Top 5 Generative AI Trends for Every Creator Alive

Or maybe you’re not even asking that question.

Maybe you’re heavily using AI in your creative endeavor already.

You guessed it!

Either way, it’s undeniable that AI is transforming industries across the board.

By 2030, generative AI is expected to be valued $78.6 billion.

Content creation, artistic ideas, fashion design…you name it. There is an AI tool or platform out there that is rocking the boat and making some big waves.

I’ll tell you about 5 of these big waves.

👆Image Created by DallE 3

1. Custom GPTs

OpenAI announced custom GPTs, also called “GPTs,” on November 2023 at their first-ever Developer Day conference in San Francisco.

Chances are you probably already use Chat GPT

Chat GPT is by far one of the most famous examples of Generative AI Trends.

So, think of GPTs like Chat GPT on steroids!

Are they any better than BARD or Bing Chat?

Bing Chat is Chat GPT 4, by the way. (Or at least it’s powered by it) …

And no, it’s not necessarily about being better.

Although, I’d have said it is better than BARD!

Unlike Regular Chatgpt, Custom GPTs Can Be Fine-Tuned to a Creator’s Specific Style or Thematic Requirements

You can upload videos, documents, and text that provide information not available online …

Whoa, that’s somewhat scary

Scary good, am I right?

Both 👀

Can anyone create them?

Anyone with a GPT-4 subscription can create their own custom GPT based on their project ideas.

There are already examples like Blog Expert and Creative Writing Coach that have built-in customization that gives you the basis of your writing before you even have to prompt the LLM.

Custom GPTs allow you to produce more relevant articles or scripts at speed.

Sweet!

Can I create one that knows my workflow

So I don’t have to repeat my conditioning prompts?

Exactly …

But you don’t always have to build the GPTs.

OpenAI will offer publicly available GPTs in a ‘GPT Store’ similar to how the App Store works.

So basically if you created a custom GPT, and you think it’s helpful to the community, you could publish it to the ‘GPT Store’ and potentially even earn an income based on the number of users.

2. Gen Video Production

Producing video content is now faster, cheaper, and simpler than ever.

Video generation combines CNNs (Convolutional Neural Networks), GANs (Generative Adversarial Networks), and transformer models to learn, compete, and capture temporal relationships in video frames for realistic synthesis.

Take a look:

👆 This is Descript for example.

One of the best tools out there along with Runway ML enables you to apply complex effects to videos with just a few clicks.

On Descript, you edit videos as if you’re editing a Word Doc. You have to see it to believe it.

It means access to high-quality production at a fraction of the time and price. 

For lack of a better term: That’s Dope! 😁

The result has been the decentralization of production power slowly but surely — with quality videos now emerging from bedrooms instead of studios.

Some of the already existing video editing tools are getting an AI boost as well.

Sensei, which is part of After Effects, can correct color, remove objects, and even auto-reframe shots for different aspect ratios.

Well in the future …

You’re gonna be able to produce videos entirely with AI, with absolutely no physical camera.

In the future?

You can do that right now with tools similar to what we mentioned above!

Doesn’t that pose some ethical concerns?

Yes, it does.

Although there are some grey areas.

Platforms like YouTube will require creators to inform viewers by labeling fully AI-generated content.

I think this is what matters the most when it comes to those regulations:

3. Generative Visual Arts

No one can deny the magic of generative art.

At the time of writing this, it is already way advanced!

For example, I just went to Bing Images, which uses DallE-3, and used this prompt:

“A moss-covered stone archway framing a view of a hidden valley, with a meandering stream and wildflowers in bloom” …

And this is what it came up with:

Looks real … but it may be slightly clunky right now similar to when Spacewar! came out in the 1960s, it shows the huge potential of the technology and how it will only exponentially improve at an incredible pace.

I must admit, it’s really fun to generate these images just from your imagination.

I got another one: “An octopus in the dark, award-winning photography and cinematic shot style, 4K”

I’ve no words …

Here is another one using Adobe Firefly.

Prompt: “Super detailed up photo of a Blue Passion Flower” …

That’s Insane! 👀

Algorithms are used to create unique artworks that can mimic various styles or generate entirely new forms of visual expression.

Okay but what can I use it for?

Research like Google’s “Imagen” transforms simple images into whatever you can (or can’t) imagine. 

AI is also changing the way art is curated and experienced. Museums are increasingly integrating AI across various aspects, enhancing visitor experiences and streamlining operations.

Refik Anadol’s Unsupervised is an AI-generated artwork that constantly changes. Exhibited in MoMA, the piece constantly evolves in response to its environment, creating art that “could exist but doesn’t” in the museum’s archive. 

Dirk Boll of Christie’s London highlights the value of AI in operations for data mining and analysis, greatly speeding up cataloging and preparation work that previously took days.

4. Multi-Modal Models

We can communicate with LLMs using text input and that’s cool and all but …

What if we wanted them to see stuff? Or hear what we say? Or talk back to us?

Well, that’s exactly what a Multi-modal Model refer to:

AI systems that can process and relate information from multiple modalities or formats, such as text, images, audio, and video.

To Accomplish Diverse AI Tasks, Multi-Modal Models Merge Multiple Uni-Modal Data Representations Into a Shared Feature Space.

Just like we have cognitive skills …

Yes, we humans as far as we know ourselves, are intelligent, it only makes sense if we wanted to build artificial intelligence that it also has similar if not more input nodes.

But this is kind of theoretical isn’t it?

It very much isn’t.

Google is the first to develop an actually usable Visual + Audio communicative AI capable of not just identifying but also reasoning about what it sees.

Take a look:

Meta, behind the scenes is also working on Multimodal generative AI systems.

In their blog post, they stated that they had already put this form of AI on commercial products such as the Ray-Ban Meta smart glasses.

5. Browsing and Coding V2

AI has changed browsing and the dissemination of information forever.

How so?

It powers recommendation systems that provide personalized suggestions and tailored search results based on each user’s interests, browsing history, and context.

One of my favorite tools for making a quick search (like factual stuff), or general knowledge is Perplexity.

It saves me those countless seconds wasted clicking away in my browsing track record 😅

Perplexity has become so famous now that it’s become a trend!

AI algorithms better understand the context, intent, and meaning behind searches and user journeys to provide more personalized and relevant results. Search becomes more of a continuous assistive experience versus one-off lookups.

But …

Ironically Google Search is also an AI algorithm, the difference is the content you get can be summarized by AI or written by Humans.

Personally, unless it’s a broad fact or a quick look-up, I’d have a hard time trusting AI-provided information because it could be obsolete or erroneous.

Here are some of the most typical things we see search engines and search-focused artificial intelligence platforms doing these days:

Okay, what about Coding?

Oh, yeah.

The elephant in the room!

Key players like GitHub’s Copilot (now powered by GPT-4) are pushing boundaries, offering capabilities like code explanation and automated commit messages.

Google’s Duet AI adds seamless integration with cloud and documentation, while JetBrains enters the arena with its own multi-model AI assistant.

These advances aim to empower programmers by suggesting code, deciphering complex codebases, and even generating tests and documentation.

The More You Rely on AI, the Less Skilled You Become at What the AI Does.

AI coding tools unleash enormous potential for faster growth via augmented intelligence. But at the end of the day, you gotta be aware of what the AI is doing.

As a programmer, maintain your skills strong, and use these tools sparingly rather than as a crutch, and give prudent human oversight.

FAQs