blankbrain

@synthwavedd It will forever be behind google's model cause copyright and stuff

PATRIOTS I AM PLEASED TO REPORT THIS WAS A FALSE ALARM I WAS USING A HALF BAKED VERSION THE PISS TINT IS BASICALLY DEAD 🥳🥳

@Elaina43114880 Copyright stuff will still hold it back

GPT Image V2 has improved the generation effect of pictographic characters (such as Chinese characters), making its rendering quality much better than V1.5. But this improvement is limited to just a few simple words. You can see from the following images that if you want to generate multiple sentences of Chinese characters in multiple positions, its performance will drop significantly. It can't even correctly generate the Chinese characters in a very simple comic image, so it's still not feasible to expect it to directly complete image generation containing multiple Chinese character sentences.... 😹 But! For non-pictographic characters (such as English), the situation is completely the opposite! You can now directly use GPT Image V2 to translate your favorite comic pictures into a language you're familiar with, while preserving the original text's font! The English version in the following images was first paraphrased by GPT-5.4 (so there are deviations from the original meaning), and then GPT-5.4 called GPT Image V2 to generate it. You can see that it perfectly transformed into an English translation picture! What shocks me the most is that not only the text inside the speech bubbles, but GPT Image V2 perfectly removed the artistic text and onomatopoeia in the image, and embedded the English translation in a similar font style, as if the original image was in English from the beginning! The workflow that used to require several minutes of manual translation and typesetting can now be completed in just a few seconds with a single prompt! 🙀

@synthwavedd U know what they all distill from . NBP

lmfao gpt image 2 just generated me an image with the gemini watermark in the bottom right completely unprompted

@Awk20000 he know his channel is diving headfirst with viewer rate . only his fans still watch and no new people watch really

Linus Tech Tips shows off his new private jet

@Hesamation it also mentions OAI doing same or similar

Redditor claims Claude Code is nerfed for Pro/Max users vs Enterprise customers and the strategy is to use the paid plan users to generate hype on X and LinkedIn so companies would reach out to them.

@AcerFur same image it fucks up the orange on top . Look closely

maskingtape-alpha gaffertape-alpha packingtape-alpha Seems like good image models on the image arena (try them out)... but they're not quite perfect just yet. Still fails the Rubik's Cube reflection test.

@bcherny another anti-consumer bs from anthropic

Starting tomorrow at 12pm PT, Claude subscriptions will no longer cover usage on third-party tools like OpenClaw. You can still use these tools with your Claude login via extra usage bundles (now available at a discount), or with a Claude API key.

@alibaba_cloud bs quality

Wan2.7-Video is live! 🚀 Stop gambling with AI "slot machines." 🎲 🔹 Instruction Editing: Change weather/outfits via text. 🔹 Consistency: Lock 5 subjects across shots. 🔹 Fluidity: 15s stable cinematic sequences. Don’t Re-generate, Just Edit. ✍️🎬 Try: wan.video/?cref=im&cinfo… Doc & API: int.alibabacloud.com/m/1000411584/

@synthwavedd its gonna be instantly censored to the ground so no thanks

this model is exceptional at realism and text - at long last deepmind have proper competition. it also has better world knowledge than nano banana pro/nb2 excited to see how this stacks up with nano banana pro 2 in the coming months too!

@KatieMiller cause OAI employee are mods aswell

Why did Reddit moderate this thread? 👀

@venturetwins looks like someone RL trained using Nano Banana Pro 4k

I'm really impressed by the Grok Imagine upgrade. There's a new "quality" mode where your images take (slightly) longer to generate, but the results are shockingly good. I've been testing it for the last few days - a couple things to try 👇

@Hadas_Gold i dont think controlling the narrative is side project for them

This was... 2 weeks ago? Now OpenAI is buying a daily 3-hour show. wsj.com/tech/ai/openai…

@scottstts Apple is primarily customer dependent while Anthropic is business dependent

For those who like to compare Anthropic to Apple, Apple has world’s best customer service and customer care Anthropic treat paying customers like trash and want them to beg for using their product

@tetsuoai its all bs quality

🚨 xAI just dropped a massive upgrade to Grok Imagine! Now you can choose between two modes: Quality (new): Produces next-level images with jaw-dropping detail, realism, and creativity, generating four high-quality images at once instead of an infinite scroll. Speed: Delivers lightning-fast generations with consistently impressive results, just as you've experienced before. way more power. just select a mode, then type what you want to imagine. here are two prompts that you can try. prompt one: A dramatic low-angle cinematic view of a menacing knight in intricate ornate black plate armor and helmet with flowing dark cape, riding a powerful muscular black warhorse in full gallop charging directly toward the viewer across a vast field of tall green grass. The knight raises a long straight sword triumphantly high in his right hand. The horse has a wildly flowing mane and tail whipped by motion, ornate leather-and-metal barding, and tack. Foreground shows extreme radial motion blur streaks in the grass conveying tremendous speed and velocity. Background features towering jagged snow-capped rocky mountain peaks under a vibrant deep blue sky with voluminous dramatic white clouds. Epic dark fantasy atmosphere, highly detailed textures on armor and horse, cinematic high-contrast lighting, dynamic heroic composition. prompt two: A serene young woman with fair skin, shoulder-length straight dark brown hair, and a calm introspective expression gazing slightly downward stands centered in a vast barren white sand desert at night under a deep dark navy blue sky, with distant low dark rocky mountains on the horizon. She wears a long flowing translucent white hooded robe with subtle iridescent holographic shimmer and prismatic rainbow-cyan reflections across the fabric, the hood up over her head. In both hands at chest level she holds a single glowing bioluminescent blue cross, jesus in the middle is emitting intense bright cyan-blue light that illuminates her robe, hands, and face. Directly behind her head and shoulders is a massive luminous circular rainbow halo aura with concentric glowing rings transitioning from inner vibrant blue and cyan outward to turquoise, yellow, orange, and red, crowned by a brilliant white starburst light at the top center casting dramatic volumetric god rays downward. Cinematic mystical ethereal atmosphere, photorealistic yet fantastical style, sharp focus, high detail on fabric textures, rose petals, sand ripples, and light effects.

@chatgpt21 why put your watermark ? you dont own the video clip

🚨 SAM ALTMAN ON HIS 2-YEAR AGI PREDICTION "It's possible we're only two years away from a world with more of the cognitive capacity in the world inside of data centers than outside of them." "It's easy to say 'Oh we're going to have super AGI here'... But if you say like the AI is going to do more of the total thinking than people are, that kind of like gets across the magnitude of what's happening." "In the short term... I think it may and likely will be quite disruptive to a lot of the current jobs if more of the intellectual horsepower in the world is GPUs. And I think like right now we've got to start talking about the principles for how we want to design an economy that works for everybody."

@ChrisGwinnLA @grok they tried distilling from NB

@ChujieZheng stop input censoring

Qwen3.6-Plus is now officially launched on OpenRouter, and still for free. Feel free to empower your agentic workflow with our latest flagship model!

@lydiahallie when did we ask for "bs" in the output

Peak-hour limits are tighter and 1M-context sessions got bigger, that's most of what you're feeling. We fixed a few bugs along the way, but none were over-charging you. We also rolled out efficiency fixes and added popups in-product to help avoid large prompt cache misses

Thank you to everyone who spent time sending us feedback and reports. We've investigated and we're sorry this has been a bad experience. Here's what we found:

@edzitron "Trust me bro"

In what universe is buying TBPN not a "side quest"? And what about this suggests TBPN will remain "independent"?

@KhyatiTrehan @FlowbyGoogle @KavyaTrehan a d*ck and its a asymmetrical balls ?

Walked to the park. Basked in the sun. Made a silly thing using phone pictures, Nano banana and Veo in @FlowbyGoogle Music by @kavyatrehan 🎵