You are a fish, you must escape the kitchen

cybercoupe

You can ask 4o for a depth map

I tried to see how Kling v1.6 would handle the trolley problem. But it just backed away slowly.

4o native image generation is the beginning. It's a seismic change in generative AI. Where is this all going? We won't need ipadapters, or controlnets, or loras, or comfy workflows, or face landmark models, or segmentation models, or niche task specific models. It'll be one model to rule them all. You'll only need a prompt, perhaps a reference, and your imagination. It's only going to get better, and it will apply to more and more mediums – audio and video are next. And it'll happen soon.

If you give FLUX1.1 a prompt like "IMG_1018.CR2" you get back images that are so very hard to tell they're AI. The realism here is kicked up a notch.

NO WAY. It did it. And, was that, actually funny? Prompt: > a man doing stand up comedy in a small venue tells a joke (include the joke in the dialogue)

👀

4o prompt: "make her [20, 30, 40, 50] years old"

I put some famous logos through ControlNet.

I asked Midjourney v5 to '/describe' some logos, to see how it would create prompts for them, and to see what it would create in response. Starbucks

My first Veo 3 gen > a video with dialogue of two muffins while baking in an over, the first muffin says "I can't believe this Veo 3 thing can do dialogue now!", the second muffin says "AAAAH, a talking muffin!"

These are all AI, I just don't know anymore.

Sora 2 fails this test: > a man counts out loud from 1 to 10, using his fingers and holding them up as he goes

🧵 A big #Midjourney thread on how to write prompts to get good cinematic images. In this thread I’ll build up a single prompt with cinematic elements, and show their effects. Each prompt will use a 16:9 aspect ratio, and to minimise variation I've locked in a seed.

I'm so impressed with how Wan 2.1 14b interpreted this image for the video

live view of my subconscious

Veo 3 does video games really well. It's like you can explore new worlds already. Prompts are all variations of: > a third-person open world video game walking around... > an fps video game in/on a...

🌻

> A close up video of a twitch streamer in a low lit room, in an ASMR style she whispers very closely to the camera, "Generating videos with built in dialogue really changes things, doesn't it?"

👀

Here's the first release of my consistent-character model on Replicate. It uses InstantID, IPAdapter, Controlnet and FaceDetailer, with SDXL Lightning. Prompt clothes and hairstyle for best consistency It's open source but non-commercial, all links below 👇

With #stablediffusion img2img, I can help bring my 4yr old’s sketches to life. Baby and daddy ice cream robot monsters having a fun day at the beach. 😍 #AiArtwork

Transiting the habitation ring (gpt-image-1 > PixVerse v4)

The full prompt for this is LONG: > An underwater scene stretches across the entire screen. Amidst the colorful reef, a small, rolled-up parchment map lays on the sea floor. Jerry, the brown mouse, swims calmly into the scene from the left, his large eyes wide with curiosity, carefully looking around while his tiny paws propel him forward gently. A small, rolled-up parchment map is laying on the sea floor. The map has a single small red x on it. An underwater scene stretches across the entire screen. Jerry, the brown mouse, pauses, his attention drawn to a small, rolled-up parchment map laying on the sea floor. With gentle care, Jerry grasps the delicate parchment with both paws and slowly unrolls it, examining its smooth surface intently. The map has a single small red x on it.<end_scene> <start_scene>A coral-covered underwater area stretches across the screen, rich with delicate coral formations in pastel hues gently moving with the rhythm of the ocean. Jerry, the brown mouse, swims thoughtfully toward a gently glowing sea anemone, its soft tentacles subtly pulsing with luminescence. Jerry's expression is curious and determined, his tiny paws methodically propelling him through the water. As he swims, he turns his head left and right, eyes carefully scanning for something important. The camera smoothly pans along with Jerry's steady forward swimming, highlighting his thoughtful and attentive demeanor as he continues his careful exploration of the vibrant underwater environment.<end_scene> <start_scene>Underwater, a close-up reveals Tom, the blue-gray cat, with a hungry and predatory expression, his sharp eyes narrowed as he keenly scans the surroundings. Tom turns his head methodically from left to right, squinting his eyes carefully as he searches for something, a cunning gleam evident in his gaze. Small bubbles of air gently rise behind him, highlighting his location beneath the ocean surface. In the softly swaying background, a green coral plant moves rhythmically with the water's flow. Tom is depicted naturally, free of any clothing, his sleek fur softly illuminated by the refracted sunlight filtering down from above.<end_scene> <start_scene>Underwater, a large purple coral sways gently and rhythmically from side to side in the mild ocean current. From behind the broad coral branches, Tom, the blue-gray cat, peers mischievously with a sly and calculating smile, his eyes following Jerry intently. Jerry, the brown mouse, swims slowly and obliviously past Tom, calmly navigating the serene underwater landscape without sensing any threat. The camera smoothly pans horizontally, tracking Jerry's leisurely movement across the scene, emphasizing Tom's stealthy observation, his sly grin deepening as he patiently waits for the ideal moment to act upon his mischievous intentions toward Jerry. Underwater, a large purple coral gently sways side to side, animated by the subtle ocean current. Jerry, the brown mouse, calmly swims forward through the tranquil environment. As he progresses, Jerry glances backwards casually but suddenly notices Tom, the blue-gray cat, swiftly approaching from behind with predatory intent. Jerry's eyes widen dramatically in shock and fear. Quickly regaining composure, Jerry swiftly darts to his right, frantically swimming away with anxious urgency clearly etched on his face. Tom eagerly pursues Jerry, propelled by excitement and determination, rapidly swimming toward the right with an energetic eagerness, intent on capturing the swiftly escaping mouse. An underwater kelp forest stretches upward, its tall green kelp strands swaying rhythmically and gracefully beneath the ocean's surface, forming dense vertical layers. Jerry, the brown mouse, swiftly swims into the dense kelp forest from the left, maneuvering with agility and precision. Close behind Jerry, Tom, the blue-gray cat, swims with urgency and determination, repeatedly reaching forward with his paws in swift attempts to capture Jerry. Each grasp misses narrowly, emphasizing the frantic chase. Jerry adeptly weaves and ducks between the long, gently moving kelp strands, utilizing the dense underwater foliage to skillfully evade Tom's persistent pursuit and escape imminent capture. An underwater kelp forest stretches upward, its tall green kelp strands gently swaying and undulating in a hypnotic rhythm underwater. Jerry, the brown mouse, quickly swims upward, skillfully weaving through the gently swaying kelp strands. Following closely, Tom, the blue-gray cat, swims upward rapidly, paws desperately outstretched to grasp Jerry, yet consistently failing as Jerry remains just out of reach. Jerry continues his swift, upward escape through the water, maintaining a focused and determined demeanor as he adeptly maneuvers through the thick kelp, always narrowly avoiding Tom's determined efforts to finally capture him, his movements quick and precise.<end_scene> <start_scene>Underwater, Jerry, the brown mouse, quickly darts to the left, swimming rapidly with agile and determined strokes. Tom, the blue-gray cat, follows closely behind, intensely pursuing Jerry in a heated underwater chase. Jerry swiftly manages to swim off-screen to the left, disappearing into safety beyond the visible boundary. Tom visibly slows, appearing deeply frustrated and exasperated, his expression showing disappointment and irritation as he realizes that Jerry has effectively escaped, once again eluding his grasp despite the vigorous effort and determination that characterized Tom's intense pursuit throughout this tense underwater chase sequence. Underwater, Jerry, the brown mouse, swiftly swims through vivid clusters of orange and purple coral formations, frequently glancing nervously behind him to gauge his distance from his pursuer. Following closely behind, Tom, the blue-gray cat, navigates carefully around intricate coral structures, swimming swiftly yet strategically. Tom exhibits a clear and unwavering determination in his pursuit, his eyes sharply focused on capturing Jerry. Jerry, meanwhile, continues moving with agility and speed, expertly utilizing the coral reef's complexity as cover, his eyes filled with anxious awareness and urgency as he strives to evade Tom's persistent and determined underwater chase.<end_scene> <start_scene>Underwater, outside the wooden shipwreck, Jerry, the brown mouse, swiftly swims in from the left side of the screen, his movements precise and determined. Approaching directly, he navigates skillfully toward a dark hole in the side of the shipwreck, which is covered partially by layers of coral and seaweed gently swaying with the underwater currents. Without hesitation, Jerry slips through the hole and disappears into the ship's shadowy interior. The aged wooden planks and moss-covered surface of the shipwreck enhance the mysterious atmosphere as Jerry carefully moves inside, suggesting anticipation and careful urgency in his movements as he vanishes within. Underwater, outside the wooden shipwreck, Tom, the blue-gray cat, swims swiftly into the scene from the left, showing determined urgency as he searches diligently for Jerry. With careful inspection, Tom peeks into various holes and crevices scattered along the damaged wooden hull, occasionally pausing to closely examine an opening, only to move on with visible frustration after finding nothing. Tom's expression gradually shifts from determined to confused and increasingly frustrated as his thorough search fails to reveal Jerry's hiding spot. His movements slow, becoming calmer and more deliberate as he methodically continues inspecting and searching around the shipwreck's exterior, determined yet clearly baffled.<end_scene> <start_scene>An old wooden shipwreck lies quietly underwater on the sandy ocean floor, partially submerged in sand, and overgrown by vibrant coral, seaweed, and algae. Jerry, the brown mouse, swims swiftly through a wide opening in the damaged hull, entering the shadowy interior of the wreck. Inside, Jerry pauses briefly, anxiously scanning the dimly-lit interior of the vessel, his eyes wide with cautious anticipation. He quickly retrieves a parchment map and examines it closely, verifying the marked red X location carefully. Confident but alert, Jerry carefully rolls up the map again and continues swimming deeper into the shadowy depths of the shipwreck, resolutely pursuing his goal.<end_scene> <start_scene>Inside the underwater shipwreck, Jerry, the brown mouse, swims cautiously toward a closed wooden door located in a dimly lit corridor, the interior filled with scattered debris and floating particles illuminated by beams of filtered sunlight. Jerry reaches out slowly and pushes the old wooden door open, the door creaking slightly as it reveals the mysterious room beyond. Carefully, Jerry peers inside, his cautious expression shifting instantly to one of joyful surprise. His eyes widen dramatically with excitement and wonder, clearly captivated by something extraordinary within the revealed chamber, heightening the sense of curiosity and intrigue within the scene.<end_scene> <start_scene>Jerry, the brown mouse, enters an underwater room within the shipwreck, revealing a magnificent wooden treasure chest resting prominently at its center, overflowing abundantly with glittering golden coins, shimmering jewels, and sparkling gems. Jerry eagerly checks a parchment map, excitement evident as he recognizes this spot as the treasure location marked precisely by the red X. His eyes sparkle joyfully, and a wide smile spreads across his face, clearly mesmerized and enchanted by the sight of the dazzling treasure. Jerry's excitement and delight grow visibly, captivated completely by the wealth and beauty discovered in this hidden underwater chamber.<end_scene> <start_scene>Underwater, outside the shipwreck, Tom, the blue-gray cat, swims anxiously back and forth, continuously peering with frustrated urgency into various openings and cracks in the wooden hull of the sunken ship. His expression is tense and increasingly agitated, clearly indicating a desperate search for something—or someone. Tom's eyes scan each dark opening repeatedly, his whiskers twitching nervously, as frustration grows with each unsuccessful attempt to find Jerry. His movements become more vigorous yet increasingly exasperated, capturing Tom's rising impatience and desperation as he remains intent on locating Jerry's elusive hiding place somewhere within the shipwreck's mysterious interior. Underwater, outside the shipwreck, Tom, the blue-gray cat, continues anxiously swimming back and forth, repeatedly peering into openings along the shipwreck's wooden hull with careful yet increasingly frustrated examination. His expression remains deeply troubled, his eyes narrowed in irritation and impatience as he struggles fruitlessly in his ongoing search for Jerry. Tom persistently examines each crevice methodically, visibly irritated by Jerry's successful evasion. Despite his determination, Tom's frustration is evident in his increasingly tense posture and rapid movements as he continues his thorough but futile search around the shipwreck, clearly determined yet baffled by Jerry's skillful disappearance within the underwater vessel. Underwater, outside the shipwreck, Tom, the blue-gray cat, swims back and forth with restless irritation. Suddenly, from a large opening in the shipwreck, a large silver shark emerges ominously, eyes locking directly onto Tom, the blue-gray cat. Tom, the blue-gray cat, instantly freezes, his expression rapidly shifting to sheer terror, his eyes widening dramatically and panic gripping his entire body. Reacting swiftly, Tom, the blue-gray cat, spins around frantically and begins swimming away from the large silver shark as quickly as possible, fear propelling his urgent escape. The large silver shark, clearly noticing Tom's desperate retreat, immediately begins pursuing, steadily increasing its speed and menacingly closing in behind Tom, amplifying the tension and urgency of the chase.<end_scene> <start_scene>Underwater, Tom, the blue-gray cat, frantically swims upward, his face filled with panic and fear, frequently glancing behind as a large silver shark angrily yet menacingly follows close behind. Tom, the blue-gray cat, swims with increasingly frantic movements, rapidly darting around coral formations and rocks in desperate attempts to evade capture. Each nervous glance reveals the large silver shark steadily approaching, its sharp teeth clearly visible and ominous. The large silver shark maintains an unwavering pace, patiently and deliberately closing the distance between itself and Tom, the blue-gray cat. Tom's desperation and terror are palpable, as his swift evasive maneuvers become increasingly erratic under the relentless pursuit of the dangerous predator.<end_scene> <start_scene>The screen is diagonally split in two halfs. On the right side of the split screen, inside the underwater treasure room, laughing happily, Jerry, the brown mouse, tosses handfuls of golden coins high into the surrounding water, watching them slowly drift downward, shimmering brightly and reflecting brilliantly around him. His cheerful demeanor and playful actions highlight the triumphant moment as Jerry's excitement and delight are vividly conveyed through his joyful smiles and carefree laughter. On the left side of the split screen, the background is different. The underwater in a coral reef, Tom, the blue-gray cat, is frantically swimming away from something, his face filled with panic and fear. The screen fades to black The screen is diagonally split in two halfs. On the right side of the split screen, inside the underwater treasure room, Jerry, the brown mouse, laughs happily as he swims excitedly above the treasure chest overflowing with glittering golden coins. In a joyous display, Jerry dives into the treasure, completely submerging himself within the gleaming coins, briefly disappearing beneath their shiny surface. Moments later, he playfully pops his head back out, smiling broadly with pure delight, clearly enjoying this magical moment of discovery and indulgence. On the left side of the split screen, the background is different, where, in an underwater in a coral reef, Tom, the blue-gray cat, is frantically swimming away from something, his face filled with panic and fear. The screen fades to black

Google's new ultra AI plan: $250 per month 12,000 credits 150 credits per 8s Veo 3 video So 80 videos At $3.12 per video Which is $0.39 per second of video

A new experiment mixing regional prompting and controlnet to create some smooth non-flickering animations. If folks are interested I'll put up a workflow video on YouTube.

New benchmark for "this AI video model is good at physics"

this is wild > a man in a music video raps to the camera about generating videos with Veo 3

I've worked with @kylancodes to put together a little dedicated Replicate model that gives you: - all the face images - a sprite/grid of all the faces - vanilla HTML, JS and CSS to render the effect - a preview video replicate.com/kylan02/face-l…

beast of a show

The live action Game of Thrones Mario Kart adaptation is wild. Winter is coming... so are the spiny shells.

The Sora 2 feed is fascinating, seeing how everyone is using it for the first time. This generation stood out a mile from all the cameos and pokemon, by jesperagain: > 1960s black-and-white BBC report on Sora 2 video generation model launch. Grainy video to match the time

> A man is running through a beautiful summer park at dawn, he is out of breath, he slows and stops, looks at the camera and says, while panting, "Run AI with an API. Use Replicate", then he carries on running. Then "Replicate" text fades into view at the end Seems like the model likes to add it's own captions too.

Controllism > Upscale > Pika Labs > Upscale video

> the distracted boyfriend meme but its anime

4o is the first model I've used that's been able to generate an output of what the pyramids might have looked like when they were first built. The pyramids have previously been an overfit object that's relatively unchangeable.

nano banana experiment: > make this a how to draw example, with 4 progressive changes showing the development of the minimal sketch (no text)

> a dishevelled tired looking man looks at the camera and gives a long very very fast unhinged rant about how AI is going to destroy us all – he doesn't stop to breath, just constant fast talking, he is on a busy impoverished US street in downtown, there is a lot of traffic noise and fast moving cars

A new flux fine tune I'm experimenting with

You can also try out vectorizer.ai to color vectorize an image. Thanks for the heads up @KireiStudios. 1. Use Midjourney with a prompt like: "an svg vector image of [subject], white background, logo" 2. upload to vectorizer.ai 3. get svg, eps, pdf … 🙌😍

> a sitcom scene in a 90s bar, the word "fofr" is in neon on the wall in the background. a couple say something and the audience laughs The acting in this one 🫣

Firefox

A few more, it's amazing how you get such different outputs with each seed.

Midjourney v6 + Stable Video 👌

👌

Scary levels of realism in pro ultra now at 4MP resolution, especially using raw mode plus the old IMG_2482.CR2 trick 🤯

It's remarkable how good kontext is at this, as a general purpose editing model rather than a specialised restoration one.

I had to laugh a lot at this one.

I trained an SDXL fine-tune on really bad early digital photos with poor use of flash and crappy lighting.

I had to try this. Gemini 2.0 Flash Experimental with image output 🤯

Tell me how you extrapolate from this.

Pepsi, lol

The accuracy in the details of #midjourney v4 is superb.

"make her 1,000 years old"

I can finally bring my monsters to life <maniacal laugh>

> the word “PIKA” where the shapes of all the letters combine to form a pikachu using only clever and artistic typography, beautifully designed flat art poster Imagen 3

It's a kind of magic

Mona Lisa with video-01

I am unreasonably excited that Veo 3 can make video game playthrough videos like this. > a playthrough of the original PS2 video game of GTA 3, the player runs and gets into a cybertruck and drives it away, the radio is playing a sarcastic ad

Listen to Kokoro-82M... 2m25s of speech generated in 4.5s on a T4 replicate.com/jaaari/kokoro-… replicate.com/p/k3cg51x8vdrg… So graceful, so precise.

This game was ahead of its time

Wow. This is has historically been a really hard one for image editing models to get right. > Add a photo of a woman standing in front of "Qwen" text but behind "Image Edit" text, in a stylish design

Use this sref code for an easy 80s movie look in Midjourney: --sref 2158083008 Everything is pretty 80s with this code, without trying.

Stable Video Diffusion (SVD) is now on Replicate: replicate.com/stability-ai/s… - Make 14 frame or 25 frame videos - Use any image, it'll be resized to make the best video - Control degree of movement Great work putting this together @lucataco93, and @StabilityAI for the fantastic model.

I've trained a new handwriting flux lora. It can do many styles. Prompt it with HWRIT keyword, give it some short text, a handwriting style and some ink and paper types. More examples and download links in 🧵

I'm impressed with how well Veo 3 handled this prompt: > The video depicts a half length view of a beautiful young woman with subtle brown eyes, and dark eyebrows, she is standing by a wall. A movie is being projected directly onto her. The movie is of an old and beautiful romance set in Italy. She stays still, but she tells us this is one of her favorite movies, projected onto her. (Do not show a projector)

prompt: > a google images search result for the query "meme"

😍 Woah, CoDeF (temporally consistent video processing) looks really powerful! Consistency is fantastic. Transform a derived canonical image using controlnet then apply to a whole video. Project: qiuyu96.github.io/CoDeF/ Code: github.com/qiuyu96/CoDeF Some fave examples in 🧵

🫡🪴

I think I hit a ChatGPT cache

You can ask Gemini to combine two images > Make an image where this woman is holding this toy car

I liked this take on the Spotify logo

Dang, I just can't get over how good this Stable Video Diffusion model is. replicate.com/p/i3p45mtb7hle…

I really like how creative you can be with Seedream 4. It seems to be pretty malleable.

everything is a restyle

Why prompt in JSON when you can prompt in yaml? ``` description: "Cinematic shot of a sunlit, empty kitchen. A sealed Chewy box sits in the center. It trembles, explodes open in one burst, and pet supplies rapidly assemble into place: food and water bowls, a dog bed, toys, and a bag of food. A dog runs in and flops into the bed. No text." style: cinematic camera: fixed wide angle lighting: natural warm with soft shadows room: modern kitchen with hardwood floors elements: - "Chewy box (logo visible)" - "dog food and water bowls" - " dog bed" - "dog toys (rope, ball, bone)" - "bag of dog food" - "wall hook with leash" - "dog (golden retriever)" motion: "box explodes open, dog items fly out and assemble rapidly and precisely" ending: "dog enters and settles happily into the bed" text: none keywords: - "16:9" - "Chewy" - "pet supplies" - "fast assembly" - "dog" - "no text" - "warm lighting" ```

🧵 Announcing Prompter, an open-source tool (and npm package) for generating and sharing prompts: prompter.fofr.ai/ Use it with Midjourney, Stable Diffusion, Gen2, anything. It's great for quickly exploring models. I've shared some of my fave prompts in the thread 👇

> make a 4 panel image, 2x2 showing her at 30 years old with different projections based on how she lives her life 🫠

4o image, with prompt: > Create a book cover for the tv show Severance, make it like it's from 1973 and slightly worn

A new prompt to add to the fofr-benchmark: > a man counts out loud from 1 to 10, using his fingers and holding them up as he goes > a man counts out loud from 1 to 10, "1, 2, 3, 4, 6, 7, 8, 9, 10", he counts using his fingers and holds them up as he goes

Here's a Mona Lisa character lora. I used `consistent-character` on Replicate to generate multiple poses for training. replicate.com/fofr/flux-mona… huggingface.co/fofr/flux-mon…

Experimenting with cutout prompts

face + style = become image I've now pushed "become-image" as a Replicate model and open sourced the code and ComfyUI workflow on Github. All links in the thread below.

I love using this prompt to remix images > An extremely unremarkable iPhone photo with no clear subject or framing—just a careless snapshot. The photo has a touch of motion blur, and mildly overexposed from uneven sunlight. The angle is awkward, the composition nonexistent, and the overall effect is aggressively mediocre

You can now use a character reference image when using the Minimax "video-01" (Hailuo) model on Replicate. And, dang it's sooo good. replicate.com/minimax/video-…

NO WAY DID IT JUST DO THAT replicate.com/wavespeedai/wa…

Flux redux just does character consistency out of the box. The outputs don't perfectly match the input person, but they are internally consistent, the person is the same person in all the outputs. replicate.com/black-forest-l…

Veo 3 fast: - 20 credits rather than 100 - Same resolution - 8s 720p in ~1m20 Seems fantastic after a few runs. I can't wait for the API. > A woman sits in a busy diner, she says: I was made with Veo 3 Fast, on Google Flow, video with audio but cheaper, same resolution, makes sense, seems pretty good.

> print this out and add unhinged red ink hand written annotations, doodles, scribbles, little cut outs if you want

Seriously, how good is Kontext? $0.04, 5 seconds.

flux dev prompt: the word “cat” where the shapes of all the letters combine to form a cat using only clever typography

😍

Zoom in. This is Hailuo 02

Google Veo-2 Minimax Hailuo Kuaishou Kling Alibaba Wan Luma Ray Runway Gen-3 Tencent HunyuanVideo Skywork SkyReels Pika Pika OpenAI Sora Stepfun Step-video Pixverse Genmo Mochi-1 Lightricks LTX-Video Haiper Zhipu CogVideo

You can use this prompt to make new emoji sets 🤯

🤯 > powerpoint presentation, the slide title says “Flux AI has new skills”, three bullet points, “good at text”, “prompt comprehension”, “amazing images”

Seedream 4 feels really fresh, especially in 4k mode. replicate.com/bytedance/seed… And the "IMG_2094.CR2" prompt trick works for generating random, realistic and high quality imagery. Are there any obvious tells in these?