There's a new and totally unique way to generate images with AI, and it totally eliminates the need for things like prompt engineering. With this new method of generating AI images, all you need is sort of a fuzzy idea of what you're looking for, and you can hone in on the

exact image that you want. The new tool I'm talking about is called Flow State, and it comes from Leonardo AI. Now, full disclosure: I am an advisor to Leonardo, and I've known about this Flow State feature for quite some time now, and I've had to keep my mouth shut about

it. It's been really hard because I thought this was going to be one of the coolest new ways to generate AI images, and I really, really wanted to let the cat out of the bag that this was coming, but I haven't been able to. But now that

it's here and it's live, I finally get to show it off. So first, I'm going to show you how to use it; then I'm going to show you a little bit about what it actually does; and then we'll talk about the cost and some of the little details that

you're probably going to want to know about it. In order to use it, you're going to head over to leonardo.ai. This option for Flow State right here—at the time of this recording, it has a little new icon on it—you can click on

Flow State, or if you simply click on image creation, you'll have the option to switch between classic mode and Flow State. Once you're on the Flow State page, it actually starts you off with some inspiration to choose from; it'll give you three prompts with example images to

You can start with one of these prompts, or if you have a fuzzy idea of what you're looking for, you can start with that prompt. Let's say I want an image of a robot wolf, but I don't really know the details. I don't know the

environment I want it in yet; I don't know the style, I don't know the colors, I don't really know anything about what I really want other than I want a robotic wolf in it. I'm going to go ahead and set my image dimensions over here at 1024 by 768. You

have the option for 256, 512, 1024, and 169, or a sliding scale to really dial in more specific dimensions. I'm going to leave all of this on automatic on the left side, and I'm just going to type 'robot wolf.' We'll click 'Generate,' and what it's

going to do is it's going to generate a sort of grid of robot wolves, and they're going to be in all different styles, all different color schemes—basically, all different vibes, lightings, shot types, and color themes. And here's what's so cool about Flow State: if I'm not really

seeing an image that jumps out at me as like, 'That's that's what I had in mind,' I just keep scrolling. It's like an infinite scroll where, as you scroll, it just keeps on populating with more images. This is why you don't really need

to dial in that prompt exactly perfectly because it's just going to keep on generating images, and eventually you're going to find one where you're like, 'That's kind of what I had in mind.' It might not be exactly, but it's pretty close. So let's just scroll here. I'm

Going to keep on scrolling; it just keeps on adding eight new images as I infinitely scroll this thing here. And let's say I really like the aesthetics of this image right here. I like the blues and purples, if you haven't noticed.

It tends to be a color scheme that I really like, but the image still isn't quite what I'm looking for. Now, if I hover over 'More like this,' you can see it tells me that the lighting on this is Light Art and the color theme is

cool tones. So if I click 'More like this' over on the left, you'll see it update the style to Light Art and cool tones color theme. And now all of the images that are showing up in my flow state grid have that kind of color scheme to

them. Once again, I can just keep on scrolling and it will generate more and more images as I scroll until I find one that's close to what I'm looking for. This one down here is pretty cool; I like that one. If I hover over the 'More like

this,' you can see it tells us what the prompt is. Each one of these is now changing to a different prompt; it's still a prompt of a robot wolf, it's still using the lighting of Light Art, it's still using the color theme of cool tones; it's just giving us different

prompts that have those sort of features in them. But I like this one, so let's do 'More like this,' and now it's going to give me a grid of images that are Light Art, cool tones, and use that exact prompt; it's just going to be a whole bunch of

Different seeds now, but the same exact prompt and the same styles, and you can see now I'm getting images even closer to the last one that I liked. And once again, you just keep on scrolling until you find that image that's the perfect one that you were looking for. Let's try

another one. This time I'll leave it as a 1:1 aspect ratio, and let's just do a futuristic alien world. I'll leave everything the default here and click generate. And now I get a whole bunch of different styles and random images of what a futuristic alien world can look

like. It sort of starts with this wide variety of images; none of them really looking that close to each other. But let's say we really like this one with the sort of red in the background and the silhouettes of the aliens. I can click 'More like this,' and you can see it

set the lighting as backlit and the color theme as vibrant. And now we have a whole bunch of images with a vibrant theme and that sort of backlit look to it. This one's kind of cool and psychedelic-looking. Let's go ahead and

click 'More like this,' and now it drills in and starts to really dial in what that image looks like. No additional prompting necessary; you can see it writes the prompts for me, it picks the lighting and the color theme for me, and it just helps me get more and more

drilled down into what I'm looking for. Now, if you already know the vibe or the lighting or the shot type or the color theme that you want, you can start with that information already in there as well. So if I jump back into the flow,

State here and start over. Let's say that for Vibe, I want it to have an acrylic vibe, and I want it to have a vibrant color theme. Let's say I want to do a colorful coral reef with vibrant sea life. We'll click generate, and all of

them are going to start with this acrylic vibe and the vibrant color scheme. So it's already kind of drilled in for us a little bit because I've already preset a little of what I'm looking for, but I can drill down even more. Let's say I like this one; if I

hover over it, you can see that it's got mystical lighting and shot type, as cute. Keeping my acrylic vibe and my vibrant theme, I can click 'More like this.' It locks in all of these elements of my style, and now all of them are starting

to look a little bit closer together. I like this one because it looks like this interesting painting art piece. If I click 'More like this,' you'll see it's just going to use the same prompt. Now I'm just going to get a whole bunch of variations using the exact same

prompt. So now the style is all really close together. It is just so cool! One of the easiest ways to generate images if you just have that fuzzy idea of what you're looking for but don't want to, like, prompt engineer it and put all the

details into it, you can sort of let this tool figure out all the details for you as you go. And they call it Flow State because you could literally do this for hours. We just drill down and drill down and drill down, and eventually you have that image that's just like, 'Mmm, that's the'

Perfect image! I was looking for that. Now, as far as pricing goes, this may change, but as of the recording of this video, if you're on the Leonardo free plan, you can generate up to 150 images a day. If you're on one of the paid plans,

you get virtually unlimited images using this Flow State feature. So, free: 150 a day; paid: virtually unlimited, at the moment. Now, that pricing could change, so if you're seeing this video far in the future after it was recorded, there's a chance that pricing has changed

since then. But as of the rollout of this new feature, for the time being, it is available at those pricing levels. Anyway, I just wanted to show off this feature to you. I know I mentioned it in Friday's news video earlier this week, but I wanted to do a little bit

deeper of a dive on this one because I just thought this was a game-changing way of generating images. You no longer have to think about being a prompt engineer or putting all the details of what you're looking for in an image; you can sort of let the image

flesh itself out based on just diving deeper and deeper. And when you find an image you like, you can save it to your library. You can generate more like this, and then, if I head back over to my Leonardo library here, the image that I just saved is right here. I

can find it again. I can generate a video from it. I can generate another image based on this image. I can get all the prompt details around it; it is there and available for me to use whenever I need it inside of my Leonardo library. Anyway, that's all I got. I just wanted to share

This one thing I wanted to make a real quick video, break this down, play with it a little bit, show it off. I think it's really cool. Again, over at leonardo.da [Music]

Perfect AI Art Without Writing Prompts

AI, No-Code, Tech, Futurism - I'm a tech nerd and talk about tech nerd stuff

Matt Wolfe

레오나르도 AI의 새로운 AI 이미지 생성 도구인 Flow State를 소개합니다.

플로우 스테이트는 즉각적인 엔지니어링이 필요하지 않으므로 사용자는 모호한 아이디어로 이미지를 생성할 수 있습니다.

발표자는 레오나르도 AI의 고문으로 지금까지 이 기능의 존재를 비밀에 부쳐왔습니다.

이 동영상에서는 플로우 스테이트의 사용 방법, 기능 및 비용 세부 정보를 다룹니다.

사용자는 leonardo.ai를 통해 직접 또는 이미지 생성 옵션을 통해 플로우 스테이트에 액세스할 수 있습니다.

흐름 상태 사용: 초기 단계 및 반복적인 개선.

흐름 상태는 영감을 얻을 수 있는 초기 프롬프트와 예시 이미지를 제공합니다.

사용자는 기본 프롬프트(예: '로봇 늑대')를 입력하고 이미지 크기를 조정할 수 있습니다.

이 도구는 다양한 스타일과 색 구성표로 이미지 그리드를 생성하며, 사용자는 스크롤하여 선호하는 미학을 찾을 수 있습니다.

이미지 위로 마우스를 가져가면 조명 및 색상 테마가 표시되며, 이를 선택하여 결과를 구체화할 수 있습니다('더 보기').

이 과정은 '미래의 외계 세계'라는 프롬프트를 통해 시연되며, 사용자의 선택에 따라 반복적으로 개선되는 모습을 보여줍니다.

사용자는 분위기 및 색상 테마를 미리 설정하여 결과를 더욱 세분화할 수 있습니다.

무료 요금제는 하루에 150개의 이미지를 생성할 수 있으며, 유료 요금제는 사실상 무제한으로 이미지를 생성할 수 있습니다.

발표자는 Flow State의 사용 편의성과 사용자가 반복적으로 비전을 구체화할 수 있도록 지원하는 기능을 강조합니다.

저장된 이미지는 나중에 사용하거나 생성할 수 있도록 레오나르도 라이브러리에서 액세스할 수 있습니다.

Perfect AI Art Without Writing Prompts

스크립트