Thanks for watching! Let me know your thoughts and comments on DALL-E 3 below! 👍

quite interesting. ur examples were really cool

@sejinAI

9 ай бұрын

Thank you so much! I'm glad you liked it. There is so much more you can do with DALL·E! Appreciate the comment and support 🙏 🙏 👍 👍

wow this is crazy. I enjoy it a lot. Wonder where are we heading with this AI thing...

@sejinAI

9 ай бұрын

Super crazy right? I'm so excited to see what DALL-E has to offer in the near future! Thanks for the awesome comment and your support!!! 👍 👍 👍 👍 👍 👍

Hi Sejin, thanks for this video. I'm on ChatGPT plus, but Dall-E 3 still hasn't been rolled out to me yet. That said, your video has me thinking that it might be a fantastic replacement for tools like Midjourney and even Stable Diffusion, despite the copyright issues. Those images of of labelled camping gear really got me thinking about how easy it will be to generate high quality images for ecom websites within the ChatGPT interface itself. Keep up the good work, this video has me excited to see what OpenAI has in store for the future.

@sejinAI

9 ай бұрын

Hey, how's it going? It's interesting that you pointed that out. It made it seem like OpenAI released it to all Pro Users, but if you dig deeper into the forum, it's more of a gradual rollout with no specific schedule. Hopefully, you'll get the option real soon! In the meantime, you can use it on bing.com/create. The image resolution and quality isn't up to MJs standard yet but I think in due time it can easily replace MJ as the best alternative despite the copyright issue. I do think the knolling design ability is pretty amazing... I'm constantly testing it out to see what's possible. Curious to see what the next update will bring! Good to hear from you again! 😃 😊 😀

@brgtechdev

9 ай бұрын

@@sejinAI Thanks for the link, I've not tried the Bing AI yet, and it looks good. Yeah, I did some reading and found out that it's a gradual rollout. I'm very interested to see what is possible in terms of automating image generation with ChatGPT's functions. I haven't used MJ myself, but I do have a functioning instance of Stable Diffusion that I use for testing, though due to my hardware, I tend to just use playgroundAI for faster results. I also have you to thank for introducing me to Runway, I took a look at their video gen and it's quite compelling. Here's hoping that OpenAI start going down that track too. Will be looking forward to your next chapter in this OpenAI series!

@sejinAI

9 ай бұрын

Hmm... Automating Image generations with ChatGPT? Would you be able to clarify on what you mean by automating? I've personally never used Stable Diffusion but I do often use PlaygroundAI. I feel like all these platforms, inluding MJ, are similar in terms of output but MJ isn't has user friendly in my opinion. I'm glad you like Runway. I think with their latest update you actually have more control over your output so you don't actually waste credits. Have you tried Pika Labs text/image to video platform yet?

@brgtechdev

9 ай бұрын

​@@sejinAII'll do my best, but since ChatGPT doesn't have a feature like this yet, what I can tell you is just an approximation of workflows using custom code programmed to run with Stable Diffusion. Basically, any type of automation is just a set of instructions containing a list. Right now, most commercial AI tools allow you to perform singular operations, such as creating a single image, or a single batch of images based on a single prompt. To be able to automate generative AI, is to be able to preload a list of multiple prompts that runs and creates multiple outputs. So for image generation, imagine taking a list of 100 generated prompts, plugging them into ChatGPT and getting 100 completely different images back (I won't touch on token limits here, but just like how Open AI allows you to now just keep resuming a long ChatGPT response, long instruction lists should not be an issue if they are implemented). Or in my case, generate 100 different website UI templates based on a list of say, 10 themes, for a picky client! hahaha

@brgtechdev

9 ай бұрын

@@sejinAI With regard to Stable Diffusion, I would recommend it if you are inclined to delve into some of the more technical aspects of image gen, but really, PlaygroundAI is the most user friendly way to interact with an SD model, though in order to make the UI user friendly, they have peeled away some of the granular controls that you would have if you were interacting with the SD model directly. Not that it matters much though, the results are similar. Yeah, I also feel that MJ isn't particularly user friendly. I don't think the Discord setup is feasible for commercial purposes in the long run, though I am aware that a lot of people make do with it because of the image quality. Tools like Runway and PlaygroundAI, and of course Dall-E with their API layer seem to be more geared towards commercial users. I haven't tried Pika Labs, but will have a look. Thanks for the suggestion!

This is misleading because it hasn't been released widely yet. Specifying that its not widely released, but rolling out gradually would better inform your viewers. As a Chatpgt plus user who has been checking his account daily for the release, this was disappointing.

@sejinAI

9 ай бұрын

Hey, I appreciate the feedback. I did some digging the other day and found out that it wasn't a full release but a gradual rollout. I'm sure you'll get it in no time! Have you tried bing.com to test it out?

I dont see DallE available on my end? I have a full sub so thats weird

@sejinAI

9 ай бұрын

Apparently, it's not a full rollout but a gradual rollout. You’ll get it soon! 👍 👍

Dall-e is terrible of censorship

@sejinAI

9 ай бұрын

Yes! I couldn't agree with you more! The biggest drawback is the copyright filter... let's hope the other platforms don't follow this route.

Still no dalle for me :(

@sejinAI

9 ай бұрын

Yea, it's a gradual rollout. You'll get it soon!! 👍 👍 👍

😨😰😱😖😤😐🤔🙂😃👍

@sejinAI

9 ай бұрын

👍 👍 👍 👍 👍 👍 👍 ❤️ 💙 💜

It's good for childish stuff.

@sejinAI

9 ай бұрын

What's your preferred platform?

Yeah, good luck trying to get simple images, everything is "unsafe content"

@sejinAI

9 ай бұрын

Hahaha I know what you mean. Let's hope the other AI imaging platforms don't adopt a copyright filter!

@WaqManVR

9 ай бұрын

They keep getting sued, that's why

@sejinAI

9 ай бұрын

Yea that's true 😔

The DALL-E that chatgps is using is terrible. Cant create a image of a man without a beard. Just horrible

@sejinAI

Ай бұрын

Hey there, just curious what did DALL-E generate instead?