I wanted to ask if we can discuss some pain points with AI, just because it doesn't do everything right or correctly. One big issue is hallucination, another can be memory, and even another is coding. Like last night I was going through putting together code for an app, but it kept iterating through the same mistake over and over again (arrgggghhhh!). In order to resolve this, I post the same issue on another channel and it seems to get through it. I lao plan to use other GPTs to see what they will do also in remedying this. Maybe a channel or conversation devoted to this will help those in need of help or who encounter similar issues (doesn't just have to be code of course, it can be anything). That way when folks stumble upon similar issues it has been discussed here. Any thoughts?

I'm adding a new lesson to the prompting course!! I realized over the last month of really working with a lot of image generators that you need a very different framework for prompting. I've put a few popular ones together to create my very own framework: SAAMPLED. S - Subject A - Action A - Art Style M - Mood P - Perspective L - Lighting E - Environment D - Details The subject is obviously what you want and the action is what they're doing. The art style can be canvas, charcoal drawing, or even specific camera shots and settings. Mood could be the mood of the subject or it could be the mood of the scene. Perspective is really important especially if you're doing anything with architecture. Lighting might be one of the most important things aspects of the prompt, especially for anything you want that's photorealistic. Environment describes everything that's around the subject - it can even be blurry! Finally details matter. Overshare a little bit on the details, be really creative and give the AI exactly what you have in mind. Example: Bad Prompt: A firefighter standing in a city street at night, photorealistic. Good Prompt: A weary firefighter (Subject) walking slowly down a rain-soaked street (Action), captured in gritty cinematic photorealism (Art Style), with a somber, exhausted expression (Mood). Seen from behind at street-level perspective (Perspective), the scene is illuminated only by flickering neon signs and the glow of distant flames reflecting in puddles (Lighting). The cracked pavement glistens under the rain in a smoky, deserted downtown (Environment). Details bring it to life: droplets running down the firefighter’s helmet, a half-burnt newspaper stuck to a lamppost, and faint silhouettes of ruined buildings blurred in the background (Details). Check out the difference in image quality!!

This is what happens when I accidentally pay an entire year's subscription while researching for the AI Tools course.

Having trouble with text in your generated images? I just played with Qwen and the results are pretty impressive!

Folks - the engagement so far has been awesome! Just a reminder - post those mini-projects as you go through the lessons! We have the posting party tomorrow, where we'll check out all the great work you are doing as well as like/comment your posts to help you level up on the leaderboard. See you tomorrow!