From: gregisenberg
Brett from Designjoy, a one-man designer, utilizes AI to produce exceptional designs and generate millions of dollars annually [00:00:00]. These AI tools are powerful and simplify the process of creating beautiful graphics [00:00:35]. The core idea is that AI creation is surprisingly easy, even for non-designers [00:00:39].
Accessibility for Non-Designers
Individuals without prior design knowledge can create significant value using AI tools [00:02:57]. While good taste helps, there is no inherent skill barrier for non-creatives to produce impressive results [00:03:21]. Most content generated by Brett is not specifically for designers but for anyone with a Midjourney or ChatGPT subscription [00:03:08].
Types of AI Generation
The current phase of generative AI includes various realms such as image generation and video generation [00:03:33].
Key AI Tools and Their Applications
Higgsfield
Higgsfield is an ideal starting point for those with no prior experience in AI creation [00:04:11]. It offers templates and presets that simplify the process of generating content, making it easy to create ads for products [00:04:19]. The tool simplifies prompt engineering by providing preset actions and effects, addressing the difficulty of “talking” to AI [00:04:41]. It can even consider product branding and text, which was historically a “problem child” for AI generation [00:05:46]. Users don’t need to prompt it, as it relies on presets [00:06:23].
Midjourney
Midjourney is primarily used for generating images [00:07:17]. It excels at artistic styling, allowing users to create visually unique images by referencing specific styles [00:08:13]. Users provide a prompt and can incorporate image prompts (to influence content) and style references (to influence aesthetics like colors, shadows, and depth of field) [00:09:50]. Mixing different styles can lead to surprising generations [00:10:21].
ChatGPT
ChatGPT is crucial for crafting effective prompts for AI models [00:10:36]. It removes the need for advanced writing skills to describe desired outputs, a task that was once difficult before its capabilities improved [00:10:40]. ChatGPT can generate multiple prompt options from a simple request, adding rich detail that enhances the final image [00:11:38]. It can even dream up entire collections of images based on a theme, allowing for varied scenes without explicit instructions for each [00:13:18].
Runway
Runway is a tool used to add motion to static images, transforming them into videos [00:14:45]. Users can upload an image and provide a prompt (often generated by ChatGPT) to dictate the movement within the video [00:16:51]. Runway can produce high-quality video output and offers options to upscale videos to 4K resolution [00:25:25]. It understands specific instructions like keeping the camera still [00:20:16].
Cling
Cling is another image to video tool that allows users to set start and end frames, enabling seamless video loops [00:17:26]. It also provides AI-generated sound options that are contextually relevant to the video [00:17:53].
Other Tools
- Magnific Upscaler: Used to enhance generated images, adding super-rich details [00:14:55].
- Photoshop: Can be used for generative fill after initial AI generation [00:14:48].
- Cap Cut: A user-friendly video editing platform for adding text layers to generated videos [00:26:50].
Workflow for AI-Generated Visuals
A common workflow involves:
- Idea Conception: Start with a basic idea of what you want [00:14:01].
- Prompt Generation: Use ChatGPT to create detailed prompts based on your idea [00:14:04].
- Style Reference: Find a desired style (e.g., from Pinterest) to use as a style reference in Midjourney [00:14:13].
- Image Generation: Plug the prompt and style reference into Midjourney to generate artistic images [00:14:09]. These can serve as base images for videos [00:07:24].
- Motion Addition: Upload the generated images to Runway or Cling and provide prompts to add motion, creating videos [00:14:54].
- Video Extension (Optional): For longer continuous shots, save the last frame of one video as an image, then upload it as the start frame for the next video, stitching them together later [00:18:04].
- Refinement: Use tools like Magnific Upscaler for detail enhancement or Cap Cut for adding text layers [00:14:55], [00:26:50].
Opportunities and Use Cases
AI tools present significant opportunities for creatives and startups:
- Branding: The biggest opportunity currently lies in branding, allowing for the rapid generation of consistent visual assets [00:22:54], [00:24:50]. Midjourney can be used to set up mood boards, ensuring continuity across generated content [00:23:35].
- Ad Content: These tools are excellent for creating engaging ads for platforms like Instagram [00:07:00], [00:26:19].
- Creative Exploration: They enable the generation of complex and beautiful visuals that were previously unachievable for many [00:27:30].
Limitations
Despite their power, AI tools have limitations:
- Web Design: AI in web design is still developing, with current outputs being “pretty mediocre” [00:22:58].
- Logo Design: AI is not yet proficient at logo design [00:24:19].
- Text and Vector Art: Text generation, while improved, isn’t always perfect, and vector art is not supported [00:24:26].
- Fickleness: Tools like Midjourney and Runway can be “finicky,” sometimes producing unexpected results, requiring users to adjust prompts and re-generate [00:21:12].
- Video Length: Current tools often limit video generation to short clips (e.g., 10 seconds) [00:18:02].
Users are encouraged to experiment and “get their hands dirty” with these tools to understand their capabilities and produce remarkable results with minimal effort [00:28:13].