Transcripts
1. INTRODUCTION: The world has been abuzz
with generated AI. Everywhere you look, there are new pieces of EHR
technology that had been embedded into
existing platforms and all new platforms. In this class, we're
going to look at Photoshop and how they've embedded regenerative AI
into an entirely new way of editing existing photos or creating new images altogether. This class is intended for
beginners of Photoshop, for people who want to
get their feet wet. And learning the
basic building blocks of how to use photoshops, new generative AI functions in creating images or an
editing existing images. It's a really simple class. It's quick, and
hopefully it's useful to white you do whether
you're a new user, Photoshop, or someone who's played around with the
different versions of Photoshop in the past and wanted to try this
new Beta version. This class is for you. It should be lots of fun. I'm looking forward
to seeing you here.
2. WHERE TO FIND THE RIGHT VERSION: One of the first questions is, you're going to have
when looking for all of these new generative
AI functions in Photoshop is where to look. It's not as intuitive
as you'd think. Rather than looking in Photoshop proper and looking at the
Photoshop app being up-to-date. You have to go into
your Creative Cloud and you have to start looking for the beta versions. So it's not another photo. It's not under any of that, or graphic design, or video, or illustration, or 3D and AR. But under beta apps, if you click on Photoshop, Beta, That's where the
generative fill functions are available inside Photoshop. So you have to download a specific beta
version of Photoshop. Maybe subscribed to
Adobe Creative Cloud. It should be available under beta apps inside of your
Creative Cloud account. You open it, you add
it to your account, and you add it to your computer, and you're ready to go. It's the very first
step is go to Creative Cloud and
find the beta apps. And that's where you'll find
this version of Photoshop
3. HOW TO REMOVE ELEMENTS: So the Beta version of
Photoshop is almost identical to the version of Photoshop you're
already familiar with. There's one really
interesting difference. It allows you to create images and edit images
using generative AI. And that's something that
really is mind-blowing. We're going to look
at several examples of how you use this, or how do you can use this? And the possibilities
really are endless. Let's start off by
looking at a photograph. Here. Here's a picture
I've taken before. It's a image right
behind my house. It's completely,
it's an open field. There's nothing edited in
this image whatsoever. And what you can do with generative AI is
pretty incredible. So you could always edit
these images in Photoshop. You could painstakingly use
tools like the Brush tool to, or the eyedropper tool
to change colors, to change pixel by pixel. But what's interesting
here is that you can edit things just by using text. Let's use the lasso tool. Some of these cars right here, I don't like these cars. Let me circle them. And then generative
fill and I'll put, remove and let the AI work. If all goes well, it should get rid of these cars. There they go, they're gone. And again, even the way that it creates the image behind
where those cars were. Now there are no
more cars there. It's pretty incredible. Let me try another one.
Let me circled this bus. And again, there's no additional
coding needed to remove. And it knows that
I'm talking about the bus and they can tell what to remove
and how to remove it. And this kind of
thing is possible. There it is. It's gone. Remove, gone. So removing items that
you do not want in an image is one way you
can utilize this sulfur. Let's take away these cars. Remove it again
if all goes well, it will just remove them. And this is trial one. There's no I'm not
trying this and advance. They're just gone. So there were four or
five cars that are on that parking lot and
now they're are not. And again, this was
possible before, but you would have had to painstakingly zoom
in on the image, pull out every pixel, and then we create
the background. And the AI is really good at figuring out what you're talking about and then learning how to edit the picture accordingly. So removing or
retouching photographs, which used to be an
extremely long process, done quite easily using
this generative AI. And it's as simple
as circling and writing like this pole here. If I wanted this pole removed. Steve, it does this remove. And you can already, I hope, see the advantages of
how you could again, retouch and edit photographs there that's done right away. And you can look at the different options
and how to remove it. So there it is created. It will give you three
different options and how to create
the background. But there, that
image right there, that looks pretty good. And even if you zoom in on it, even if you were
to zoom into 200%. Like, it's hard to tell that
those cars wherever there. And again, that's done
all doing the using the generative process,
which is incredible. It just removes things by just writing
them word for word. So that's one thing
we will look at. Let's go back to 30%. There's the original
photograph the way I had it size before. In the next class. The next part of this class, we're going to look
at how you can create things that don't
exist using the AI
4. ADDING ELEMENTS: In this class, we're going to
look at how you can create elements and objects that don't exist just by
writing them in. So again, we're using
the same image of this open field that we
used in the last class, in the last section
of this lesson, in the last lecture, in the last lesson
of this class. And that we've already
taken these cars out. There are cars here before. It doesn't matter if
they were here or not. But now we're going
to see what we can create in this image. You could, you could cut
and paste things before. But here using
generative AI by itself, you can edit this image
and it will do it with light and texture and
make it look pretty, pretty good just using typing. So we'll use the
lasso tool again, that's a really good way
of selecting an area. So let's say in this image, let's say I wanted to put
a skyline, a city skyline. That's the wrong one. I'm going to use this, this
image, this tool there. We'll draw this thing in here. Okay? Drew a circle there. Any shape you want
but generative fill, Let's put city's skyline
and press Enter. Wait a couple of seconds. And what should happen is
the AI will figure out a, an image or an edit to your image that will meet the criteria
that you defined. There we go, city skyline. Let's see. Let's say that. I mean, that doesn't exist. It's pretty incredible. And all of the
images that creates for you are images that are
already royalty cleared. So you can use them in your, in your final deliverable. But let's say this guy, I don't like it gives you three options. So three separate options for every generative piece
that you're talking about. And in this, that
looks pretty good. And if you zoom in, like it creates it in a way that is quite, quite believable. And again, I can edit even
the reflections that are used are pretty appropriate
tool we're talking about. So that's a pretty
incredible piece of sulfur. And then even a
piece of technology. Even within these, even within the image
that you're creating, you can say a shorter tower. Let's see if it will
shorten that tower. Okay, so that gives me
what I wanted it to be. And that makes a crane. So it doesn't always hit it
the way you want to do it. In this case, I'm going
to take that out. But you play around with things and you can draw
them in and out. But let's say I wanted
to zoom back out. And I wanted to create other
elements in this image. That would be
interesting for me. Let's say I wanted to create
a piece of water right here. So let's go to the Lasso Tool. Draw in an area. Will generative fill, put water. And we'll see what
it comes back with. And by playing around with two mentions and
playing around with the way you draw
areas of interests. So that doesn't look as
realistic as I want it. There's some analysis. These are all waters, but none of them really hit the element that I want to use. So let me try redrawing the
field, redrawing the area. Let's try there. Or you can even make a
pond and say a pond. And the AI has the ability to understand in real language
what you're trying to do. All of this was possible before, but you'd have to find the
right image, place it in. That looks decent. And again, you can play around
with things until you get the right the right fit. That's not what I want yet. So let me play around. Try to draw it here. Pond. And hopefully what we'll
do here is it will adapt the image around that as well to make it look
more realistic. Okay?
5. EXPANDING THE CANVAS: Sometimes the images you take, sometimes the pictures you have, or things you want to make bigger or have
different dimensions. And what generative AI does is allows you
to expand what is shown in an image to an extent that is kind
of beyond comprehension. So again, using the same image that we've been using so far. So this image of a
plain blank field, I've taken out the cars. But what you can do here is
you go 12345 tools down. So this is the crop tool, expands the edge of an image. And let's expand this a bit. So let's put it from here to here to there. Okay, so that's expanded image. Then what you would
do is you would select the original or the
thing you want to expand it. You inverse the selection. So two over you
invert the selection. So now it's selecting
that area in the border, outside that white border. And now I can put up top, now, I can generatively fill
and we'll leave it blank. We'll say, just figure it out. Now it's taking that
original image that was in the frame so that,
that vertical image. And it should expand
it by what it thinks. If the pattern continues. Here we go. This is incredible because
none of this stuff exists. And it gives you a
couple of options too. So a couple of options of
how it thinks it might go. In one of the options, the building on
the right side of the frame is built
up on another. There's a smaller
building just on another. There's kind of an open field
6. CORRECTING BLEMISHES: But sometimes the generative
AI is not perfect and you get these kind of aberrations that you
want to smooth out. So in selecting the
expanded Canvas, you got this area here. That is, It's a
weird white line. And so what we'll
do is we'll draw a, will draw over that. We'll select it and
we'll generatively fill that within the
lines of selection. And so we'll let the
AI and figure out what are some of the patterns
of fallen there. And that should smooth out that area that looks
like a vertical line. He did it. Okay. That's better. Still is missing a piece here. So let's smooth out
the whole thing. Let's re-select this. And let's smooth out this area. And it takes all the data
and we're trying to get that one little piece
to smooth out a bit. And there it is, smooth out. That looks pretty good. And even that horizontal line across the horizon is
consistent throughout. There's another white line here. So let's take this line
and let's smooth it out. And also let's get
that vertical line on the horizontal line out too. So we'll do a degenerative
fill this whole area. And what it should
do if all goes well, is it should smooth
out the whole sky. And this was very
possible beforehand. But with generative technology,
with AI technologies, you can do this so
quickly just by using text with no
additional skills. Now when I take this image, I can draw in other elements
that I want to see. So going back to that
initial example, I can dry up for
city skyline here. And much of this
image never existed. And I actually use very limited, almost no visual graphics skill to create. What I'm saying. Hey, that's weird. I'm not sure where that came
from, but that's kinda cool. Looks kinda like Philadelphia,
kinda like New York. I'm, I'm gonna go
for this first one. Hutton as cool as well. Let's make this a parking lot. Make this whole area
into a parking. And let's see what
comes back with. Here it goes, it's
going to create an image of what it
thinks is a perfume lot. That's fine. Again, it's interesting to play around
with all these potentials. So this is what you can
do with this software. Where it gets really amazing is when you work with people. In the next class, we'll look at how to use this on humans
7. HUMAN SUBJECTS: When you're working
with images of landscapes or of settings,
that's one thing. But when you're
working with people, it can be very
difficult to trust AI to maneuver the data, because there's a trust factor of wondering whether
all the real or not. So let's take a picture. This is an image of me sitting
right here, right now. And what you can see
is with generative AI, you can add this picture just
as easily as other things. You can do it pretty amazingly. Let's start off with, let's say on my arm. Circle my arm here. Let's add a tattoo. And what I find interesting
about the way that it does this is it's able to
figure out the position, the amount of light. And it can build in
using all this data. Sudden, it looks
relatively reasonable. I don't know what
these are tattoos up, but that's kinda cool. It's pretty simple. It's printer attached to, either attached to them and
say, oh, tattoo. And it even gets
the angles correct. And this is just by
writing in one word. So what's cool
about their lungs? That looks relatively real. You can noodle around with
it and play with that. Let's add. These are parlor tricks, but you can see where this
could become useful. Beard. Don't know how great those
particular periods are, but you can play around
with it until you get one that looks about right? Remember these beards? Let's change the
shirt I'm wearing. So you can really play
with all these factors. Crusher, he's able to figure out what
position my body isn't. Yeah, it is. That's amazing because
that doesn't exist at all. I can draw on
glasses onto myself. Sunglasses. And this is with no, I keep saying this
over and over again. But the Sioux with no additional graphic
design knowledge, this is just typing in words and watching
how they appear. I think the frame which I drew the glasses
wasn't so good there. But nonetheless, you
also can utilize things like the object select tool. And now I can do the inverse of this. And I can put neon, Tokyo, whatever that means. Let's see, plays and keeps the right
dimensions and so on. What's amazed me about
the software is that it's incredibly good at making creative possibilities
come to life with limited knowledge of the how-to is the maximum
light and color. And that's where I think this
is incredibly interesting. I've only started playing
around with this, and it's unbelievable
the possibilities. I actually find that much of the Photoshop skills that I had learned in the past have been entirely replaced by this
new generative AI input. It's worth trying for yourself. In the next class, we're going to sum this all
up and we're going to build a lesson that allows you to play with an image
that you already have and to adapt it
using generative AI
8. CREATE A PROJECT: Last lesson of this course, I'd like you to go into your
computer and find an image. It doesn't matter of what. It can be. Any, anything on your computer, an image that you know, that's really simple to use. And I'd like you
to adapt it using generative fill on Photoshop
beta in three ways. One is I want you to expand
the canvas and fill it in. However you may like to. I'd like you to remove
elements from the photo. And three, I'd like you to add elements adaptive in some way. It doesn't matter how exactly, but I want you to play
with some of these tools only using generative
fill to see how powerful this can be in your daily
workflow on Photoshop. It's best if you're a novice
in Photoshop and don't use any other tools other than
the gender to fill aspect. On Photoshop Beta. I'm interested to see
what you come back with in sharing this shear, the original image, and then
share how you've changed it. Hi, You can put, you can upload these into the
project folder below. And thank you very much for taking this quick
introduction to Photoshop beta and to using
generative fill on Photoshop