology to boost productivity. With a seamless transition from the web application to the desktop, Chat GPT is becoming more agentic, truly stepping beyond Q&A to assist users in a wide array of tasks.

Exploring the New Features

The fully native Chat GPT desktop app for Mac is highlighted for its light-weight, resource-efficient design that operates in its own window, making it easily accessible with just a keyboard shortcut. The ability to effortlessly work with applications on one's computer, pulling in context to streamline tasks, is a game-changer.

The discussion also delves into the integration of Chat GPT with code in IDEs like Xcode. By using advanced voice mode, the model makes coding suggestions, streamlining the coding process with live interactions. Features like Advanced Data Analysis and Chain of Thought models enhance the user experience, making Chat GPT a versatile tool for a broad range of use cases.

Leveraging Chat GPT for Writing and Research

The versatility of Chat GPT extends beyond coding to writing tasks. By introducing support for popular applications like Apple Notes, Notion, and Quip, users can now harness the power of Chat GPT for writing projects. The ability to work seamlessly within these applications, aided by the model's advanced capabilities, opens up new possibilities for content creation and research.

A fascinating demonstration involves leveraging Chat GPT within Notion to enhance a walking tour script. By requesting information in the document's context, conducting web searches for factual accuracy, and mimicking the user's style, Chat GPT proves to be a valuable writing companion.

Embracing Interactive Voice Mode

The introduction of advanced voice mode takes user interactions with Chat GPT to a whole new level. By engaging in a conversation with a special guest, such as Santa, users can receive live feedback and suggestions for their projects, adding a unique and engaging dimension to content creation.

In conclusion, the desktop apps for Chat GPT are set to redefine how individuals work, write, and research, offering a seamless and agentic experience that enhances productivity across various tasks. As these new features roll out, users can look forward to a more interactive and efficient workflow with Chat GPT.

As we anticipate the release of these exciting features, the possibilities for creativity and productivity are endless. So, gear up to experience the power of Chat GPT desktop apps and unlock a new realm of efficiency and innovation in your daily tasks.


In a world where technology continues to evolve, the integration of AI models like Chat GPT into desktop applications opens up a realm of possibilities for enhanced productivity and creativity. Whether coding, writing, or conducting research, the seamless interaction with Chat GPT promises a future where assistance is just a keystroke away.