← все видео

FREE YouTube Thumbnail Master Class - How to Use 1of10 to Make Thumbnails

Roberto Blake · 2026-01-22 · 2ч 6м · 2 436 просмотров · YouTube ↗

Топики: craft-roberto-blake

Аудио ещё не скачано.

📝 Summary

model=deepseek-v4-flash · prompt=summary-v7 · 21 245→4 226 tokens · 2026-05-29 07:10:58

🎯 Главная суть

Инструмент 1of10 позволяет за 60–90 секунд генерировать готовые к публикации миниатюры YouTube, которые профессионал делал бы 45–90 минут. Ключ к качеству — не магия AI, а правильные исходные данные: собственные качественные фотографии на чистом фоне, продуманные позы, физический реквизит и хорошее освещение. Комбинация человеческого вкуса и фотографического фундамента с вычислительной скоростью и масштабированием AI даёт результат, недостижимый ни для одного из них по отдельности.

Почему 1of10 даёт плохие результаты у большинства пользователей

Основная ошибка — использование инструмента без качественных исходных материалов. Пользователи загружают снимки, сделанные наспех: плохой свет, размытый фон на улице, селфи с искажённой перспективой. AI не может из ничего создать отличную миниатюру. Секрет, который демонстрирует Roberto Blake: он импортирует в инструмент собственные студийные портреты, сделанные на белом или чёрном фоне, с использованием физического реквизита (огромный карандаш, иконка YouTube, реальные объективы камер). AI перекомпоновывает существующие качественные элементы, а не генерирует их с нуля. Когда лицо и фон уже отличного качества, результат получается почти неотличимым от работы профессионального дизайнера, и зритель не замечает, что это AI.

Как делать фотографии для миниатюр: настройки камеры и смартфона

Для получения "хрустально-чётких" портретов, которые можно использовать в 1of10, нужно выставить следующие параметры (даже на телефоне):

Физический реквизит как основа уникальности

Многие думают, что миниатюры Roberto Blake созданы целиком в Photoshop, но на самом деле значительная часть элементов — физические объекты, снятые на камеру. Он использует реквизит, купленный на Etsy и Amazon: гигантский карандаш, несколько копий пластиковой иконки YouTube разных размеров, YouTube-кнопку, настоящие объективы Sony G Master. Это делает миниатюры уникальными — AI никогда сам не придумает такой реквизит, и результат не будет похож на генерацию с нуля. Реквизит придаёт характер и запоминаемость, а AI затем может заменить текст или перекомпоновать композицию, сохраняя физический объект.

Скорость AI vs качество профессионала: конкретные цифры

Roberto Blake, имея 10+ лет опыта в графическом дизайне и собственные шаблоны, тратит 45–90 минут на создание одной качественной миниатюры вручную (фотография + Photoshop/Canva). 1of10 выполняет ту же работу за 60–90 секунд и генерирует не один, а 4–5 вариантов. Для не-дизайнера ручное создание может занять ещё больше времени с менее предсказуемым результатом. AI не может превзойти человека по вкусу и качеству единичного результата, но превосходит по скорости и масштабу: он выдаёт десятки вариантов, которые можно сразу A/B-тестировать, в то время как дизайнер за час делает один. Следовательно, выгодно потратить время на фото-сессию и подготовку ассетов, а генерацию и итерации поручить AI.

Работа с промптами: как управлять AI, как фрилансером с языковым барьером

Чтобы получать предсказуемые результаты, нужно формулировать запросы максимально конкретно, как если бы вы давали задание удалённому фрилансеру, который плохо понимает контекст. Примеры неточных промптов: "сделай мне красиво", "добавь эффектов". Точные промпты: "измени фон на YouTube-интерфейс, без канального арта", "оставь только меня и текст, убери всё остальное", "используй шрифт Poppins", "убери красное свечение с лица, оставив всё остальное". Можно указывать, что не нужно менять: "оставь мою фотографию нетронутой, поменяй только текст". AI (как и человек) лучше справляется, когда ему чётко говорят, что оставить, а что заменить.

Редактирование сгенерированных миниатюр: текст, композиция, шрифты

Инструмент позволяет изменять уже сгенерированную миниатюру с помощью кнопки "Edit with AI". Примеры команд, которые дают результат:

Результаты для продуктовых/обзорных каналов

Для каналов, не использующих лицо автора (обзоры техники, товаров), 1of10 работает особенно хорошо. Когда в инструмент загружены реальные фотографии продуктов (например, объективов Sony с чётким логотипом G Master), AI выдаёт миниатюры с точным отображением продукта, фона и текста. Качество таких миниатюр часто оказывается настолько высоким, что 3 из 4 вариантов можно использовать сразу, без доработок. Это особенно ценно для авторов, которые не хотят или не могут появляться в кадре.

Каналы с персонажами и иллюстрациями (Lo-Fi музыка, анимация)

Если у канала есть уникальный персонаж, нарисованный художником (например, женский аниме-персонаж для Lo-Fi канала), загрузив его иллюстрацию в 1of10, AI сохраняет все ключевые детали дизайна (заколку-панду, форму очков, куртку, шорты, кружку) и генерирует новые миниатюры в том же стиле, но с другим контекстом. Это открывает возможность: за 60 долларов (стоимость комиссии художнику) получить множество вариаций миниатюр, которые затем можно даже анимировать с помощью Google Veo 2/3 и создать целое видео для фастлесс-канала. Иллюстрация + AI = неограниченное количество визуального контента с сохранением авторского стиля.

Избегаем "генерической" миниатюры: роль уникальных ассетов

Основная критика AI-миниатюр — они выглядят "пластиковыми" и однотипными. Решение: загружать в 1of10 собственные фотографии с уникальным реквизитом (физические объекты, детали одежды), специально снятые позы и ракурсы. AI, имея такой фундамент, не "выдумывает" элементы, а перекомпоновывает их. В итоге зритель видит не стоковый образ, а реального автора с узнаваемыми атрибутами. Roberto советует иметь капсульный гардероб для съёмок (12 одинаковых чёрных футболок, несколько копий одной кепки) — это упрощает создание множества портретов без лишних раздумий и облегчает вырезание в фотошопе из-за однородного цвета.

A/B-тестирование миниатюр: быстрая итерация как главное преимущество

Когда AI выдаёт 4-5 вариантов миниатюр за минуту, автор может сразу загрузить их в YouTube Studio для A/B-тестирования. Это кардинально меняет подход: раньше на A/B-тест уходили часы или дни (сделать одну миниатюру вручную, потом вторую). Теперь можно за 15 минут сгенерировать 10-15 вариантов, отобрать 3-4 лучших и запустить тест. Если CTR не устраивает — повторный промпт и новая серия за минуту. AI отлично справляется с рутиной создания вариантов, оставляя человеку задачу отбора и утверждения.

Как подготовить кадр до генерации: советы по позам и выражению лица

Вместо того чтобы генерировать лицо с нуля (AI до сих пор иногда искажает черты, делает "расо-свап" или странное выражение), лучше заранее отснять 50-100 портретов с разными эмоциями: удивление, шок, гнев, радость, отвращение. Снимать на белом фоне, на расстоянии 30 см от стены, с хорошим светом. Использовать маслопромокательные салфетки для устранения жирного блеска на лице (нос, лоб, щёки). Носить очки с антибликовым покрытием или наклонять голову так, чтобы отражение не портило кадр. Если AI всё равно испортил лицо на сгенерированном коллаже, можно попросить инструмент удалить человека из композиции, а затем вставить своё идеальное фото в Canva или Photoshop за 2 минуты — это быстрее, чем править AI до идеала.

Пример работы с разными типами контента: практические результаты из трансляции

Roberto продемонстрировал генерацию для трёх разных каналов:

  1. Канал о технике (Creator Gear Guide): загружены фотографии реальных объективов Sony + фото автора, промпт "Best Sony camera lenses for studio photography 2026". Результат — 4 варианта, из которых 3 выглядят как ручная работа профессионального дизайнера. AI сохранил логотип G Master на объективе, так как исходное фото было качественным.
  2. Основной канал (Roberto Blake): промпт "How to grow a YouTube channel from zero subscribers 2026" с использованием фото с кнопкой YouTube. Один из вариантов получился идеальным с точки зрения композиции, но лицо было неудачным. Команда "Удали моё фото, оставь весь остальной макет" сработала — AI убрал автора, и Roberto вставил своё идеальное лицо в Photoshop.
  3. Lo-Fi музыкальный канал (ZenBuster Music): загружен арт персонажа (аниме-девушка), нарисованный художником за 50-60$. AI сгенерировал 4 миниатюры с сохранением всех деталей (заколка-панда, форма очков, кружка), что позволило бы затем анимировать их в Veo и получить целое видео.

"Ревизии" в 1of10: как исправлять недостатки без перегенерации

Раньше главной слабостью AI-инструментов была невозможность точечных правок. Новая версия 1of10 позволяет давать уточняющие промпты к уже готовой миниатюре. Примеры из трансляции:

Почему не нужно бояться AI-миниатюр: зритель не замечает разницы

Большинство зрителей YouTube не отличают AI-миниатюру от ручной, если она качественна. Roberto приводит пример: он сгенерировал миниатюру с собой, где AI идеально передал лицо и реквизит — никто из зрителей не заподозрил, что это работа алгоритма. Важно не зацикливаться на философских спорах "человек vs AI": миниатюра — не произведение искусства, а средство заставить зрителя кликнуть. Если AI делает это быстрее и не хуже, то 2 часа, которые вы экономите на ручном создании, лучше потратить на качество самого видео: сценарий, монтаж, звук, исследование темы. Конечная цель — увеличение просмотров и удержания, а не демонстрация того, что вы "всё сделали руками".

Дополнительные функции 1of10 за пределами генерации миниатюр

Инструмент является частью пакета, который также включает:

Советы по организации гардероба для съёмок миниатюр (профессиональный лайфхак)

Roberto использует подход "персонаж телешоу": у него в гардеробе висят 12 одинаковых чёрных футболок и 5 копий одной кепки. Это не прихоть — однородная одежда упрощает последующую обработку (вырезание из фона в Photoshop и Canva происходит за секунды, не нужно бороться с узорами или мелкими деталями). Авторам, особенно женщинам, рекомендуется создать "капсульный гардероб" из 5-8 нарядов, которые будут использоваться исключительно для съёмок на камере. Не нужно беспокоиться о совпадении одежды в превью и в самом видео — зрители этого не замечают, как не замечают изменения костюмов между сценами в сериалах Netflix. Экономия времени на решении "что надеть" — прямой выигрыш для креативного процесса.

📜 Transcript

en · 16 608 слов · 212 сегментов · clean

Показать текст транскрипта
is the one of 10 thumbnail tool. I'm going to walk you through a workshop of how to do this for yourself. If you want to get one of 10, you can go ahead and do that using my link. You could also use code Roberto at checkout. They'll give you 20% off. So 20% off at checkout. This tool is going to hit a little different for you because I'm going to be using it in a way that most of you are not because most people using these AI tools are doing them wrong. I'm wondering why they're not getting good results. The way that I'm using this tool is that I actually imported and a lot of you know this about me. A lot of you know that I have taken a bunch of quality headshots. You can do this if you just have a clean white wall. I have a white backdrop in my studio in my basement. I also have a black backdrop. You guys also know I have a green screen. So I use those and I take advantage of those. But if you have a clean background and if you utilize props, you guys know I'm a big fan of using props. So if you use posing, props, and photography, and then you use the one of 10 AI thumbnail generator, if you're using props, physical props, see, it's not digital, it's real. and you use proper posing techniques, facial expressions, all of those things, and you do good photography with good lighting, you know, I'm a stickler for good lighting, then you will be able on your own to craft better YouTube thumbnails. And so I wanted to show you this. So using the one of 10 tool, what I did was I uploaded real photos and it gives me really good results. I'll give you a primary example. We'll switch to the creator gear guide channel, which is supposed to be a photography channel that I'll eventually revive. And then what we'll do here is we'll actually feed it. We're going to feed it some of my favorite Sony G Master lenses. So we selected those. We're going to... select a camera body, and then another camera lens. We're going to hit done. Then we're going to select a photo of me that we're going to use. We'll just select this one. This is a handsome photo of me, if I do say so myself, that has me holding the camera. And we're going to the best lenses for... Studio photography in 2026. And we're going to have it generate some thumbnails for us. And you're going to get to see what it makes. Now, while it's doing that, which will not take long. In fact, how long is it taking? You can see it's actually going pretty quickly. So while it's doing that. I don't mind answering questions in the chat that you guys have about this because my goal here is to show you a process where you can generate thumbnails or a very close approximation of them very quickly because I'm a skilled graphic designer. I can do a lot of what these AI tools are doing, but I cannot do it fast and I can't do multiple iterations of it fast. This would take me by hand dozens of minutes. possibly an hour hour and a half to be able to make multiple thumbnails even with my skills my photography my templates to get to the same results take me a half hour to 90 minutes per thumbnail and uh that's because i know what i'm doing and that's to get very good quality thumbnails and then i'd still have to do more work after that for a b testing for a lot of you if you use this instead of spending 90 minutes to get a result And that's if you had my skill. If you are not a professionally trained graphic designer, it's going to be much, much more difficult for you. And after 90 minutes, you may not get something you're satisfied with. But here's the thing. This is going to do in about 90 seconds what a skilled graphic designer can do in 90 minutes. I mean, and it's wild. It's one of those things you have to see it with your own eyes to actually believe it. But I'm going to remind you, the trick to this is that I inputted my own photography, which is really good. For a lot of you, you need to be taking better photography. What I recommend is that you get a friend or a family member. For those of you who don't have professional camera gear, if you have a real YouTube camera, if you have one of these real YouTube cameras like what I use, even if it's a Sony a6600 or 6100. or a Sony ZV-E10. If you have a real camera, I recommend you have a friend or a family member, even if they're not a professional photographer, take some headshots of you that you can use for your thumbnails in different poses with different facial expressions and so on and so forth. If you don't have a real camera with your phone, instead of you just taking selfies, have somebody take close up and wide shots of you against a clean wall. clean white wall, you know, nothing in the wild, and just do different poses and different facial expressions. If possible, use physical props as well, and then put those and load them up into one of 10. Then the other thing that you're gonna wanna do is you're going to want to, oh, I ran into a little error, but that's on me. I think I just have to refresh it. What you want to do is you also, if you're a product reviewer or even if there's a specific pattern or background or something else you want to use or an asset that you have, upload those to one of 10 as well. In my case, I own real Sony camera lenses. I got the actual product shots of all my Sony lenses, a couple of my Sony cameras, and I upload them to one of 10. So when I want to get shots that involve those, it doesn't have to do generative much of anything for the actual physical photo of me or the camera lenses. It can just recontextualize them. It can pose me a little differently or change the expression. And so it works out really well overall. And I can show you some of, I think this one's almost done, but in the meantime, I could also probably just show you. some of the other ones it's generated in the meantime because again a lot of these actually came out pretty good so if i were to show you um some of these you would actually think that it's very similar to something i would make so the thing is i can literally make this it's just that it would take me about roughly 45 minutes to make this one of 10 generated it in about 60 seconds but it was using the photo assets i gave it plus my real photography uh so soul um soulless barbara says so you create a uh database of assets then prompt it to create what about adding the text yeah it does the text based on what i prompt and i could also uh have it edit and change the text if i wanted to as well um in the prompting which i'll show you here in a bit that i can um change this So if I go here and I go edit with a I can go Change the text to say I'm gonna tell it to up accidentally hit Something too soon there. So I'm gonna tell it to change the text to street photography instead of street Kings and then it's going to change that because by the way It's done this before for me where all that changed was the text that I can literally get to where all it will do is just change the text So in this case it changes up that I did this earlier and I got to change the text without changing much of anything else Because you can literally just give the prompt oh and change nothing else when you ask it to do something I'm gonna see if giving it one less camera lens here fix this problem or whether i have to refresh it but yeah so um it was able to do that for example um but this one i don't love it um i like the pose but i don't like the text or the background image at all make something that shows more of the youtube interface and uses Less text and so it's probably just gonna reroll this one for the most part assuming that I'm not having like a browser issue or anything like that. So it's basically on reroll this one And again that could take up to a minute or so All right, so watch this I asked it basically for a completely different image and look at that it gave me something Completely different here. I'm gonna ask it This is This is close, but I want a more generic YouTube homepage without Channel art. I also would like some graphics as well that demonstrate auditing and so it's going to Do some things there And it's basically going to reroll. But you see how quickly this just gets me to a better result. And again, it maintains this like photo that would just great photo with this prop and this pose that I use that's very thoughtful. I can even change the composition. Like I could tell it to put me on the left instead of the right if I really wanted to. Yeah, so this is actually a lot better. I like this direction, but I want... Black and white and red theme for it that matches more of the YouTube Static and also make it a bit more futuristic That's re-rolling that but on my other generation rolled an error But that's probably because that's why I get for trying to do two things at once probably also, this is I should have made a... Actually, I'm going to tell it. Let's do that again. But let's leave the photo of me the same. Eliminate the channel audit text. That'll get me closer to what I want. Again, sometimes you have to specify and tell it to leave things alone. that you don't want changed, and just make sure it changes the thing that you want. So we'll give it a minute there to re-roll. But as you can see, it doesn't take long for this to come up with some pretty good stuff. So yeah, that's actually not bad in terms of a design. Again, I just don't fully know how I feel about this one as a concept, but at least it would give me an option. So I do like that. As you can see here, I evolved this one from where it started to here all the way to here where i had it do a little bit of facial retouching because i just felt like my eyes weren't bright enough and i think this pops more again we'll probably literally use this one for the friday channel reviews refresh here for this one it had a little bit of trouble with my face and i think that was just an error so we're gonna re-roll that one And I think if I'm being completely honest, I think that Even though my photography here was really good. I just think that there's just some times where depending on How you look it could be a little off with facial features I'm actually gonna try one that's product only and I want to show you guys one that's product only so we're gonna just pick these Sony lenses and I'm going to show you one that even actually doesn't rely on my facial features or anything like that. Best Sony camera lenses for portography in 2026. So hopefully it rolls a little better with that one. But it can, again, use your face. And it can do it really well because you see it's actually done it for me. uh several times here so it's actually working pretty well for that um again sometimes it just depends and it'll go a little far depending on just how you look in terms of the facial accuracy sometimes but the other good news is i could prompt it to literally remove me and then if the only issue is how i look and i like the layout i could just have it remove me give me that and then i could go into canva photoshop adobe express and then add myself into it Boom so look at the portrait the ultimate portrait kit that one's really good Portrait Kings best so like this these are actually pretty good These are actually perfect for the aesthetic of creator gear guide if I'm being honest I probably could use three out of the four of these Just as is so for if you guys if any of you do product stuff or if you're doing things that don't require your face and you just had product images. And again, you can see how accurate it is because it has the G icon from the Sony G Master lens. That's also because, again, I put accurate product photos in here. I mean, would you guys believe how well this does? You're asking, can I give it brand guidelines? Actually, I have done that before with... Some more prompting here where I gave it some prompting around wanting You know some of this green color here What I'm gonna do here is I'm gonna tell it I like the green glow on my Oh Portrait, but I don't want it as pronounced I want the background to contrast with more of a yellow glow within the shelves, if anything. And I want the text lock up the flushed photo unchanged. So I can give it brand guidance and I can give it art direction and Get these things out of it and again I can also literally tell it to change things with the text as you're about to see and again when your ingredients are good when you use good photography and Also when you input your channel and has previous things you've done with your channel as another reference guide Then it tends to work out fairly well. So as you can see it took the direction fairly well here uh i am gonna ask it though make the text flush left but keep it on the right side so we'll see if it um accommodates that because i don't want it uh flush center for the text um yeah i can probably tell it what uh fonts to use to be honest with you uh it's struggling with the text placement here we are gonna try something else though Can you change the text to the to use the Poppins font otherwise the rest unchanged? So again, you got to work with a little bit We're gonna see if it can change the text to the Poppins font if I told it to something more like Oh change this to a decorative or a handwritten font or marker felt it would probably work fine, but again this for a lot of you will be much easier than messing around in photoshop or canva on your own but if you also wanted to at least see okay what is a good um you know uh image look like and then you wanted to recreate it in canva or what have you then you could so you'd want to be able to change those things this one actually came out pretty good just by itself so as you can see like one of ten can generate pretty good overall thumbnails as long as you feed it good assets and decent prompts and like I said it can even do it like I did here with just doing gear stuff if we change over to doing something for the Roberto Blake channel and then we use a photo of me let's use a photo of me Here at the YouTube play button yep, and it will basically just use that and then we'll do how to grow a YouTube channel from zero subscribers in 2026 and we'll see what it comes up with for that but again if you literally Start with the foundation like I told you of good photography, this is gonna go a long way. So what I recommend to a lot of you is if you're gonna use the one of 10 thumbnail generator, best thing to do is to use it right. One, connect it to your channel. It'll use your existing channel for some of its references, but then go in and do what I did and upload photography of yourself, well shot, well lit. For your photography, I'm also gonna recommend that you guys manually adjust the settings for posing for your photography um i think you guys should shoot at f1 uh sorry f4 shoot at f4 or even f8 for your photography that you use for your youtube thumbnails here's why by shooting at f4 or f8 and just using really really bright lights what's going to happen is you're going to get these perfectly tack sharp images where everything about you is in focus You also want to shoot against a clean wall, but you want to shoot against a clean wall and not be right up against the wall. When you guys are shooting your thumbnails and your headshots and everything like that, a lot of you, you're way too close to the wall. If you have more than one light, put a light behind you so that it can eliminate some of the shadows so you don't have to worry about that or get rid of that or have that affecting the shot, especially if you're wearing dark clothing. Even in your videos, sometimes it helps to have backlighting or an accent light to separate you from the background a little bit more, especially if you're wearing darker clothing, like I tend to do. So you wanna make sure you have good lighting. I recommend also trying to get as much of the shot right in the camera for your thumbnails. So again, try to do some posing. Try to get some good lighting, even lighting. If you can, make sure you have some type of diffuser on the light. Don't just blast light at yourself. Think about things like reflections. A lot of times I try to angle the lights and my head so that there's less glare reflection from my glasses. Even in some of the shots, you can see I try to minimize the glare on my glasses. You'll want to have to try to edit that out in post, even if you are using AI. So you want to avoid that. And I would recommend most of you don't shoot your YouTube thumbnails against a green screen. Green screen video is one thing. For photography, it's a bad idea most of the time. Just try to get a clean wall, a white wall if you can manage it. Get a clean white or gray wall to shoot against. Do that. Don't stand too close to the wall. Maybe be a foot away from the wall if possible. You can be a foot away from the wall, that's probably for the best. So you want to do that. Then I would say, like I said, shoot at F8 or F4. Then for shutter speed, if you're going to use manual shutter speed, whether you're using a phone or using a real camera, for the shutter speed, I recommend most of you use a shutter speed of 1 over 250 or 1 over 500. It'll help with the camera shake and shutter when you push it down, especially if you have a friend or family member shooting for you and they don't know what they're doing. If you just dial in a lot of these settings, it'll work. Try to get the ISO low. So you want to use bright lights to get the ISO low. This is true whether you're shooting on your phone or a camera for the shots that you're going to use for building your YouTube thumbnails. And so you want to do that and you want to do some posing. So those are things you're going to want to do, frankly, to try to make all of this work. Now, you can technically, and again, it depends on what you want this to all look like, per se. But you could technically do a sketch of what you want your thumbnail to look like, and then you could also prompt instructions. But I would caution you that you could do that. But I would also caution you to probably try to use real photography assets as well. And if you combine those, then the results will tend to be better and more of what you like. If you need to change the pose of the person or kind of do a body double. And this will get you closer. And this is how big YouTubers are doing it. Because as you've seen from me showing you guys what one of 10 can do. It all makes sense how these big YouTubers are getting their thumbnails. And again, I'll answer any of your questions in the chat. I do not mind. We got plenty of people in the chat tonight. Across all of our platforms, it looks like there's 103 of you watching across all platforms. So that's pretty cool. But yeah, if you have questions about this process, if you have questions about the 1 of 10 tool, if you have questions about... thumbnail design in general, you can leave them in the comments and I will try and get to as many of them as I can. But yeah, I do make almost all of my assets myself from scratch. However, there are times where I'm behind on a video because I'm behind on a thumbnail. This would be a faster process to get to it. And in fact, there are a couple of thumbnails from something like this that I would absolutely 100% not hesitate to use. You saw how good a lot of them came out. So let's see. Home Rapid Repair in the chat says, question, if I use a touch of AI to make small B-roll shots, should I tick the AI box in the video description? From a disclosure standpoint, sure. But honestly, in my opinion, if it's less than 10% of the video, I really don't think it's worth bothering with, especially if it's something. that most people would not notice but if you want to be a stickler in terms of just disclosure or you feel like your audience will feel some kind of way if you aren't going to be a stickler about the disclosure then you can go that rate but i think it really what youtube means when it tells you to disclose that is if you were doing something that where it would matter genuinely matter if um where it would genuinely matter to know if something was real or not if you're just using it for incidental b-roll and it represents less than 10 of the overall video personally i don't really think it matters but if you just wanted to say well if i use it at all maybe i should check that box you can but at the same time a lot of the audience is just so aggressively and ideologically anti-ai that they'll just assume that most of your video is ai instead of five seconds out of a 20-minute video. Question, does it have or have you used the feature to suggest the thumbnail from the title hook of a video? Actually, all of the thumbnails are from me doing exactly that when I'm prompting it. So when I'm typing these things, that's literally using the title or the hook of the video. to dictate the thumbnail in addition to me using um an asset from the photography now technically if you had graphical assets you could use those as well depending on your style of video um but yeah that's what i'm doing and i'm doing exactly what you're asking here so good question miami cloud doug is exactly correct so if your stuff they really when they say the ai disclosure they're talking about deep fake content they're talking about ai voiceovers They're really talking about that stuff where there's an expectation of realism or a human being. If you are literally doing this, guys, for like five seconds of B-roll where you just need to show somebody using a hammer, that's not really what that box is for. So I wouldn't worry about it. It's the same thing like when people ask, hey, if I got a free product, do I need to check that this is a sponsored video? And the answer is like, you didn't sign a contract. You didn't get paid money. You got sent a free product. It's like so you don't really need to check the sponsored box So again a lot of people but again some people feel some kind of way but a lot of people From my point of view just take it a little too far So here's what it generated for the Prompt here on how to grow a YouTube channel from zero subscribers I think some of these are actually pretty good I just think that the bottom two for the bottom two The design is good. The photo is not so what I would do in that situation is Edit with AI so what I'm gonna do here is I'm gonna say everything about everything here is perfect aside from the photo of me please remove The photo of me and leave the rest Me YouTube be specific and leave the rest of the layout intact Hopefully it'll take that guidance because the thing is if I do that I can just swap myself out for uh for this In photoshop now technically I could just take this in Photoshop, use the quick selection tool around this, and then ask it to do a removal, and then I could just put myself in its place. I could do it that way. My hope is that this gets me pretty bloody close, and look at that. So now I could just take this, and I can go into Photoshop, and I can add the perfect photo of myself to this, just like that. and have it. Now, for those of you who don't use Photoshop, you could obviously do the same thing in Canva and you would get that result. You'd even ask it to remove the arrow. Then once you have it, you could put all that back in. So, yeah. So, I mean, so I think this is like a really good example of how even if it doesn't get something perfect, but it gets everything else perfect, you can literally prompt it with the edit thumbnail tool, which is, by the way, it's still in beta. So this could only get better. This can only get better in terms of the 110 tool. Yeah, so again, we've got a link to that in description. The mods are also linking in the chat. So Miami class question, follow up. How does the AI thumbnails perform compared to thumbnails generated the old fashioned way back in my day, back in my day. So here's the reason I brought up this tool and why we're doing the presentation and demo of it tonight. We're doing the workshop tonight. Most of y'all are not really good at YouTube thumbnails. Whenever I do channel reviews, one of the things we keep running into is that you guys are making 1 out of 5 thumbnails by hand the old-fashioned way. You're making 2 out of 5 thumbnails the old-fashioned way. The 1 of 10 tool gives you a way to get multiple thumbnails that are going to be a 5 out of 5 or a 4 out of 5 every single time or are going to get you so good that all you would have to do is drop a decent photo in. and otherwise the thumbnail is done. So that is why I'm demonstrating this tool for you guys because most of you, your content is actually good. Your content is fine. But the thumbnails are so bad that you're not getting the views your content actually deserves. So a lot of you, you're not getting the views your content actually deserves because the thumbnail is bad. Now, you could test an A-B test. If you guys are wondering about it or you're worried about it, you can make a thumbnail the old-fashioned way. Then you can make one with one of ten. A-B test them. Do that for five or ten videos. See what happens. See what people prefer. But if they don't click, they don't watch. And a thumbnail is make or break. And a lot of you, you just don't have the time or skill to become a graphic designer. So, again, if any of you had to make one of these thumbnails, you'd also be spending like an hour trying to make something and still coming up short. Even as simple as this is, most of you would not be able to get to this result in an hour if you tried. Like, for example, I'm going to show you one of the thumbnails it came up with. So let's look at this one. As simple as this is, as simple as a thumbnail as this is, most of you wouldn't get to this in an hour. As simple as this is, a lot of you struggle with the layouts of where to put text in your thumbnails. Most of you would not be able to get to this result. So I know a lot of people, there are people turned off by AI, but I think it's been proven that most viewers on YouTube are not turned off by AI. Again, this result, as simple as this is, I could probably do this in 15 to 20 minutes. Most of you probably couldn't get there in an hour. And that's an hour that you could have been editing a video and making a higher quality video. So again, I know that there is some people are like, well, I wanna make everything human by hand or whatever. But for a lot of you, you can't make a thumbnail as good as some of these. If you spent an hour or two hours doing it, you just wouldn't get the same result. And that two hours could have been spent filming more content. That two hours could have been spent writing a better script, writing better hooks. That two hour could have been spent doing research. That could have been spent doing a higher quality edit. Why spend more time at the thing you're bad at just so you can brag on saying, well, a human being did it. A human being presenting a two out of 10 or a one out, sorry, not two out of 10, a human being presenting a two out of five or a one out of five result when AI can do a three, four or five out of five. and do it in a 10th of the time, it's not meaningful that, oh, a human being did it. If we were talking about you doing illustration and trying to genuinely make art, I could understand that. But a YouTube thumbnail is not art. It's not an art portfolio. It's not part of an art project. It's not part of an art gallery. A YouTube thumbnail has one job and one job only. Can I get people to click on a video? And if you do it just right, again, a lot of people will not even know this is AI. I mean, for the most part, if this had gotten my face perfect, you guys wouldn't know the difference between this and a thumbnail that I made myself because it's that close. If it came down to it, if it came down to it, most of you, for this one, for sure this one, not that one, for most of you, if you didn't know that one of 10 made this, you would literally just assume that this is a youtube thumbnail that i made you would just assume that this is a youtube thumbnail that i made it's that good um we have um note here from a fellow graphic designer um so tavo says i'm a graphic designer with that experience not able to create 10 options only two or three and keep going with the next video yeah exactly So just this speed and variety and being able to test thumbnails, A-B test thumbnails, get a higher click-through rate, that's better. And saving the time, saving the time and having the multiple options, and then being able to go back and spend more time on your video editing, spend more time on doing sound design, spend more time on the retention part of the video. If you can get the click, now you can spend more time on the retention. But a lot of you are struggling just to even get people to click on your videos, and this is your answer. The Artist Haven says, this would be good for artists or creatives that don't have their end product in hand before the video is recorded. At least you'll have a plan that you can edit afterward. Yeah, of course. You can always edit or change things afterward. And again, sometimes this is good for layout. Sometimes it's just, oh, this is what something should look like. For a lot of you, you could use this. If you still wanted to craft things by hand, you could at least know, okay, I know what it's supposed to look like. I'll do the mechanical myself. but at least you would have the layout correct. Miami Cloud says, hey Roberto, thanks for the detailed and real take on this. Yeah, absolutely. Again, I just want more of you to succeed with your content and you could do this better if you understood how to do it properly and that it's not about AI versus human, it's about both of them. The thing that I can do better than this obviously is my own photography my graphical assets and it is picking up and learning from what the style has been successfully on my best videos and it's um taking that into consideration plus my prompting so it is the human and the ai working together because I cannot be as fast as the AI, nor can I produce the variety as quickly. So it has two things on me, if nothing else. It has speed and it has scale. And no matter how hard I work, no matter how talented I am, I can never match the speed and I can never match the scale. That's what AI does well. However, when it comes to taste and things that look good and some of the base things, it is better at those things because it has my assets to work with. So it has my original assets that I made by doing my photography, doing my posing and using my physical props and products and doing my product shots. Those elements, those assets, I did that. And then I'm giving the direction as the art director and doing the prompting. Neither of us could get the best result possible on our own. I could do one result and it could take me an hour and it'll be really good. It may be as good or better than the AI. The AI can get a bunch of stuff fast, but can it be as good as me? When we work together instead of against each other, we get a result that's really good and we get a lot of them and we get them faster. And that's where people need to stop fighting with AI and need to learn that a hybrid solution is really the answer. So by using a hybrid solution, we can just get to a better result. How can you create a thumbnail for my videos after posting? Well, the thing is, you can always change your thumbnails in YouTube even after you've posted, and you get to A-B test your thumbnails even after you've posted if you go into YouTube and you go into the YouTube dashboard. This is going to be more so on desktop, so if you've only ever been doing YouTube from your phone, you may not be aware of some of these options. Most of you need to start using the YouTube desktop more. for the dashboard because then you can be a power user because the phone cuts like 80 of the features that youtube actually has so a lot of you are just too used to posting from your phone by the desktop more powerful so mobile first or mobile only youtubers that use their phone for everything they're only getting 80 of the uh features that youtube has or sorry they're getting less than 80 of the features youtube has yeah if you're struggling with views i think this would help the majority of you The majority of you struggling with views. This would absolutely help you because you could make thumbnails faster and you can make a variety of them and you can do them very quickly. And again, some of them can be even more advanced. I mean, this one, this one is pretty good and I could still manually kind of do this. I would have to do the pose. But again, if I wanted to and I wanted to do the posing myself. I could get to this result, but oh my God, would it take me much longer than this did to get there. So what I'm better off doing is using AI for something like this because it gets me there. There's only minor tweaks I would have wanted to do on that if I'm being honest. Again, this one is not bad. There are minor tweaks I would do to this, but this is actually really good and it's a different style. It could be, it could, it might A-B test better for all I know. This one's actually pretty good. There are some changes I would make here. I could literally tell it. Let's see. I could tell it change the top most YouTube thumbnail. Oh, sorry. YouTube play button icon to be white instead. like the left most change nothing else and let's see how close that gets us so again i'm giving it the kind of direction that i would have to give to a freelancer um yeah especially someone overseas i think you should treat all ai tools like they're an overseas freelancer where you have to use precise language and be very clear and tell you exactly what you want in in detail And that's how I think we should use AI when we're directing it, if I'm being honest. I think everyone's problem is they keep using AI like it's a Google search box. I think you should use it like you're actually giving somebody instructions and knowing that there may be a language barrier. If you go from that premise, I think it makes more sense. Because then if you were talking to someone where there's a language barrier, you'd have to use incredibly precise language. And yeah, the revisions are pretty quick. Usually, again, they take around one to two minutes at most. Some of them move a little faster than others. I think some of this is frankly hampered by the fact that I'm live streaming at the same time and the fact that I'm using Firefox instead of Chrome. This all works so much better with Chrome, but I'm a Firefox guy and it's for privacy reasons. and memory leak reasons and stuff like that um it doesn't help that i have like 30 tabs open either if i'm being honest so i'm sure that's no small issue that i have 30 tabs open one of them's chat gpt that's probably not helping anything so like what if i close the overwhelming majority of these tabs that's probably for the best i wonder if it froze but again it could literally just be all these bloody tabs i have open I'm still closing tabs as we speak. I'm not joking. That's how bad it is. I'm like really bad when it comes to having a million and one tabs open all at once. Let me just refresh this or come back to it. But as I said, oh, I have another example for you. So something I did was I fed it. I fed it one of my own design layouts and it came back with a unique take on it. And this was actually really good to the point to where if I eliminate the text and this right-hand graphic, this concept is incredibly usable and dynamic, and I think it would do pretty well for something on my channel. Even if I don't do a phases video or a steps video or a stages video, I think that this would do really well. Miami Cloud says, thanks. I'll check it out. I have some older videos that I need a thumbnail makeover on. I haven't had time, but this would help me so I could focus on creating new videos. Yeah, that's the thing about this is the speed. The speed is incredible. And again, just getting to a good result. Yeah, you change the thumbnail when you see the views are low. Yeah, absolutely. So like I said, I'm a graphic designer who can do things by hand, but at the same time, I just cannot work as fast as this. but if i put my same assets that i would be using as a graphic designer which is what i do with this right is i can take um the assets such as i can go ahead and prompt it real advice for small and all right so if i go and i take that then i take this And then I have it generate something we're gonna see what it does there, but again it's how quickly and how Capable this is especially with the new Model that they have here and then I've also shown you what can be done with the revisions because that was the struggle before to Generating something that looks good generating something custom this so it's not generic and then generating the revisions and the variations Because everyone's complaint about AI thumbnails is, oh, they all look the same. But these do not look generic at all. You guys can tell me in the chat or in the comments. Be honest with me about how you feel these look as I'm doing them for these different channels. So hopefully this one doesn't crash out on me. Again, my own fault for having so many tabs open and using Firefox instead of Chrome. Oh, excuse me. But yeah, I'm also happy to answer any other questions you guys have about this. Because like I said, I think if more of you were doing this, instead of struggle busing in Photoshop or Canva, knowing you don't have the background, knowing you don't have the skills, that's no offense to anybody. I spent years getting these skills. If you just focused instead on, can I make really good photography with poses and props of myself? Can I get good product shots? Can I get, you know, good assets? And then can I bring them in here? And then can I... Give it some directions. You can instantly generate much better images for your YouTube thumbnails. Hey, Angela. Yeah, Angela and I actually just had a coaching call. It was either this week or it was last week. Actually, it was last week. I think it was before I went to Florida. Yeah, the generated thumbnails look great. Maybe this is something I can use to revamp my thumbnails. Yeah, absolutely. Ryan DePaul says. Yeah, absolutely. You can actually prompt it for that on your side. On your side, you would be able to prompt it for that because... again it gives me a very different for for my different channels it gives me different looks because for the photography channel for the creator gear guide channel it gives me something very different than what it gives me for the roberto blake channel and it's because i used product photography um of the sony cameras and you can see it gives me a much different look different color scheme even a different approach to the fonts because i understand it's a different channel And it goes more photo-centric than design-centric when it's trying to accomplish that, even based on the prompt I give it for what the title is. For example, for the wildlife stuff, look at this. This is what it gave me on the wildlife stuff. So again, it's a completely different look for whatever channel you're doing. And so you don't have to worry about, one, everything being generic. It's not going to be everything generic. Number two. It's going to go off of the basis of what the channel is about and even what you previously have done. And then again, if you do what I do and you upload assets into it, it will be able to use those assets to get you a better result. And this will work for gaming. This will work for beauty. This will work for lifestyle. This will work for vlogging. This will work for talking head. This will clearly work for product stuff. If I switch over to my music channel, if I switch over to... zen buster music and i go lo-fi beats to help you and then what i'm going to do differently here is i'm going to upload a new image so it's going to prompt me to upload an image um i have to find an image that i can use for this that fits the lo-fi channel because i think i downloaded an asset that i had made for the lo-fi channel anyway um yeah i actually do have an asset made for this that was done by an artist so i can use that did it upload hang on i might have to re-upload it yeah i had an artist hand draw um something for my lo-fi channel so if i just pick that image file is too large maximum size five gigabytes hang on So we picked the Zenbuster music channel. We gave it the title lo-fi beats. Actually, I'm going to go further. I'm going to say cozy lo-fi beats to help you relax. Give it this image, and we'll see what it does for our lo-fi music channel, just to show you how radically different it can produce results. and that it can generate like a softer and maybe even a female aesthetic since we're gonna be giving it a female character that was designed based on my sisters if they were anime characters oh wow these came out good these came out so good that we could probably take this into google veo3 from here and literally generate the whole bloody uh video for it, and keep everything about this image the same, but remove the glowing aura from the character. Yeah, this can clearly generate any aesthetic you need to, but again, the key to it in my view, the key to it in my view, is the assets that you start with. And again, when you combine human work With the AI, you get better than what both did. Because remember, I took something I had commissioned, right? I paid like, I think 50 or 60 bucks for this commission, maybe more, years ago. And I took that artwork that was commissioned for my music lo-fi channel. And then I threw it into this and I got four variations that are all really good. And I could literally, at this point now, animate those variations. off of that character that I actually own copyright free and clear because I hired an artist for it. And then I could not only generate the thumbnail for it, I could literally generate the whole video based off of the image that it created. And there you go. It removed the RF around the character and it intuited to have the steam coming up from the teacup or coffee mug there. And this came out perfectly. So again, this can definitely handle a female aesthetic. So, you know, Angela, you're still in the chat. It can definitely generate a female aesthetic. And it actually can do really good. And it's good at variety. And it will take the context and themes of the channel into consideration. So again, we got some really good stuff here. And it used the character art that I commissioned a real artist for. And it got a really good AI off of that. And it wouldn't have been able, I think, to do that if I just tried to prompt it with text alone. That's why I am telling you all that if you use assets and then you base it off of that, it will do better. Oh, it finally generated the YouTube one. I hate that when it loses for a minute and rarely do this, but it's again. The new version is still working itself out where it won't get my face right. And in this case, it literally race swapped me, but we'll just get rid of that one. But other than that, the rest of these are kind of in line with a video thumbnail I would make for my channel. Like this one, this one is something y'all would actually expect from me and you wouldn't know it was AI anyway. Like you guys would believe that I made that myself. And actually I have everything I need to make this exact thumbnail except I'd have to like probably generate that broken chain. That chain would probably be the main thing and then probably the broken glass but I could probably find broken glass in Adobe stock. But I could get really close to this myself without AI. It would just take me longer and that would be that. This one, this is a good concept but again, It lost my face. But this is an amazing concept, even with the holding the 01. So all I would do is probably have it remove this from here. Because again, and also the edit thumbnail is in beta. When they take this beyond beta, and I'm gonna make some recommendations to the team for this. It'll probably be actually easier to do a head swap where I need it to, but I could literally take this into Photoshop, do a head swap. I don't need to do a body double. I could literally go into Photoshop, just do a selection mask, do a head swap with Photoshop's AI, and I would have a perfect thumbnail from this because this concept is strong. This is really good. So, yeah, let's see. Hey, bro, just heads up. If you're having trouble with the one of 10 link and using the discount code, try taking the HTTPS out of the URL and try just typing in HTTP colon slash slash and see if that changes anything and see if that works for you. Let's see. So let's try this. I'm gonna paste this URL in the chat for you guys. And let's see if this works better for you guys and see if you can use the link. So let's go ahead and try that. And mods, if you could copy and paste this link that I'm sending and see if this works better for everybody who is trying to use it. And remember, even if you just go to oneoften.com and you use code Roberta at checkout, it should also work, but try this and see. If not, I'm gonna definitely message the oneoften team. and let them know that a bunch of you were coming through tonight and you were having trouble with the um code and the affiliate page if you guys are still having problems oh this one's actually pretty good and again it's close to in line with exactly what you guys would expect from me all i would do is tell it to take some of the red off my face in fact actually if i prompt it um leave leave everything else alone and just take the Red go off of face. And then I'll probably tell it to reverse the arrow and have the arrow going up instead of down. Yeah, I'm gonna have the arrow go up instead. All right, so literally, all right, it did the perfect retouching there. It still kept the reflection, it still kept the reflection, you guys. Keep everything the same. Just have the arrow pointing. Upward instead it's still crashing at checkout. Okay for anybody Yeah for and Doug if you could if you could DM the one of 10 team and screenshot and let them know Then that'd be great too. Thank you so much And if anybody like if anybody's on X if you let the one of 10 team know you're trying to Check out using Roberto's code, but you're getting an error. Let them know and see if they can definitely fix that. But if you use code Roberto at checkout, you should get 20% off. If it's not working tonight, keep trying later this week and see if they've managed to fix it. If you're watching the replay of this, hopefully it's been fixed by then because I don't know what's going on on their side with that, but that would make sense. That would make sense. But yeah. Boom, we were able to make two tweaks to this in real time. And in seconds, we got the result that we wanted. It kept the reflection in the thumbnail. It took the red glow off my face and that came out perfectly. We flipped the arrow. We did all these things that would have probably taken us a lot more time and we were able to do them very quickly. Yep, and I'm gonna try and see if I can DM the one of 10 folks over an X. Just one second. All right, I've let them know. But thank you for looking out, Doug. Thank you for looking out, everybody in chat. Appreciate you. We'll try and get you squared away so you can get this tool and get your discount ASAP. We're working on it. But thank you for letting me know. But yeah. But overall, I think that, again, we're just seeing in real time how good this can be. Let's try this one. I'm gonna try with the same character. I'm gonna try cozy lo-fi beats to help your anxiety. Guys, we've literally tried product-based thumbnails. We've done how-to, talking head, personality-driven thumbnails. And now we've done character-driven thumbnails. And so you see it's not just one type of thumbnail. that this is capable of doing. It can easily do a lot more. And then I'm gonna show you how to do a storytelling thumbnail that doesn't use a human. I'm gonna pull one of the graphics that I made and I'm going to find that here and I'm gonna send it from my phone to a desktop and then we'll upload that and we'll use it. to do a storytelling thumbnail in a minute on one of my other channels. While this is generating, I'll send those to the Mac Mini. I wonder if this one is hanging. It could just be taking a while to come up with the concept. But in the meantime, I've downloaded my other assets because I'm going to upload a new asset. I'm going to show you guys how to do a storytelling style thumbnail that doesn't use a face. or a product, but we'll use a graphic. And I guess I can take your questions in the chat while we're waiting on that one to generate. Again, it's possible that that's hanging and I'll have to refresh. Yeah, we can take a couple of your questions here in the chat. I'm going to switch this over to demo, showing you guys a storytelling thumbnail and see how that goes. I know it's here somewhere. That's not it. That's it. I probably could have labeled this. better if i wanted to find it now that i'm thinking oh there it is there's what i need that's the one and generate all right so we're gonna let that asset generate a new set of thumbnails while i take your uh questions here because we're gonna pull off a storytelling thumbnail off of that for those of you who do things like commentary and you cover celebrities or you're covering Public figures this can actually be really good for you because you can always get a really great photo of them and then boom That's how people are winning in commentary news That sort of thing reaction videos even and so that's doing really well for people in that for those of you who reaction faces This could be a gold mine for you. So yeah, just a lot of opportunity here I do think this will be game-changing for quite a few of you. Yeah, I think Angela this this could Potentially be very very good for you Yeah, the photos are so crispy yeah with me again like I told you guys earlier the photography aspect of this is key and what I do is and you guys see that I have good cameras good lenses, but even with your phone there's a way to get this like I said the settings you want to dial in are F4 or F8 to get really crisp sharp photography with everything in focus because When you're trying to get these shots for the thumbnails, you don't need a blurry background. You need to shoot on a clean background. A clean background. Usually a white wall. Not stand too close to the white wall. Maybe stand a foot away from it. If you can use a bunch of bright lighting, do that. But balance it out. You know, you don't want... And then also consider... And I should have done this before I went on camera. A trick for you guys for good headshots and photography is to use these. These are... oil blotting tissues okay these are organic oil blotting tissues um you want to get these so what these are going to do is you're going to use these and i learned this in my time in photography but also um working with people on production uh video production is you want to use these oil blotting tissues if you can't use powder and makeup and a lot of times you can't and this is fast portable and cheap This will take a bunch of the oil off of your face and take a lot of shine off your face You want to use it definitely on the bridge of your nose Around your mouth your cheeks right up under your eyes, and of course your forehead So all these places where you would have shine and oil as you can see even with me on camera right now You can see that actually takes quite a bit of shine right off me right then and there So you know it takes it does a lot very quickly just a couple of dabs will do you So you want to do that and that'll actually make your head shots and even your on-camera presence a little better Sometimes I forget I have these at the desk sometimes I forget before I go on camera as you can see That makes a huge difference, right? Doesn't it make a huge difference? Right off the bat with this all this bright light a lot less shiny Right lot less blown out in certain areas, but just because I did that One thing and so there are little tricks Yeah, there are little tricks you can do Like that a lot of people don't know about they don't have a background photography Again why have you for thumbnails you don't have a background photography or design if you at least do a little bit of photography And you get at least the lighting good you get your pose is good and again you learn to incorporate physical props Like I do Physical props I got these off Etsy Etsy is great for physical props Amazon as well. That's where I got my giant pencil. So as you can see, I have physical props. This isn't even digital. This isn't Photoshop. See, like there's only so much actual Photoshop and magic wizardry that I'm doing in the thumbnails, believe it or not. A lot of people think it's all Photoshop or all AI generative. What I am doing is I'm using physical props and it makes my thumbnails unique and different. And a lot of people don't realize that. I was also one of the first people to start doing the pose with my YouTube play button as well. Out of all the YouTube educators, I was one of the first people to make that a normal thing. I may not be the absolute first, but I was one of them for sure. I definitely am the guy who uses the giant pencil to do my thinking pose. And again, physically holding the YouTube icon, I'm one of the only people that went through trouble getting a... I have multiple physical sizes of this as props. And so it's physical YouTube play buttons that are actually in my thumbnails. It's not all Photoshop or AI. It's enhanced from those things, sure. But I start with a physical prop in most cases. Sometimes I'll just use a placeholder prop. And then if I need to Photoshop or AI something in, I can have it just change what the prop in my hand is. But a lot of times I'll actually use a physical prop. So props, poses, and photography are a big differentiator, and that's how you end up with unique images. Speaking of unique images, we just got back some of my results here. Look at this. Aside from the fact that it totally lost the plot on race swapping me here, and it got my face wrong, but it's a good expression, I can just swap that out in Photoshop. But these two, these two came out perfect. These two came out perfect. So we're gonna Maybe I can keep that one because I can fix actually I'll keep those two because I can fix them later with Photoshop But with this one the change I'm gonna have it make is Leave everything else alone but change the text to Make sure it's capitalized so it takes it I don't want to say algorithm broken because that just sounds too dramatic. The facial expression is dramatic enough to sell it. So, yeah. So, boom. It made the changes I needed right then and there. Yeah, that's dramatic. But it definitely gets the point across. And I'll have it reduce the text size by about 15%. I think it's just covering up a little too much of the YouTube play button there. And again, it's gonna do this pretty quickly for us. Again, I think this came out really good for a storytelling style thumbnail. Yeah, I think that's pretty good. Did it get a little smaller? Yeah, it got a little smaller. Have I tried pixels? I played with it a little bit, but I only did it on a trial version, so maybe I didn't give it enough of a chance. But again, the thing about 1 of 10 that you guys need to understand is that We're demoing the thumbnail tool today because it's the most important thing most of you need to utilize. And then I want to teach you proper preparation for it by using photography. And I want a chance to tell you guys more about photography because it's it's a foundational skill for being a YouTuber. Honestly, if you're going to be in your thumbnails is or if you're going to do product stuff is you want to get the photography assets, right? So I wanted to take the time to tell you about the photography, posing, props, and get that right, and then use the AI tool to help you use that as a foundation and then build on it. But 1 of 10 is a complete packaging system that solves most of y'all's biggest problems. You can find the right ideas and topics for videos and validate them and say that they will actually, here's why this thing would get views. You can validate the idea of your video topics using 1 of 10. You can then build the assets and then use the AI tool and generate a really good thumbnail off of that. You can find a good title structure and inspiration for that and know what's working already in your niche off of one at 10 as well with the title generator. We can generate variations of your titles for A-B title testing. So you have the ability to build and generate thumbnails for A-B testing. titles for avi testing you can validate your video topics you can also figure out the timing and see what's popping in your niche now and what's good right now or what's constantly performing and is evergreen so one of ten is a system for you to get um topic title thumbnail and timing correctly and it will do well yeah the play button looks great because the play button is actually the asset that i uploaded and then again like i told you the other assets that i uploaded were my own photos so the reason the play button looks good is i want to show you guys what i did i uploaded this asset of the thumbnail and so it had that and it had reference photos of me to work off of so if your assets are good as your foundation if you have good graphics if you have good photo assets and then you do a good prompt, then you will get a good result from this. That's the key. And it will figure out the rest without you having to be a graphic designer or a digital artist. That's the key to this. In another demo, another time, we'll actually try out this sketch feature and the buy swap feature because I need to play with that a little bit more myself. But as you can see, this does a really, really good job. So this one's going to generate another set of assets to tell this particular story. So, yeah. Oh, excuse me. I'm glad the stream is holding up considering all the weather issues and the fact that earlier today I actually was having mad internet issues earlier today because of the weather with these storms. Generally, a little worried about the overall storms. What's going to happen? Oh, with this one, this one came out really good. The rest of them kind of lost track of my face, but I could just swap myself in Photoshop with that for a lot of them. But the layouts are really good for the rest of them. This one is probably the closest one to actually doing it. And that would be a believable facial expression of me. But I could also recreate that. Seeing the layout now, I would have inspiration if I wanted to recreate that myself in Photoshop. and i have the assets same thing with quite a few of these i could get really close or again if i didn't want to have to go through all the trouble in photoshop i could um do the same poses i might even have similar poses to this because in this case i just use this asset if i give it the reference and i tell it um if i give it say this to work with We could probably see what would happen here. Now it has something to work with. But yeah, how are you guys enjoying the stream and seeing this play out in real time with real images and what it would do for my different YouTube channels? We really love you guys' thoughts on that. HipHopLobbyist says, hey, my mom and I were literally talking about you right now. Going to book that session. Yeah, absolutely you should. Oh, these came out pretty good. I like this one on the right a little bit more than the one on the left. Oh, I like these two. I like both of these on the right more than the one on the left. And this one, I could just photo swap a more similar photo of me for this, and it'd work out perfectly. So I could just head swap myself there, and it'll work out perfectly. This one's actually not that bad. This one's actually not terrible. I don't even know if I need the in 2026. Now it's going to just randomly remove something because I typed it too soon. I hit enter too soon. That was my bad. But I think if I remove the in 2026 that that one's actually really good. But yeah, again, I'd love to answer some questions for you guys if you have any questions about this, like I said. And I'm going to show you another one for the lo-fi channel. I'm going to bring in a new asset for that. I'm going to find... another jpeg that i had an artist make for that to again show you you know what happens when you have um the hand-drawn work like you know if you want to use it in this particular way if you have an artist that can do illustration for you again you can use illustration and you can benefit from that and then take it further with the ai Let's use this. This is one of the first versions of this. And so for the LeFi channel. So we'll have it generate something off of that for the Zenbuster channel. But it's just really interesting that, again, this is fully capable of getting multiple styles of thumbnails correct for different channels and applying them. Yeah, a lot of people don't think about um the ai stuff and again a lot of people either um overestimate overestimate what it can do or they underestimate what it can do but once you get a little bit more creative with it it works out huh it did a really good job oh my god it did a good job with these uh the two at the bottom actually all four of them this one contextually um i just don't love that it's a standing one but it's fine but could from a practicality and a contextual standpoint, these three are great and are usable. So usable to the point, it got the details right. It got so much of the details right. It got the panda hair clip from the original character design correct, which was a very big and specific detail that I had in the drawing on purpose. It got that right. I believe it got the shape of the glasses right. Yeah, no, it got the shape of the glasses. It did it. The jacket, the shorts, it got it all right. It maintained the character fidelity spot on. The character fidelity and continuity was spot on. I don't know why I'm surprised, but I am a little surprised at just how crazy good it came out. I could literally take this into Google Veo from here and have it animated. For like next to nothing in cost. For my account. For my Google Veo credits. In Veo 2 or 3. I think I have access to both models. Which is really good right now. Take the music I had my ghost producers make. Which was human. And so because remember. I had a human draw the character design for me. Based on. References and I designed it again to be like well, what if my sisters had an anime character that looked like them, right? So I had a illustrator that paid multiple times To do character art for me I've actually even paid an animator before to do this But now I would just pay an illustrator and then I would just have AI do the animation because it's just gonna be faster more accurate more of what I want and cheaper by far so ironically cheap fast and good AI is getting very close to cheap, fast, and good on too many things to ignore now. And so now, all right, it gave me the thumbnail off of the original character art in a different pose in a different context while maintaining character fidelity. And from here, I could even literally have this animated. And then I did pay a ghost producer to make actual music, not AI generated music. And then, okay, boom. There we go on my music channel, my faceless music channel. So, boom. It would have that. So, this is really good. I'm probably most likely going to end up using this, to be honest with you. I'm going to end up using this. I'm pretty sure. Because I'll just have Veo, Google Veo animated. And YouTube can't go, oh, that's this or that because I'm using their product. So, again, this is an example of what you can do with these kind of tools. So, with one of 10, I was able to... You know do something really cool here and I can still edit variants of this Remove the wave form and leave everything else Unchanged and it's gonna remove that little wave form there because I don't need it because if I actually want the thing that I would do manually that I would do for these is I would actually do the animated wave form in Adobe After Effects myself because i can because i actually know after effects i should do a tutorial on this channel for um generating an after effects wave form if i do i'll probably also um make some kind of download template for it um i've been actually thinking of selling a few of my after effects files in a bundle really cheap um so it's something i keep thinking about so we'll see if i actually uh do it because again Exporting my After Effects files and then just letting you guys buy them for a couple of bucks wouldn't be the worst idea. It's just, I don't know how many of you actually use After Effects or would care, but I mean, I have some really cool After Effects files. Like I have some of these waveforms in After Effects generated. I have some particle effects and bokeh background effects generated. I have some smoke effects generated to where you could customize it, but you wouldn't have to start from scratch yourself. so it might save you 45 minutes to an hour because you just have my files so i don't know if that's something you guys would care about but we'll see um i have been planning to do like one or two after effects tutorials ironically i'm not great at the basics of after effects yet i'm good at the generative parts of after effects so i'm good at intermediate and some advanced things and then for after effects i really suck at the basics that's the irony Because I just went to VFX I went to VFX in After Effects first and Skipped over the basics to be honest with you. I just wanted to do visual effects and compositing Because back in the day that's kind of just the thing I wanted to do VFX compositing rotoscoping keyframing I was doing all that stuff But I end up skipping over a lot of the basics so ironically I can I can generate and I can map special effects features, and I can do a couple of superhero effects in real life in After Effects, but some basic things like text masking can't do it to save my life. So this is kind of an interesting piece of lore about Roberto is that a lot of times Roberto can do intermediate to advanced stuff and not the basics. That's terrible. But... To be fair, in my other Adobe programs, I can do all the basics and a lot of the intermediate and advanced stuff. Photoshop probably being one of the best examples. Illustrator, InDesign. For those, I can do all the basics and then intermediate and then some advanced stuff there. Lightroom as well. And boom. It took away the waveform, preserved everything else. Shavendra says, used to do After Effects, not anymore. As After Effects trying to utilize free as cost not to use. Yeah. And it happens sometimes. Aside from Blender, I don't know what would be a free alternative to After Effects. And I'm not sure that Blender would be truly comparable, if I'm being honest. But yeah, there we go. We get another one of these type of thumbnail sets out of it for a music channel. So again, it just shows you the variety of if we do different channels. we get different results and we can do all kinds of different styles and then also what the editor is capable of. Again, I think a really great example for a lot of you is literally just the product shots. The product shots I think are really healthy for a lot of you because there's some of you that you don't do talking head, you don't do animation and faces channels, but you do do product related stuff. And just realizing that if you were to want to really get strong thumbnails fast in a product niche for product review stuff, this works. And it just is so clean, it's so fast, and it works out great. It's just incredibly strong at this. And when I was looking at this for my camera channel and the wildlife stuff, I was really impressed. With just how good it can be and what it was able to come up with I mean that that's good. I feel like Feel like this one was really good. I don't think that one was bad I think this one's actually really good I'm being honest. So again, I was really impressed with the updates to the 1 of 10 thumbnail tool Cannot stress enough how far we've come with regard to these to where, again, I just feel like it borders on what I would do or could do myself in many cases in Photoshop. And I think that what you guys are seeing in terms of what's capable of, even in revisions, revisions was the one thing I was like, the AI tools are not good at yet. And now we're getting to a place to where you can prompt the revisions pretty effectively and it works out. So I think it's actually getting there. And I think this is much closer to the kind of thumbnail that you would actually even expect from me. So I think it works, if I'm being honest. So again, I could see this for most of you. Instead of the way you are struggling currently with a lot of your thumbnails, I think a lot of you would benefit from going this direction. Because I think a lot of you are struggling for an hour. or 20 minutes or 30 minutes to get a thumbnail that doesn't look nearly as good as what this could do for you. But I think the key to it, like I said, is getting the assets right, whether that's getting the graphic assets, getting the products, the physical products, and getting product shots of that, even if it's from the website, or even your personal photography. If you get your personal photography right and then you use that, that's gonna be the key, and you wanna designate and select which photo of you you want it to use. Yeah, a lot of smart people sometimes have a problem with basic things and then they're very good at advanced things I tend to be one of those people Yeah, so we'll do a little bit more Q&A and then we'll wrap this up at roughly the two-hour mark here So I'm happy to answer any and all of your questions about this. Do I have any eye drops around here? My eyes are drying out I usually keep eye drops around my desk for exactly the reason that I'm staring at a screen so long My eyes will tend to dry out, but I did not plan accordingly. I have my lens cleaner for my glasses, and yet I did not bring my eye drops for my eyes. Yeah, I agree, double pulse. If you use assets that you understand as the foundation, then you will absolutely get better results from the AI. I think everyone's problem is they overestimate AI and they underestimate it. And I think that the key, like I said, is I can do some of these things, but I can't do them nearly as fast. So it's the difference between five minutes and 50 minutes. And the difference between getting one result in 50 minutes versus four or five usable results in five minutes. I cannot compete on speed or scale. And even with talent and quality, I can't overcome speed and scale. So why would I? I would just combine the two. So I would put my effort into, okay, the photography is going to be me and it's going to make it feel real. And that's going to make it feel like, oh my God, AI, as long as it's accurately me. And then I would say on top of that, there are certain assets that I can make that the AI wouldn't even make or think of that I know that I want or that I've created. And then I can use those assets as elements. And then as long as I have that, then I'm going to get a result that nobody else is going to have because I have something custom. And then for the photography, like I said, just even having physical props for the photography on top of the photography, the posing, the props are the key. Props are the key. So there's your personality having that going for you. Those things all give you the uniqueness because, again, what people are worried about with the AI tools is generic, generic, generic. Oh, my stuff would look like somebody else's. But you can avoid all of that if you focus on personality, props, poses, and photography basics. So Miami Cloud says, what if you don't have the photo assets you have? Will Canva type stock photos work to get started with custom poses of me? You could, but. Then it becomes a little bit closer to more generic. So you want to start creating the assets that would make sense for you if you can. So you could do what you're suggesting, but those are more readily available. The quality and variety might be debatable. So you could start there to get a hang of it. But I would recommend getting closer and closer to having custom assets. And leaning heavily into the photography side of things. And when possible, commissioning custom assets. Or again, going so far as to get unique prompts from Amazon and Etsy. Because I think that matters. Because again, uniqueness. Again, Roberto and his giant pencil. It works. So I think that that helps considerably. Terms of that and then I like I said little things with the photography even the little things like knowing about the blotting wipes and knowing that that takes the shine off of Aspects of your face Knowing about the angles for the lighting if you wear glasses for the glare on the lenses just different things like that so Understanding the photography aspect of it can definitely help a lot in terms of the quality of these things That's why I wanted to lean into that Building your own portfolio of headshots and poses for your YouTube thumbnails. Having the different expressions like, you know, happy, shocked, surprised, angry, disgusted. You know, things like that. Just having a variety of facial expressions. Game changer. Just to have the assets. Yeah. So I think that that would matter. I would do a variety of body poses that you could definitely use. And then I would also do a variety of facial expressions. I would do different angles, getting your different sides. I have like hundreds, hundreds of photos when it comes to that. If you're the kind of person for whom you think wardrobe matters, you could do some wardrobe changes. Me, I have like kind of built my brand going to more to the, aside from my stage presentation, where I change up the clothing and accessories a little bit from time to time. I've gone to my Steve Jobs era of basic black turtlenecks, black long sleeves, black short sleeves, jeans, dark jeans, usually black, gray, or blue, or my business khakis, or my khaki cargo shorts, cargo pants for why I'm traveling with photo gear in my pockets. So then I'll wear that for utility purposes. So I have a basic look that I use. And that helps me just kind of keep style and continuity. Also makes it easier to cut out things in Canva and Photoshop, to be honest with you. The hats especially help with cutting stuff out there. So again, it helps. But for those of you who you know, you like to dress and dress differently and have different looks or whatever, then yeah, take a lot of different shots with your looks, but try to lock in a couple of signature looks. Like I recommend that YouTubers, this sounds crazy to take it this seriously, but just set aside a part of your wardrobe. where maybe you just have, you know, five or eight looks that you always use on camera, and these are kind of your on-camera outfits, and they're your wardrobe like a character in a TV show, and you just have kind of a wardrobe like a character in a TV show. I literally swear to God to you that in my closet, there are just 12 of these exact same black shirts all hanging up on a rack. And they're like five copies of the same black turtleneck all hanging up on a rack. And then there's my various merch and create awesome hoodies all on their own side of the closet. I swear to you that that's the case. And then after that, I just have like my jackets and I have multiple copies of the same hat. I have about to keep them fresh. And because sometimes they get dented, I have like five copies of this hat. So I literally. did exactly what they do in television shows and cartoons is I literally just made a character wardrobe and locked it in and said, okay, that's it. Now I don't have to think about it. I have going out clothes for like going out to fancy dinner, travel clothes, things for when I'm at events, when I'm not speaking on stage or for when I leave the hotel room, I have stuff for that. And then I have my suits obviously for when I need them, but for being on camera, I literally have just kind of locked in a very specific capsule wardrobe for being on camera for my YouTube content and for my thumbnails. And that just makes life easier. I found that that just makes life incredibly easier as a content creator to just do it that way. And then when I want to accessorize, it's mostly, like I said, it's mostly just jackets for me at this point. Aside from any of my wristbands, watches. um apple watches anything like that again it just it simplifies the entire thing now obviously for female content creators you might need more variety but again my suggestion for that would be a capsule wardrobe so that your swaps are a little easier and that you can even make them seasonal and so then it reduces a lot of the decision fatigue for what to wear on camera but i would also caution you that You also don't have to wear the same outfit on camera as you are in the thumbnail. I want to make that clear. People do not care. They don't. That's overthinking it. They do not care. They might not even remember by the time they click on the video, if I'm being honest. Or they might not. They might be like, oh, maybe they wear that at some other point in the video because maybe the video is filmed over more than one day. Hell no. So they don't worry about it. They don't worry about it any more than they worry about it when they turn on Netflix. I swear to you. So do not overthink that at all, especially for the female content creators. Do not overthink the continuity between the outfit or your hairstyle in the thumbnail versus in the video. Just don't worry about it. Yeah, absolutely. Like, I just believe that, like, if you're going to do anything AI, build it on a human foundation. Because, again, I just think that when we combine the authenticity and the intent. and emotional intelligence on the human side, maybe even some of the craftsmanship on the human side, but then you use the speed and scale of the AI side, that's unbeatable. I can't beat a robot on speed and scale. I just cannot. Nobody can. No one can outspeed and outscale a robot. You just aren't going to be able to do it. But you might have better taste. So if you can combine your taste with the robot's speed and scale, you're going to have a better product than either could produce alone. That's what I believe. So I think that it just makes us better. Like if I code something by myself, it might never get done. So what's the point? But if I vibe code it, it may not be perfect, but it's usable and it's there. And then if I handed it off to somebody to just clean up efficiency and security, you'd have a great MVP. So like that's where why people are vibe coding now is because Ideas don't come out fully formed. You guys remember? Aside from Stanley and a couple of y'all and Doug and a couple of y'all, a lot of you will not remember or understand how bad the original YouTube looked and how bad the functionality was. Ideas don't come out fully formed. It took well more than a decade to get to a really good usable version of YouTube. But again, now look at how fast we can get great products now. So it's about speed. scale and less cooks in the kitchen to get to the result. Lot less cooks in the kitchen to get to the result now. And that's the same for your YouTube thumbnails. It's gonna be the same soon for a lot of the editing tools. We used to have to have somebody do all this clipping and editing and take a long time and then get approvals to get short form clips out of our long form or out of our live streams. Now we all just use Opus Clip and it's great. So think about that. I mean, it's just better. Speed, scale, and then you just add some human taste and curation to it and you're fine. So again, I wanna encourage y'all and caution y'all to just not be on the AI hate train and to just understand how to use it to make yourself better, the better your workflow, to buy back your time so that you can spend more time on the human craft part. Because again, if you don't have to agonize over something, like again, one of 10 helps you not agonize over the thumbnails. Now you can be a better writer. Now you can be a better video editor and storyteller. Now you can do more research. Now you have more time to do ideation on your topics. Now you have time to do these other things that you can be good at, whereas more time won't always make you a better thumbnail artist. So that's kind of where I land with it. So again, thank you to our friends at 1 of 10 for sponsoring tonight's workshop. Again, you can check them out in the link in the description, but also Again, if it's having issues or errors, let them know. Let them know that Roberto sent you and that you're trying to check out and get the discount. And I'm sure they'll stay on top of fixing that. But other than that, y'all, it's been great hanging with you. I enjoyed answering your questions. Thank you to everybody in the chat. Thank you to all the mods for holding it down. We'll be back on Friday. Weather and God willing, you know, we'll be back on Friday. For YouTube channel reviews. I think that's gonna be really good I think it's gonna help you guys out again Our friends at 1 of 10 will be helping us out sponsoring that now you guys know like what you need to do to fix your thumbnails, right? So I think this is gonna be incredibly helpful for a lot of you So I can't wait for that again Hopefully we're all safe. The weather is good and we'll be able to do that But I will hopefully see you guys back here on Friday evening So stay awesome and we will see you next time. Take care and be safe. Becoming a content creator can be a very lonely process and adventure, but it doesn't have to be. When I first got started going from zero to 100,000 subscribers, I had to do it all by myself. I didn't have a community of like-minded and like-hearted creators to talk to or to ask for advice. And I certainly didn't have a mentor to guide me along the way. I built the Awesome Creator Academy Pro Group because when I was getting started, what would have meant the world to me is not to have to struggle alone when it comes to being a full-time content creator. If I had had a community of creators that I could talk to and share information with, I would have avoided a lot of mistakes in my early career, and there would have been opportunities with brand deals where I would have gotten a better understanding of how to negotiate, and I wouldn't have undersold myself. I wouldn't have been taken advantage of in those early situations. so in the awesome creator academy pro group myself and our co-instructor andy rivera we meet twice a week for office hours with our creators and we share information we talk about how we're building our creator businesses and we offer constructive criticism and critique on the things that we're working on for our projects we'll be part of a community of usually around a hundred like-minded creators who are taking this seriously just like you are you also have all access as a pro group member to all of our past, present, and future courses, training, templates, digital downloads, and swipe files. You have full access to our creator blueprints where you'll see our systems and the frameworks that we use to execute. You know, you always fall or rise to the standard of your systems. A lot of people really underestimate how important that is. We have digital download templates that can help you so you don't have to start with a blank canvas. These done for you, files are actually incredibly helpful especially for things like your thumbnails your media kits and anything you need for your branding as a content creator to look professional we have email swipe files that help you with reaching out to brands we have great information in helping you price yourself and build your packages appropriately and remember as part of the pro group you have the opportunity to ask questions from myself and all the other creators participating and so this means that if you need to ask questions about negotiation or a brand deal there's something you're not sure about trying to find a better monetization opportunity with merch or affiliates you have people who have done it before who are already making money that you can reach out to for advice as i said when i was starting out and i was flying solo this is the thing that i wish i'd had the opportunity to be a part of and it would have made all the difference in the world in helping me be an even more successful content creator if you have questions you can always reach out via support we're happy to help you figure out if the pro group is right for you also we do have the opportunity to become a vip member if you also would like quarterly one-on-one coaching with me directly now go out there and create something awesome today and hopefully we'll see you in the pro group take care

⚙️ Pipeline jobs

StageStatusAtt.UpdatedError
download done 1/3 2026-05-29 07:07:39
transcribe done 1/3 2026-05-29 07:10:14
summarize done 1/3 2026-05-29 07:10:58
embed done 1/3 2026-06-30 06:45:37

📄 Описание YouTube

Показать
Use 1of10 to Make YouTube Thumbnails even if you don't have design skills. In this YouTube Workshop I will show you how to use 1of10 to Create Better YouTube Thumbnails and how to use 1of10 as YouTube System for Ideation and Packaging so you can get more views on your videos.

USE CODE: Roberto   at checkout to get 20% Off for Life!
https://geni.us/get1of10com

#SPONSORED By 1of10 the Best Tool for YouTube Strategy and Ideation.