Transcripts
1. Introduction: Generative AI is here
and it's going to take every illustration or writing on crochet job away from us. Well, maybe not, but
it is time for us to prepare ourselves so that we
know what generative AI is, what it can do, and how we can use it in our creative work. Hello, I am a language
model developed by OpenAI. Welcome to generative AI
for creative professionals. My name is Jacob, my goal and I work as
a service designer where I spend most of my
time designing services, but to a very high degree, rely on machine learning
and AI to function. I also co-host the podcast designing the robot revolution, where we talk a lot
about how creatives can deal with thrive
in a world where generative AI gets more
and more capabilities that we associate
with creativity. In this class, we'll
explore generative AI and its impact on jobs in
the creative industries. We will dive into
the limitations and strengths of
these tools and we'll provide examples of
how it can be used to increase productivity
in simple tasks, such as writing
standardized e-mails and summarising texts
like news articles. We will then move into a little bit more
complex topics like how we can combine
the unique elements that only humans can
bring to the table, such as abstract creativity
and emotional intelligence. And then combine that with the generativity eyes abilities to achieve remarkable
boost in creative output. This class is perfect for
designers, developers, copywriters and art
directors who want to understand how to work
with generative AI and how to use it to
become more productive while still maintaining the
authenticity of their work. Throughout this class, we'll
cover the following topics. Basic understanding
of generativity, AI and its limitations. How to use AI to do repetitive
and boring tasks for you. And then how to utilize generative AI and be
creative process. Lastly, we will talk
about the future of AI in the creative industries
and how I think that our relationship with
AI might develop. In-between the lessons, I'll
give you some exercises that you can do to better understand
how all of this works. So are you ready to
learn how to harness the power of generative AI
in the creative process. Let's get started.
2. Class project: Let's dive into the projects and the exercises that we'll
be doing in this course. The exercises are designed to let you get a good
understanding of what is possible to do with generative AI and how
to get good results from using this AI as a partner in order to get
good creative results. The first exercise will
focus on getting access to the tools that we'll be
using throughout this class. I will be using
open-air ice chat, GTP, and a dolly, but feel free to experiment with any other tools that you find interesting. Most of it's applicable
to any tool. This exercise will guide
you through setting up your account and then
understanding the tools, basic features, the
following exercise. After that, we'll
focus on generating just simple social media
content using generative AI. You will learn how to create
prompts that can be used to generate texts that is
suitable for social media. Then we'll look into how
to refine that a little bit to make it more
authentic and human-like. The third exercise
will include creating your own unique
image by using one a tool to create the text prompt for the generative model
that will create the image. Lastly, we'll be
writing a short story. We'll do that by starting
out to outline everything that we want to
happen in the story. And then moving
forward to refine it until we have something
that we are happy with. Each exercise will be designed
to build on the previous. So by the end of the class, you'll have a basic, fairly good
understanding of how to use the generative AI in
the creative process. Then you of course,
need to practice. After each exercise, please
just upload your work, provide a short description
of how much editing and postprocessing work you had to put in to get the
result that you like. I can guarantee you that it's
much more than you think. I will make sure to
give feedback and engage in discussion
on the topic. Don't feel stressed about
doing all the exercises. If you feel like that
takes too much time. But I do encourage you to do one or two
of the exercise to familiarize yourself with
the technology and take the chance to engage with me and the other students
in this course. It's a really
valuable setting to just get some good feedback
on what you're doing.
3. Introducing Generative AI : Welcome to the first lesson of generative AI for
creative professionals. In this lesson, we'll
introduce you to generative AI and its
potential impact on you as a creative generative
AI refers to using artificial
intelligence that has the ability to
generate texts, images, or other forms of media. One popular example of generative VI is the
chat bar chart GDP, which can generate
endless text streams. However, with the rise
of this technology, there are some concerns about its impacts on jobs in
the creative industries. One concern is that generative
AI tools like chat GDP, could potentially
replace human creatives and tasks such as
writing and design. However, it's important to note that while generative AI is cool and it can be used to
automate specific tasks. It's very unlikely, at
least in the near term, to replace creative professionals
doing complex work. On top of that,
generative AI has the potential to, if used well, to assist creatives and
increase their productivity, we can multiply our
creative output in order to get to those benefits and avoid the dangers of generative AI. It is however important for us to understand its limitations. I will not really
go into a lot of technical detail in this course, except for this one word that I think we
should be aware of. The type of generative
AI that we are looking at for this course is
called a transformer. They are a type of AI
model that is trained to understand and generate
human-like texts. These models are trained on large quantities of human
written text or images. Then it transform it to
quote unquote new things. This type of ai has
the ability to combine words in a new way that feels
meaningful to us as humans. The most basic way that I can explain how it does
that is that it calculates what words are likely to follow
after another one. Given a certain context. I asked Chuck GDP to
complete this sentence. I am a and I, it gave me the response. I am a language model. And I generate human-like texts based on the input
provided to me. I then changed my question too. Can you complete this sentence as a medieval fisherman? I am. And I then I get
something very different. It responded, I am a
medieval fisherman and I catch fish in the sea using
nuts and fishing roles. I never said anything about nets or fishing rods or even the sea. But based on my question, it made the calculation that
I would like the sentence, I am a medieval
fishermen and I catch fish in the sea using
nuts and fishing. This ability to create meaning
for us based on context is really useful and
they've enables generational a lot
of material quickly. We can then use that material to create really cool stuff. However, because of
how these models work, we need to be careful of how we, as the creatives, think about what an AI
presents us with. If we ask the chat GDP or some
other AI tool a question, the answer we will get is
simply a regurgitation. That is the repetition
of information without analysis
or comprehension. And what we get back is a
statistical approximation of words that is likely
to make sense to us, the reader, the AI, does not understand the
context or meaning of what we have are asked or what the
answer it gave me means. Let me repeat that and I'll
put it up on the screen here. The AI does not understand
the context or meaning of what we asked or what the
answer it gave means. It is not intelligent, it's not sentience or in any way aware for you as the creative, this is a good thing. We can use our soft
human brains to bring valuable knowledge and
input into this process.
4. Exercise 1: In this exercise, you will
set up your account for the tools that
you're going to use throughout this course. These tools include
Open AI's chat, GTP and Dolly, which
I will be using. But you're welcome
to experiment with any other tool that
you find interesting. Go to the website, open ai.com and
create an account. You can pause the video
and go and try that out. I will be here
when you get back. I have also uploaded a list of the steps for each exercise
in the resource section. If you need any help, just reach out to me
now, I'll help you. Once you have created
your account, you can access and use
these tools such as chat, GTP, and it takes some time to familiarize yourself with
the features of the tools. You can experiment with
the tools by running some sample requests such
as asking the question, could you complete
this sentence as I am and I observed the results? I want you to try asking the
same question several times. Just change out the
context as I did ask it to be a knight
or a space marine. Another thing that you can
try is taking an article, read it through, and then ask
the tool to summarize it. This will give you a basic
idea of how competent this tool is that sort of understanding things
which is not. Share your experience in the project section by
creating a short video, uploading a screenshot,
or just writing in the project description about your experience
using these tools. In the next lesson,
we will dive into the specifics of generative
AI and provide examples of how it can increase
productivity by removing tasks we would rather
someone else do. Remember that generative AI is a tool like any other tool. It should be used to enhance human creativity
and productivity. And to do that, it's essential that you
understand the tools, capabilities, and limitations, and how to use it
most efficiently. As always, feel free to share
your thoughts, questions, and findings in the
class discussion. I'll provide feedback. Thank you for joining
me in this class. I'll look forward to working
with you in the next lesson.
5. Automating the Mundane: Lesson two, automating the
mundane with generative AI. A big part of any
creative job is in fact mundane and pretty
standardized tasks and not at all very creative. In this lesson, we
will discuss how to use generative
AI to help with or even automate these
non-creative tasks so that we can free up more time to
do the really good stuff. One of the main benefits of generative VI is the ability
to very quickly create a reasonable approximation
of something a human has written based on whatever material you
provided as contexts, e.g. generative AI can automate
tasks such as writing standardized e-mails or
summarizing notes from a meeting. While straightforward, these tasks often
requires some labor, time and attention to complete in a clear and
understandable way. Generative AI can assist you in completing them
more efficiently. I have an example
of something that I would like someone else to do. I got this email from
my builder robert. The subject line was
fence repair appointment. Hello. I hope this e-mail finds you. Well, I wanted to confirm
that I will be coming by on Monday to repair
the damage to your fence. If you have any
questions or concerns, please don't hesitate
to reach out. I look forward to
getting your fans back in good condition. Best. Robert the builder. I think this is great because
I want my fancy repaired, but I have some questions. When on Monday will he
come Can he estimate how long the job will take and what's this
going rate per hour? In these types of tasks, the AI can do a pretty good job. I'll just go into chart GDP. I'll paste my email
and my questions into the chat bot together
with the words respond to. Then what I get back is this subject pens
repair appointment. Hi, Robert. Thank you for confirming
the appointment for Monday. Can you let me know what time on Monday you will
be calming also. Could you give me an estimate on how long the job will take? What is your hourly rate? Thank you. And then my name. In this example, you are
defining what should go into the e-mail and the
way I just spits out a likelihood
response based on that, it might look like it's
not that much help. But for me, a big part of
the work when responding to this type of e-mail
is structuring it. Responding in a way that
ensures that it has a clear tone and that it's correct in the formulation
using a sharp point, you'll still need to
check the email for errors and other problems
before sending it. And you're really not forget to actually add your
name to the bottom. But for the most part, the AI does a really solid job. But when it does this
type of task that doesn't really require real emotion. By using generative AI to automate non-creative
tasks like this, we can free up more time for the important things that
require more of a human touch. In this lesson, we discussed
how generative AI can increase productivity by
automating non-creative tasks. Remember though that
generatively I should be used to assist in these tasks rather than replace
the human labor. In the next lesson, we will dive deeper into
the specific use cases of generative AI in the
creative process and provide examples of
how it can be used to improve workflow and
increase productivity. Thank you for joining
me in this lesson. I'm looking forward to
seeing you in the next.
6. Exercise 2: In this exercise,
we'll be focused on generating content
for social media. We're not aiming for
anything crazy here. We just want to create a
shout out to something like a book or a movie that we like because we think our friends would like it too when
we want them to see it. This will be quite a
bit more complicated than the example of the
email in the last lesson. So expect to spend some
time making adjustments to prompt and proofreading
the content that comes out. First. Find the description
of something you like, like the summary of the
back of a good book, the official
description of a movie. I'm going to use the
description of this course and then write a list
of how the things, in this case my
course made me react. I'm going to go for it got me relaxed about my
future as a creative. I now understand
generative AI better. And I got excited to use
AI in my creative work. Okay, so we write prompt,
something like this. I have watched Red
taking part or whatever, the title of the thing
it made me feel. And then the list
that we just created. Can you write me a
Facebook post about it? Here's a summary of the thing, and then that is my prompt. Here you can see an example
where I've just added all of this stuff into the chat
GTP, generative AI. So basically, I've added, I've, I've taken part of the
Skillshare course, gender to be a for
creative professionals. I've added the lists, and then I ask the AI to
write a Facebook post. And then I've added the summary that I have in my
course description. Just going to press Enter here. You can see that it starts to
generate else. Be this up. There we go. My result just finished taking part
in the Skillshare course, generative AI for
creative professionals. And I cannot express
enough how much I learned. The class is learned by Jacob marginal and experienced
service designer and co-host of designing the
robot revolution podcast. And it's explored the impacts of generative AI on the
creative industries. I gained a better
understanding of the limitations and the
strings of these tools, and I got excited
to integrate it into my own creative work. This class is a must take for any designer, developer,
or copywriter, or even art director
who wants to increase their productivity while still maintaining the
authenticity of their work. Give this course
to try and boost your creative output
with the help of AI, the hashtag, AI Skillshare,
whatever comes next. Maybe this is not perfect, but it doesn't have to be. It's a good start.
Probably the tone for me. I would take it down a little bit just to make
it sound like me. I think some people would get worried if I just posted this. This is a bit over the top, but it looks like perfect
starting point for something that would take a good
while to do on my own. So now I would like for you to try and do the same
thing that I just did. I'll see you back here in a bit. How did this work for you? Did you get a good result? Would you ever consider
using something like this as a starting point, at least for a
social media post. Did you get something
that you could just post? Share your experience in the project section by
creating a short video, just uploading a screenshot or writing in the project
description about your experience using AI tools to make a social media post.
7. Enhance your Creativity with AI : In this lesson, we will
finally be discussing how to use generative AI to
enhance creativity. In the last session, we
discussed how we could free up time and energy to put into
that creative process. However, there are also ways to use generative AI in the
creative work directly. Generative AI cannot
replace human creativity, but we can use the technology
and the tools that spring from it to assist and enhance
the creative process, helping us create more and hopefully better
creative content. I want to reiterate a few
points from the first lesson. I will struggle with
tasks that require common sense or an understanding of the real-world contexts. An AI does not possess
qualities such as abstract thinking or creativity and
emotional intelligence. Most importantly, an AI has no sense of self or
awareness of others. For simple tasks that doesn't require creativity
or understanding. Generative AI, as we
explored in less than two, does a decent job. For more complex tasks, the relationship
between humans and generative AI becomes
much more interesting. All the qualities that AI is lacking are things that we humans are
really, really good. We have the ability to develop
the intent and the ideas for creative content
by injecting our human elements
such as agency, perspective, and emotions
throughout the writing process, we can make something
really good. Conversely, in AI, has access to a world of information is
not always up to date, but it's always
more than we have. The AI also has unlimited patience
and an ability to take impulses from us and run with it in a
way that would make any improvisation theater
actor green with envy. These characteristics makes it a fantastic creative partner. Always encouraging, always helpful, but
thoroughly unimaginative. We can use an AI to create pretty good drafts
that we can refine. I have found two
ways of creating these drafts for more
engaged content. One is close to what
you try it out in the last session with a social media post by
using a very solid prompt. A full body of texts
that gives the generative AI more contexts. This requires you to have
a lot of content already. Something comprehensive
to start from. The other way is to engage in a discussion with
the generative AI to slowly sculpt the output we're
looking for by using consecutive
intentional prompts to guide the generative AI, we can create something
pretty compelling and that aligns with
our intent and ideas. By then rewriting the resulting
outputs quite heavily, you can create texts that
is authentic and human. Usually, when I want to
create this type of content, I combine these two generation
methods and then get to work on the heavy editing and reworking required
to make it my own. Usually, I am happy
with the results. In this lesson, we've
discussed how to use the generative AI to
enhance creativity by developing the idea and the intent for the
content using prompts, guy, the AI model, then injecting human
elements such as agency and perspective
and emotion, you can ensure that the output
is authentic and human.
8. Excercise 3: For this exercise, you will
generate a visual to go with the social
media content you created in the
previous exercise. This can be very tricky to get the AI to create a
satisfactory image. It takes practice just to get it the way you want it to look. First, review the tax that you created in the
previous lesson. Think about the key
elements that you want to represent in the image
that you want to create. Write those down in a list. Now, we will use a
text generator like chat GDB to create prompt
for the image generator, TTP, and ask it to
generate a prompt for an image generator based on
your social media prompt. Some key elements that you
want the image to contain. I'm going to ask
it to be cartoony, to contain one
person and I guess one robot celebrating
academic achievement. Finally, use Dali to
create your image. Take the chat GDP output and use it as an input into
the image generator. Start with the prompt that
you got and keep refining it until you get the result
you are looking for. This is not easy, especially
with people in the image. But if you can practice, you can make the computer
do some really cool things. Alright, It's just see
you can get a sense of how I work on this
with image generators. They usually I start
out with charging to p or a text generator
to create a prom, I wanted to image generation
problem for an image or personal robots celebrating
academic achievement. And how that prompt got through HTTP, Right? I think this is
pretty good actually. Alright, so I'm going
to just manually edit this to
something like that. Just take away some of
that stuff and then this here and see what we come up with. I mean, this is kinda
quirky, a little bit fun. Maybe I want it in a, another thiol. So takeaway. I'm just going to add that. Carl Larson. Alright, uh, Carlos,
I'm painting of a robot and a
person celebrating academic achievement is quite different from the first result. I think this one is
actually lovely. So I'm going to use
that one for my post. And I always keep this. I think it's important
to show that this is by no means original
art painted by me. I think it's important to
be transparent of that and especially to keep a record
of what you have done. The level of customization
possible is only limited by your skills,
tools, and curiosity. You can learn a ton of great photo editing and illustration techniques
here on skill, share your experience
in the project section by creating a short video,
uploading a screenshot, or writing in the project
description about your experience using the
AI tool to make an image. In the next lesson,
we'll discuss how generative AI might affect
creative industries. The people working within them.
9. AI and the Creative Industries: The future of AI in the
creative industries. We've seen in the
previous lessons that generative AI
has the potential to assist in the creative process
and increase productivity. However, it's
essential to consider the potential for AI to disrupt
existing ways of working. One way that this new technology might play out for
creatives is that it becomes so easy and
cheap to create content that we are simply
out-competed by computers. And in some ways, I can
believe this reasoning, but only for content that doesn't require
that much thought. Simple e-mails and
simple social media content meant to promote
one thing or another, is likely candidates,
at least in my mind, to be taken over by AIM. And that very soon, if it hasn't already. The counterargument to this is that the eyeballs
and ears that we have to consume this tsunami
of content will be limited. They are the same as
before, and therefore, these new capabilities will make people demand
high-quality content. High-quality content
that can only be created via the collaboration between
computers and humans. Creatives need to develop skills such as creative prompting, editing, AI outputs,
and fact checking. Highly valued creatives
will be experts at interacting
effectively with AI. Additionally, abstract
creativity and emotional intelligence will
become even more valuable. They already are today. Creatives who have a
deep understanding of their field will be
better equipped to use generative AI in the way that enhances their work rather
than replacing them. While AI will play a role
in the creative industries, there will still be
a need for skilled, knowledgeable or
humans to control and drive that creative process. It is important for creatives to understand the
capabilities and limitations of generative AI and know how to use it to improve their workflow and
increase productivity. All right, This
concludes our class on generative AI for
creative professionals. I hope that you've enjoyed this class and that you've
learned a lot from it. Remember that generative AI
is a tool like any tool. It should be used in a way that enhances human creativity
and productivity. I would be so appreciative
if you could take the time to leave a review for me after completing the class. Thank you for joining me. Keep being creative and have
a curious, wonderful day.
10. BONUS: Image captioning using Google Bard: I received this
question from Samantha. I thought that this is really an interesting
question and it leads to us looking at some new interesting
developments on Google Bard, which wasn't really a thing publicly available when
I made the course. So I thought I'll just make an extra little lesson here and see what you
think about that. If you want to. More questions to this course, send me in the discussion
forum here on the page. I will see if I can answer them in text or if all
just make a video. If I think that it might be something that is interesting to a lot of people that can make another one
of these videos. So here's the question
from Samantha. Hi Jacob, is there
an AI app where I can upload my own
drawings to and have the AI generate short social media captions for me plus potential hashtags. My initial thought was
that, no, Not really. I haven't seen anything
that it works great. I've been fiddling with Google's AI image
recognition API, but it's very clunky and it's very hard to do anything
useful with that. But then I thought,
hey, wait a minute, didn't hear something
about Google. Bard being able to do something
like this because you can't upload images to ChatGPT
and Abbott, look at it. But you can actually do
that with Google Bard. So I thought, okay, let's see
what we can do with that. And I am by no means a mind. My drawing skills are
not excessively good, but we can try something here. And you see here is me
playing around with some podcasting advice might be interesting for you
guys. I don't know. What you can do is you
can drag an image here. This is my doodling. I've drawn for a talk. I've made a drawing of me on
a horse as a self portrait. Can you analyze this image? You can see there me on a horse. Let's see what Bard
can say about this and see if we can generate
anything interesting. So, sure, the image she sent me as a child's drawing
of a man writing a donkey. The man is a stick figure. Drawing is very simple, but it captures the essence
of a man writing a donkey. Miles posts is relaxed. I think this example is a great example of a
child's imagination. Here are some specific things. Okay. So cool. We have now a description
of the image. It's not what I, what I
was trying to achieve. I was trying to go for Let's see if I can
open this up here. Yeah, you can see to me this is obviously a self-portrait
of me on a horse. I think everyone
can agree on that, but we can now take this, see if we can have it
generate some social media. Can you write about this image? A self for make light of Bard thing is made by a child. Also generate hashtags. And we can see what
it comes up with. Sure, Here's social media posts about the image as
a self portrait, making light on Bard thinking that it was
made by a child. I'm not sure what Bard things
about my drawing skills, but I'm pretty proud
of this self portrait might be a large language. Alright, fair enough. There's some editing
needed if things that it should talk from the
perspective of Bard. But that's not really
what I'm going for. I guess we can just fix that. It's from the perspective
of the artist. Not Bard, the model. Sure. Okay. I'm definitely not a
child no less true. Sure. Okay. So you can see here, but let's go back to the
question from Samantha. Hi Jacob, is there
an app where I can upload my drawings
and how the AI generate a short
social media caption for me plus potential hashtags? Yes, you can actually do
that with Google Bard, but you run the risk and this is true for everything
that you do using AI, you run the risk of
losing your intent. You know much better
than Google Bard does. What you're trying to
say with your image. You know much more
what it is about your, your experience that led
you to create this image. That being said, if
you do what I did here and you modify the
output and from Google Bard, I think you can get away
with a lot using this tool. I think that it's
really useful for the scenario that
you outlines Amanda. And I think we can experiment with this and get
a lot of cool new inputs. But yeah, I'm curious to know what you think about
it when you've tried it. And specifically, if anyone
wants to try this on some more competent
Art and upload the Art itself and the social
media that comes out of it. I think you can also try
to work a little bit with ChatGPT and Bard together because it seems
to me even though Chat Gpt doesn't have this ability to
upload images to it. It's better creating
the social media posts. So maybe you can take
the output from Bard, put that into ChatGPT, and then you can output
something that is pretty good. I'm going to play more with
this I haven't previously, but I think it's really cool. If you like this sort of extra added on little example
of something special, please send me
more questions and maybe I can try something out. I can create some
examples out from your questions and that
will be really FUN for me and potentially pretty
engaging for you as well. Also, if you liked this course, please give it a
review that really helps fund to make
these courses. And I'm happy to get
your feedback and input. Alright, have a wonderful day. Thank you very much. My name is Jacob Magnell
and this has been a special lesson on Bard creating a image
caption and some hashtags. Have a great evening.