Elon Musk, AI Leaders Issue WARNING - Labs Must PAUSE AI Experiments - Open Letter
Link to Paper: https://futureoflife.org/open-letter/...
The post argues that AI systems with human-competitive intelligence can pose significant risks to society and humanity. Although acknowledged by top AI labs and extensive research, there is a lack of planning and management in the development and deployment of these systems. The post raises questions about the potential consequences of allowing machines to flood information channels, automate jobs, and develop non-human minds that may eventually outnumber and outsmart humans.
The authors call on AI labs to immediately pause for at least six months the training of AI systems more powerful than GPT-4 and jointly develop a set of shared safety protocols for advanced AI design and development. These protocols should be rigorously audited and overseen by independent outside experts. AI developers should work with policymakers to accelerate the development of robust AI governance systems.
These governance systems should include new regulatory authorities dedicated to AI, oversight and tracking of highly capable AI systems, and liability for AI-caused harm. The authors believe that society can enjoy a flourishing future with AI if the development and deployment of these systems are managed and planned with commensurate care and resources.
22
views
Stable Diffusion 2.2 XL Is Here And It Is AWESOME! - Try It Free!
Link to try Stable Diffusion 2.2 / SDXL here: https://beta.dreamstudio.ai/ or https://beta.dreamstudio.ai/generate
The video is a demonstration of the new 2.2 version of Stable Diffusion, or what's tentatively titled XL or "SD XL." The presenter explains that the new version is much quicker and can create images a lot faster than previous versions. It can also determine the style of the image based on the prompt given.
The presenter demonstrates how to access the new version through Dream Studio AI and how to use it to generate images. The presenter shows several examples of the generated images, including a photorealistic image of an attractive woman. The presenter mentions that the new version is still in beta and can only create lower resolution 512 by 512 images. Overall, the presenter is impressed with the new version and recommends testing it out.
16
views
Canva AI - Create A Lesson Plan For Any Class In Under 2 Minutes - Detailed Tutorial
The video tutorial demonstrates how to use Canva's new Magic Write AI functionality to create lesson plans quickly and easily. The presenter mentions that if you're a teacher or professor who wants to create a lesson plan without having to write it out and go through the process, then this tutorial is for you.
The presenter shows how to create a lesson plan in under a minute using Canva. First, go to Canva.com and click on "Create a Design" and then select "Docs" to create a design doc. Next, click on "Add Magic" and selects "Magic Trick."
The presenter writes in the lesson plan topic, such as "Grade 11 lesson plan about Macbeth and the three," and then clicks enter. After a few seconds, the presenter is presented with a lesson plan template that includes the title, objective, materials needed, procedure, assessment, and citations. The presenter notes that you can edit and adjust the template as needed. The presenter concludes by stating that Canva is an incredible tool, especially for teachers, and hopes that this tutorial helps.
14
views
1 CLICK A.I. Video Stylization IS HERE! Turn Any Video Into Claymation, Watercolor, Charcoal, Etc!
Link to RunwayML: https://runwayml.com
In this video, the presenter demonstrates how to use the free website Runway ML to transform any video into claymation, using pre-made presets. They show the steps of uploading a video, selecting a preset, and adjusting the strength of the effect.
The presenter also mentions other preset options, such as watercolor, origami, pen and ink, charcoal sketch, and cloud scape. They conclude by showing the results of the claymation effect on a boring real estate video and offer viewers the option to download the transformed video.
13
views
AI Text To Video Is GOOD And FREE Now! Create Free AI-Generated Videos In 2 Minutes!
Plazmapunk Link: https://www.plazmapunk.com/app
In this video, I show how to create a free text-to-video AI video using a free online tool called Plasma Punk AI. The tool allows users to create a 60-second video with customizable music and styles and provides a selection of pre-built options to choose from. Users can also upload their own music and choose from a variety of styles such as architecture, painting, and cartoon.
5
views
Create Stunning VECTOR ART from Midjourney V5 Images - Step-by-Step Tutorial
Link to Vectorizer AI: https://vectorizer.ai/
In this YouTube video, the creator demonstrates how to turn an image or a piece of art generated in Midjourney V5 into vector art, all for free and without using Adobe Illustrator.
The creator first shows an example of a bulldog logo type created in Mid Journey, which is then vectorized in Adobe Illustrator, resulting in all the paths being visible in the Layers panel. To create vector art, the creator first saves the image they want to use onto their desktop, then goes to a website called Vector Riser (in beta, currently free as of April 2, 2023), which can turn images into vector graphics.
The creator drops the image onto Vectorizer AI's website - https://vectorizer.ai/ - which quickly and accurately turns the image into vector art. Finally, the creator checks that the resulting file is an SVG (scalable vector graphic) file, and demonstrates in Adobe Illustrator how the art is now completely movable and editable using handles.
4
views
How To Vectorize Any Image In Seconds - Free AI Tool
Link to Vectorizer AI - https://vectorizer.ai
Ever wondered how to take a regular image - either a .jpg or .png and turn it into vector art? In this tutorial we take a regular image and turn it into vector art with vectorizer.ai
10
views
Midjourney V5 New "Describe" Feature - Determine What Prompt Was Used and Style Match Any Image
In this video, you will learn how to use a new feature in Midourney Version 5 to reverse engineer, image to text, the prompt for any image or style. By using the "describe" command, you can upload any photo and it will generate a set of prompts that could be used to create an image in a similar style. In this tutorial, the example used is a famous image taken at the end of World War Two, which has become an iconic image of the 20th century.
The video takes you through the process of generating prompts using the "describe" command and how to use the "Imagine this" and "Remix" boxes to make changes to the prompts and generate new images. The video shows you how to select the prompts that you like and create images in black and white or in color.
The video also shows you how to make changes to the prompts by adding or deleting aspects that you like or don't like. This is a great feature that can help you create images in the same style as your favorite photos or art. It's a must-watch for anyone who is interested in art or photography and wants to learn how to create images in the same style as their favorite artists.
/describe" is a new feature in Midjourney V5, a platform that allows users to generate images using artificial intelligence. When using "/describe", users can drag and drop an image and the platform will provide several possible prompts that can be used to generate an image in a similar style.
The feature can be used to reverse engineer the prompt for any style, whether it's a photograph or an artwork. Users can make changes to the generated image using the "Imagine this" box, such as deleting or adding elements and adjusting the aspect ratio.
Once the user is satisfied with the generated images, they can use the "Remix" prompt to make further modifications and generate a final image. Overall, "/describe" is a powerful tool that allows users to explore different creative possibilities and generate unique images.
8
views
Midjourney V5 - How to Find Job ID and Seed Number - 2 Methods
In this video I show you two methods to find the Job ID and Seed Number in Midjourney Version 5.
4
views
Midjourney V5 New Feature - Permutations - Detailed Tutorial And Example Prompts
Example Prompts:
french bulldogs running in a race, kinetic photography style --ar {16:9, 9:16, 3:2, 2:3}
french bulldogs running in a race, kinetic photography style {natural, artificial, dramatic, foggy} lighting
french bulldogs running in a race, kinetic photography style --v {2,3,4,5}
french bulldogs running in a race, kinetic photography style --chaos {0, 30, 60, 90}
french bulldogs running in a race, kinetic photography style --stylize {0,25,50,100}
{art deco, abstract} {dadaism, futurism} french bulldogs running in a race, kinetic photography style
========
In this video tutorial, the speaker discusses a new feature recently released by Midjourney which allows users to run multiple variations of a single prompt using permutations. This is currently available for pro subscribers of Midjourney Version 5.
The video starts with an introduction and a brief explanation of the new feature. The speaker then proceeds to give several examples of how to use permutations to enhance the journey experience.
In the first example, the speaker shows how to use permutations to change the aspect ratio of a prompt. They use the prompt "French bulldogs running in a race because they're cute and potato-like kinetic photography style" and add permutations using the markup language within curly braces to create four different aspect ratios. The speaker then shows the prompt being run four times with each variation using a different aspect ratio.
In the second example, the speaker shows how to use permutations to add different styles of lighting to a prompt. They use the same prompt as before and add permutations using the markup language within curly braces to create four different lighting styles. The speaker then shows the prompt being run four times with each variation using a different lighting style.
In the third example, the speaker shows how to run a prompt in four different versions of Midjourney V5 at the same time. They use the same prompt as before and add permutations using the markup language within curly braces to specify which versions to use. The speaker then shows the prompt being run four times with each variation using a different version of Midjourney.
The speaker ends the video by showing an example of how to use permutations to stylize a prompt. They use the same prompt as before and add permutations using the markup language within curly braces to create four different stylizations. The speaker then shows the prompt being run four times with each variation using a different level of stylization.
Overall, the video is a comprehensive guide to using permutations in Midjourney Version 5 to enhance the journey experience.
19
views
Meta Joins The AI Race With Segment Anything (SAM) - Cut Out Any Object From Any Image - 1 Click
Link to Segment Anything: https://segment-anything.com
In this video, I introduce a new tool called Segment Anything or Sam, created by Meta AI - the company that makes Facebook. It allows users to cut out anything from an image by segmenting all the different parts of the image so that they can be isolated. Viewers can follow along by going to segmentanything.com and clicking on the demo. They can either use one of the pre-built images or upload their own image.
I upload an image of hipsters doing hipster things, and the tool starts to identify and select the different objects and subjects in the image. I demonstrate how to cut out an object by clicking on it and using the Cutout Object feature.
I also show viewers how to use the Click on Everything feature, which scans the entire image and selects all the different objects, allowing users to cut out all of them.
9
views
Midjourney V5 Tutorial: How to Use the REPEAT Feature (-- repeat) to Run MULTIPLE Batches Quickly
The video is a tutorial on a new feature called "repeat" on Midjourney V5. The host mentions that the feature is currently available for Pro Plans and will soon be rolled out to standard plans. He then proceeds to demonstrate how to use the feature with an example. He types in "English Bulldog running using a kinetic photography style" and adds "--repeat 5" to indicate that he wants to run five batches of this prompt simultaneously.
The host then introduces another feature called "chaos," which allows for experimenting with different image results by setting the level of randomness. He enters "--chaos 100," indicating the maximum level of randomness. The video shows the prompt being processed and processed five times, and the results are different each time due to the use of the chaos feature.
The host concludes by highlighting the power of the repeat function, which enables running many different prompts simultaneously. He encourages viewers to use the feature and reminds them that it requires fast mode rather than relaxed mode. The video ends with a summary of the commands used to access the repeat feature.
6
views
The Future of VFX is Here: Meet the 1-CLICK AI That Will Transform the Industry! Wonder Dynamics
The video shows a demo of a new technology by Wonder Dynamics that enables replacing a live-action actor with CGI in a scene with just a single click. The technology is browser-based and easy to use, making it accessible to anyone who wants to use it.
The video shows the process of uploading a video, selecting an actor to replace, and choosing a CGI template to replace them with. The technology automatically identifies the actor's position and creates a mask around them.
The user can then choose from various CGI templates to replace the actor, and the technology will render out the scene with the new CGI actor in place. The demo shows that the technology is in beta testing and will be further developed to offer more features like camera tracking, alpha masks, and mocap Blender.
13
views
How I Created 40 Different Image Variations From a SINGLE PROMPT - Midjourney V5
Use this at the end of your prompt:
--chaos 100 --repeat 10
--------------------------------------
This is a tutorial on how to use the chaos feature in Midjourney Version 5 to generate 40 unique variations from a single prompt. First, type in a general prompt, for example, "cute Bulldog", to ensure that the variations will be diverse. They then introduce the chaos feature by typing in "--chaos 100" which instructs the program to apply maximum chaos with no restrictions. I use the "--repeat 10" command to run the prompt 10 times to generate 10 separate 4x4 grids with maximum chaos applied to each (40 total unique variations).
After clicking the green box to agree to run the prompt 10 times, I showcase the 40 variations that were generated, which include French bulldogs, English bulldogs, puppies, and mixes of various breeds. I also run a more specific prompt, "walking mech robot with hexagonal weapons walking in a cyberpunk wasteland", to show that the chaos feature can work with specific prompts as well.
The variations generated from the specific prompt include different types of robots in different settings, showcasing the flexibility of the chaos feature.
8
views
Activate Anime GOD MODE With Midjourney Niji 5 - Detailed Tutorial + Prompts Included
Prompts
Prompt 1: (SUBJECT) portrait, saturated plain colors, american scene matte painting, matte drawing, detailed, by artgerm and skottie young --niji 5
Prompt 2: a girl in traditional japanese clothing, with white wolf beside her, standing in front of a magical torri shrine, anime, Don Bluth, artgerm , by Studio Ghibli — ar 3:2
Prompt 3: (Subject) in the style of (Anime Movie, Show)
E.G. a strong prince in the style of Ghost in The Shell
===============================
In this YouTube video, the presenter shows viewers how to activate the new Niji 5 Midjourney mode, --niji 5, in Mid-Journey version 5. The presenter claims that Niji 5 is an excellent feature and demonstrates its capabilities by showcasing images created using the new mode. The tutorial covers how to activate Niji 5, including two methods to do so, one being typing in --niji 5 and the other through settings.
The presenter then provides viewers with an example prompt that they can copy and paste into Midjourney to create content in the same style. The prompt is for a warrior or prince portrait that is saturated, plain-colored, American scene, matte painting, matte drawing, and detailed. The presenter runs the prompt to showcase the results, and viewers can see how well the new mode has created the portrait.
Next, the presenter provides viewers with another prompt that they can copy and paste into MidJourney Niji 5. This prompt is for a girl in traditional Japanese clothing with a white wolf standing in front of a magical Tory shrine in an anime Don Bluth style. The prompt also includes artists such as Art Germ and Studio Ghibli, two renowned names in the market. By running this prompt, viewers can see how well Niji 5 creates content in this particular style.
10
views
How To Create Your Own (Free) Stock Photos With Midjourney V5 and ChatGPT
ChatGPT Prompt: Here is a prompt formula for creating stock images for Midjourney Version 5: A photograph of a [subject] [engaged in an action] with [background context], during [time of day] with [type of lighting] and shot with a [type of lens], shot with a [type of camera], using [name of lens]
Write "I understand" if you understand this formula
=============================================
The video is a tutorial on how to create customized stock images using Mid-Journey version 5 and Chat GPT, a free chatbot by OpenAI. The tutorial explains how to use a prompt formula to create a specific image with a subject engaged in an action, background context, time of day, type of lighting, lens, and camera.
The host gives an example of a prompt for a woman working in an office chair reading a tablet, with a modern office background, shot with a wide-angle lens, Canon EOS R6 camera, and F/2.8 24-70mm USM lens. The tutorial also explains how to adjust the prompts and create variations of the final image.
The goal is to create customized, high-quality stock images that are exactly what you want, saving money and avoiding boring, creepy, or irrelevant images.
7
views
Transforming Logo Design with AI and MidJourney - Step-By-Step Tutorial
Free AI Vectorizer: https://vectorizer.ai/
Free (Optional) Background Remover: https://www.photopea.com/
----------------------
This tutorial video is all about creating professional logos with the help of Midjourney V5 and a free AI tool called Vectorizer AI. The creator of the video promises to teach you how to become the best logo designer using free tools. The process of creating logos is demonstrated with an example, where an image of a logo is taken from the Midjourney website and then it is dragged and dropped into Photoshop. The background of the logo is removed and the image is put into Vectorizer AI to create a vectorized result. The final result is then thrown into Adobe Illustrator to prove that it is vector art.
The video then moves on to a step-by-step process for creating logos from scratch. The first step is to type in /imagine in Midjourney and then hit enter. The creator then proceeds to explain how to decide what type of logo to make, and in this case, a zombified logo of a black and green bear's head designed in an E-sports illustration style. The logo is then copied and pasted into Midjourney and the options available for the logo are explored.
The creator clicks on the top left logo and shows how it can be upscaled and variations can be made. For this video, the logo is deemed satisfactory and saved to the desktop. The creator then heads over to photopr.com, which is free software to make further changes to the logo. The process is demonstrated step-by-step and you are shown how to save the logo.
8
views
Blue Willow - How To Find Your Images / Create Your Own Discord Server
Welcome to another Blue Willow tutorial! In this video, I will show you how to find your artwork even if it's zooming past you on the screen. I'll cover two things: first, I will demonstrate how to find your latest art by clicking on the inbox and selecting mentions. Second, I'll show you how to create your own server and add the Blue Willow bot to it so you never have to deal with this problem again.
By creating your own private server with your own private Blue Willow, all the craziness ends today. To get started, click the plus button and add a server. You'll have a few options, but I recommend creating your own server and giving it a unique name.
Once you've created your server, go back to Blue Willow and scroll down. Hover over Blue Willow with the bot and you'll see a checkmark. Click on the bottom left, select "Add to Server," and choose the Blue Willow server you just created. Authorize it to access your Discord account and you're done! Blue Willow has been added to your server.
Now you can generate art or images inside your custom server by simply typing in a prompt like "imagine French bulldog" and hitting enter. It's that easy! Thanks for watching this tutorial and we hope this helps you solve your artwork finding problem.
7
views
Create Jaw-Dropping Fantasy Art with Midjourney V5 and ChatGPT - Step By Step Tutorial
Chat GPT Prompt:
Here is a formula for fantasy style Midjourney Prompt.
A [fantasy art style] illustration of a [majestic adjective] [subject] [engaged in an action scene] with [fantasy background] in the style of [fantasy artist] view from [perspective] --ar 16:9
Do you understand?
===========================================
Massive Fantasy Art Prompt Ideas (List)
Fantast Art Styles: High Fantasy, Low Fantasy, Gothic, Steampunk, Urban Fantasy, Surreal Fantasy, Comic Book, Anime/Manga, Dark Fantasy, Mythic Fantasy, Historical Fantasy, Retro Fantasy, Heroic Fantasy, Fairytale Fantasy, Epic Fantasy, Contemporary Fantasy, Cyberpunk Fantasy, Magical Realism.
Majestic Adjectives: Regal, Majestic, Glorious, Splendid, dreamy, Magnificent, Resplendent, Exalted, Grandiose, Noble, Sublime, Heroic, Legendary, Mythical, Divine, Heavenly, Celestial, Radiant, Ethereal, Transcendent, Imposing, Monumental, Spectacular, Ornate, Opulent, Luxurious, Iconic, Empyrean, Monumentous, Heroical, Majestical, High-flown, Imperial, Sovereign, Splendiferous, Kingly, Noble-minded, August, Stately, Exquisite, Dignified, Illustrious, Grand, Royal, Commanding,
Subject: Dragon, Elf, Dwarf, Fairy, Mermaid, Unicorn, Giant, Griffin, Centaur, Wizard, Sorcerer, Witch, Warlock, Demon, Angel, Ghost, Zombie, Vampire, Werewolf, Shape-shifter, Chimera, Kraken, Cyclops, Goblin, Orc, Troll, Hydra, Phoenix, Basilisk, Minotaur, Satyr, Harpy, Nymph, Siren, Banshee, Specter, Genie, Djinn
Action scene: Fighting, Fire-breathing, Spell-casting, Bow-wielding, Hammer-wielding, Sword-wielding, Shield-wielding, Stealthy, Charging, Swooping, Conquering, Summoning, Riding, Entangled, Enraged, Erupting, Summoned, Defending, Haunting, Rampaging, Hunting, Transforming, Rampant, Attacking, Throwing, Axe-wielding, Smashing, Exploding, Firing.
Fantasy background: Mystical forests, Enchanted gardens, Medieval castles, Gothic cathedrals, Desert wastelands, Snowy mountain peaks, Alien planets, Underwater worlds, Steampunk cities, Post-apocalyptic landscapes, Magical portals, Otherworldly dimensions, Mythical lands, Dark dungeons, Fabled cities, Fairy tale cottages, Haunted mansions, Lunar landscapes, Celestial skies.
Fantasy artist: John Howe, Alan Lee, Brian Froud, Michael Whelan, Frank Frazetta, Boris Vallejo, Julie Bell, Luis Royo, Kinuko Y. Craft, Donato Giancola, Todd Lockwood, Larry Elmore, Keith Parkinson, Brom, Chris Achilleos, Greg Hildebrandt, Tim Hildebrandt, Darrell K. Sweet, Ted Nasmith, Dan Dos Santos, Vance Kovacs, Greg Ruth, Sam Weber, Bastien Lecouffe Deharme, Karla Ortiz, Annie Stegg Gerard, Peter Mohrbacher, Tommy Arnold, Richard Anderson
Perspective: Close-up, Wide-angle, Overhead, High-angle, Low-angle, Bird's eye view, Worm's eye view, Point-of-view, Oblique, Foreshortened, Dutch angle, Split screen, Freeze-frame, Slow motion, Tracking shot, Panning shot, Zoom in, Zoom out, Follow shot, Crane shot
========================================================
This YouTube video is a tutorial on creating amazing fantasy art and illustrations using Midjourney version five and Chat GPT. The creator shows off some of their previous works, which include dark elves, spider queens, under dark, wizards, dragons, and battles.
They then credit Paul DelSignore for inspiring this tutorial. The creator explains that they will be using a specific formula for a fantasy-style Midjourney v5 prompt that involves a majestic subject engaged in an action scene with a fantasy background and a specific viewpoint. They copy and paste the formula into Chat GPT and ask it to give them a prompt for a dragon fighting a wizard.
ChatGPT provides a prompt, which the creator copies and pastes into Midjourney, and they modify it as necessary.
108
views
Level Up Your Anime Art Game: New Midjourney V5 Niji Feature EXPRESSIVE STYLE - Tutorial
Niji Bot link: https://nijijourney.com/en/
In this tutorial I show you how to activate the new Niji 5 Expressive Style functionality for midjourney v5. I show you how to add Niji to your own discord server.
Niji 5 - Expressive style has more realistic eyes, high chroma, and more mature feelings. I run 5 random prompts to compare and contrast the regular Niji 5 to the new Niji 5 Expressive Style
8
views
Midjourney V5 - How To Emulate ANY Photographer's Style With One Prompt + Reference Sheet
Prompt "In the style of" or "photo by (Name of photographer)" or "portrait photo by ____"
This YouTube video is a tutorial on how to create images or photos in any photographer's style with Midjourney and how to use the MidJourney version five Photographer's Reference Sheet. The video starts with the presenter introducing themselves and explaining what the tutorial is about. They explain that they will show how to create images in a photographer's style and how to use the reference sheet.
The presenter then demonstrates two techniques. Firstly, how to create an image from scratch in a particular photographer's style, and secondly, how to take an existing image and apply a style to it using the MidJourney. They use an example of an Asian lady reading a book and show how to apply the Brandon Waffle style to it.
The presenter also shows the reference sheet, which contains over 370 of the best photographers of all time, and how to use it to select a photographer's style. They explain that it is easy to use and should only take a couple of minutes at most.
7
views
Create Amazing VECTOR Logos With Midjourney V5 - Step By Step Tutorial
Link to Free AI Vector Tool: https://vectorizer.ai/
In this quick tutorial I show you how to design a logo in Midjourney. I then show you how to upscale it and create variations (if needed). From there, we put the image into a vectorizer called "Vectorizer AI" and we turn the image into vector art / vector logo.
Finally, we download the SVG File (Scalable Vector Graphic) and that's it, we have a professional vector logo!
4
views
Try AutoGPT For FREE - Create Your Own NO CODE AI Agent - Detailed Example
Link To Godmode: https://godmode.space/
In this video I show you how to access and use AutoGPT for free using a website called Godmode.space.
We go through a detailed example of how to create an AI agent with Godmode.
6
views
Midjourney V5 - The Secret To Creating CONSISTENT Characters - Step By Step Tutorial
Link to Reference Website: https://www.midlibrary.io/styles
Ever wondered how to make a consistent character in Midjourney version 5? Creating characters that look the same in multiple expressions and poses is easy with these steps.
6
views
AutoGPT In 4 Minutes - Is This AGI? Try It For FREE In Your Browser! Step By Step Tutorial
Link to AgentGPT: https://agentgpt.reworkd.ai/
Welcome to our latest video where we dive into the world of AutoGPT and explore its incredible AI capabilities, all in under 4 minutes! In this video, we'll demonstrate the powerful browser-based version of AutoGPT on AgentGPT, which is free to try.
Follow along as I set up a personalized AI agent and witness it create a joke maker right before your eyes!
Timestamps:
00:00 - Introduction to AutoGPT on AgentGPT
00:25 - Setting up a personalized AI agent
01:01 - Creating a joke maker with AutoGPT
01:20 - Witnessing the AI agent in action
02:04 - Exploring the tasks performed by the AI agent
02:51 - Understanding the importance of setting a hard limit on API usage 03:07 - Reviewing the progress and discussing potential improvements
Join me on this exciting journey as I (try) unlock the potential of AI AutoGPT. Don't forget to like, comment, and subscribe for more AI and tech-related content! 🔥👍
11
views