Sabitlenmiş Tweet
Benny
300 posts
Benny
@hemzadev
AI Guy | Exploring AI and how it’s transforming the world.
Miami, FL Katılım Ekim 2024
90 Takip Edilen97 Takipçiler
@stitchbygoogle Another reason why we love you @Google @_davideast, solving the real problems we actually face. DESIGN.md is going to save us so many hours of "guessing the hex code." Amazing release!
English
Today, we’re open-sourcing the draft specification for DESIGN.md, so it can be used across any tool or platform. We’re also adding new capabilities.
DESIGN.md lets you easily export and import your design rules from project to project. Instead of guessing intent, agents know exactly what a color is for and can even validate their choices against WCAG accessibility rules.
Watch David East break down this shared visual language in action👇. New capabilities and links in 🧵
English
Just generated this App Store-style concept with ChatGPT Images 2.0
x.com/OpenAI/status/…
OpenAI@OpenAI
Introducing ChatGPT Images 2.0 A state-of-the-art image model that can take on complex visual tasks and produce precise, immediately usable visuals, with sharper editing, richer layouts, and thinking-level intelligence. Video made with ChatGPT Images
English
Benny@hemzadev
I’ve been testing the new ChatGPT 2.0 image model and honestly… it feels a bit scary. The biggest difference is not just “better quality”, it’s consistency. On the first try, the model already keeps shapes clean, text readable, and details intact. No more weird distortions, no melting objects, no random artifacts that ruin the whole image. Before, you had to generate 10–15 versions just to get something usable. Now, you can actually get a solid result almost immediately. Where this really hits is in product mockups and design exploration. You can quickly test ideas, visualize concepts, and iterate without losing time fighting the model. It’s not perfect, but it crossed a line. This is starting to feel like a real tool in a design workflow, not just something you play with. Definitely worth trying if you’re into UI, branding, or product design. x.com/OpenAI/status/…
QME
Just tried gpt-image-2.
It is really good. OpenAI is finally leading the image gen again.
OpenAI@OpenAI
Introducing ChatGPT Images 2.0 A state-of-the-art image model that can take on complex visual tasks and produce precise, immediately usable visuals, with sharper editing, richer layouts, and thinking-level intelligence. Video made with ChatGPT Images
English
Benny@hemzadev
I’ve been testing the new ChatGPT 2.0 image model and honestly… it feels a bit scary. The biggest difference is not just “better quality”, it’s consistency. On the first try, the model already keeps shapes clean, text readable, and details intact. No more weird distortions, no melting objects, no random artifacts that ruin the whole image. Before, you had to generate 10–15 versions just to get something usable. Now, you can actually get a solid result almost immediately. Where this really hits is in product mockups and design exploration. You can quickly test ideas, visualize concepts, and iterate without losing time fighting the model. It’s not perfect, but it crossed a line. This is starting to feel like a real tool in a design workflow, not just something you play with. Definitely worth trying if you’re into UI, branding, or product design. x.com/OpenAI/status/…
QME
For some reason, gpt-image-2 really sucks at generating images of Sam
English
Benny@hemzadev
I’ve been testing the new ChatGPT 2.0 image model and honestly… it feels a bit scary. The biggest difference is not just “better quality”, it’s consistency. On the first try, the model already keeps shapes clean, text readable, and details intact. No more weird distortions, no melting objects, no random artifacts that ruin the whole image. Before, you had to generate 10–15 versions just to get something usable. Now, you can actually get a solid result almost immediately. Where this really hits is in product mockups and design exploration. You can quickly test ideas, visualize concepts, and iterate without losing time fighting the model. It’s not perfect, but it crossed a line. This is starting to feel like a real tool in a design workflow, not just something you play with. Definitely worth trying if you’re into UI, branding, or product design. x.com/OpenAI/status/…
QME
Benny@hemzadev
I’ve been testing the new ChatGPT 2.0 image model and honestly… it feels a bit scary. The biggest difference is not just “better quality”, it’s consistency. On the first try, the model already keeps shapes clean, text readable, and details intact. No more weird distortions, no melting objects, no random artifacts that ruin the whole image. Before, you had to generate 10–15 versions just to get something usable. Now, you can actually get a solid result almost immediately. Where this really hits is in product mockups and design exploration. You can quickly test ideas, visualize concepts, and iterate without losing time fighting the model. It’s not perfect, but it crossed a line. This is starting to feel like a real tool in a design workflow, not just something you play with. Definitely worth trying if you’re into UI, branding, or product design. x.com/OpenAI/status/…
QME
Benny@hemzadev
I’ve been testing the new ChatGPT 2.0 image model and honestly… it feels a bit scary. The biggest difference is not just “better quality”, it’s consistency. On the first try, the model already keeps shapes clean, text readable, and details intact. No more weird distortions, no melting objects, no random artifacts that ruin the whole image. Before, you had to generate 10–15 versions just to get something usable. Now, you can actually get a solid result almost immediately. Where this really hits is in product mockups and design exploration. You can quickly test ideas, visualize concepts, and iterate without losing time fighting the model. It’s not perfect, but it crossed a line. This is starting to feel like a real tool in a design workflow, not just something you play with. Definitely worth trying if you’re into UI, branding, or product design. x.com/OpenAI/status/…
QME
This is the single prompt I used:
Create a futuristic glassmorphism user interface designed for an augmented reality headset (similar to Apple Vision-style spatial computing), floating seamlessly in a real-world environment. The UI should appear as a transparent, curved holographic panel suspended in mid-air, with soft depth and subtle perspective distortion to simulate spatial presence.
The interface represents a premium streaming platform (Netflix-inspired), featuring a modern layout with sections like trending content, recommendations, continue watching, and categories. All UI elements should feel immersive and spatial, slightly layered in depth, with foreground and background separation.
Use a glassy, translucent aesthetic with blurred backgrounds (frosted glass effect), smooth gradients, and subtle reflections that react naturally to ambient lighting. The UI should integrate perfectly with the environment behind it, allowing the real-world scene to softly bleed through the interface.
Lighting should be cinematic and ambient, with soft shadows, gentle highlights, and realistic reflections on the glass panels. The color palette should lean toward dark mode with deep blacks, charcoal tones, and hints of neon accents (electric blue, soft purple, or teal) to emphasize interactivity and futuristic design.
Typography should be clean, minimal, and highly legible, with a premium tech feel. UI elements like buttons, cards, and thumbnails should have soft rounded corners, subtle elevation, and micro-glow effects to suggest interactivity.
Add depth by layering UI cards in a slightly curved layout, as if wrapping around the user’s field of view. Include hover-like states, glowing focus indicators, and soft motion blur hints to imply interaction without showing actual movement.
The background environment should be softly blurred (bokeh effect), suggesting a modern living room or workspace, reinforcing the mixed reality experience. Ensure the UI feels anchored in space with realistic perspective, reflections, and light interaction.
Overall style should be ultra-modern, minimal, and immersive, combining glassmorphism, spatial UI design, and cinematic rendering, with close attention to detail, realism, and premium product design aesthetics.
English
I’ve been testing the new ChatGPT 2.0 image model and honestly… it feels a bit scary.
The biggest difference is not just “better quality”, it’s consistency.
On the first try, the model already keeps shapes clean, text readable, and details intact. No more weird distortions, no melting objects, no random artifacts that ruin the whole image.
Before, you had to generate 10–15 versions just to get something usable.
Now, you can actually get a solid result almost immediately.
Where this really hits is in product mockups and design exploration.
You can quickly test ideas, visualize concepts, and iterate without losing time fighting the model.
It’s not perfect, but it crossed a line.
This is starting to feel like a real tool in a design workflow, not just something you play with.
Definitely worth trying if you’re into UI, branding, or product design.
x.com/OpenAI/status/…
OpenAI@OpenAI
Introducing ChatGPT Images 2.0 A state-of-the-art image model that can take on complex visual tasks and produce precise, immediately usable visuals, with sharper editing, richer layouts, and thinking-level intelligence. Video made with ChatGPT Images
English
ChatGPT just killed half the design industry overnight.
This new image model is not normal.
• photorealism is scary good
• characters stay consistent (!!)
• it actually understands what you mean
• text in images finally works
This isn’t Midjourney vibes anymore.
This is production-level output.
Startups don’t need designers for 80% of tasks now.
Just vision + prompts.
If you’re not using this yet… you’re already behind.
English
$META TO INSTALL TRACKING SOFTWARE ON U.S. EMPLOYEE COMPUTERS TO CAPTURE WORKFLOW DATA FOR AI TRAINING -INTERNAL MEMO
META TRACKING TOOL TO CAPTURE MOUSE MOVEMENTS, KEYSTROKES AND SNAPSHOTS OF WHAT EMPLOYEES SEE ON THEIR SCREENS -INTERNAL MEMO
English
Meet Kimi K2.6: Advancing Open-Source Coding
🔹Open-source SOTA on HLE w/ tools (54.0), SWE-Bench Pro (58.6), SWE-bench Multilingual (76.7), BrowseComp (83.2), Toolathlon (50.0), Charxiv w/ python(86.7), Math Vision w/ python (93.2)
What's new:
🔹Long-horizon coding - 4,000+ tool calls, over 12 hours of continuous execution, with generalization across languages (Rust, Go, Python) and tasks (frontend, devops, perf optimization).
🔹Motion-rich frontend - Videos in hero sections, WebGL shaders, GSAP + Framer Motion, Three.js 3D.
🔹Agent Swarms, elevated - 300 parallel sub-agents × 4,000 steps per run (up from K2.5's 100 / 1,500). One prompt, 100+ files.
🔹Proactive Agents - K2.6 model powers OpenClaw, Hermes Agent, etc for 24/7 autonomous ops.
🔹Claw Groups (research preview) - bring your own agents, command your friends', bots & humans in the loop.
-
K2.6 is now live on kimi.com in chat mode and agent mode.
For production-grade coding, pair K2.6 with Kimi Code: kimi.com/code
-
🔗 API: platform.moonshot.ai
🔗 Tech blog: kimi.com/blog/kimi-k2-6
🔗 Weights & code: huggingface.co/moonshotai/Kim…
English
Daily model releases
Daily security incidents
GIF
Kimi.ai@Kimi_Moonshot
Meet Kimi K2.6: Advancing Open-Source Coding 🔹Open-source SOTA on HLE w/ tools (54.0), SWE-Bench Pro (58.6), SWE-bench Multilingual (76.7), BrowseComp (83.2), Toolathlon (50.0), Charxiv w/ python(86.7), Math Vision w/ python (93.2) What's new: 🔹Long-horizon coding - 4,000+ tool calls, over 12 hours of continuous execution, with generalization across languages (Rust, Go, Python) and tasks (frontend, devops, perf optimization). 🔹Motion-rich frontend - Videos in hero sections, WebGL shaders, GSAP + Framer Motion, Three.js 3D. 🔹Agent Swarms, elevated - 300 parallel sub-agents × 4,000 steps per run (up from K2.5's 100 / 1,500). One prompt, 100+ files. 🔹Proactive Agents - K2.6 model powers OpenClaw, Hermes Agent, etc for 24/7 autonomous ops. 🔹Claw Groups (research preview) - bring your own agents, command your friends', bots & humans in the loop. - K2.6 is now live on kimi.com in chat mode and agent mode. For production-grade coding, pair K2.6 with Kimi Code: kimi.com/code - 🔗 API: platform.moonshot.ai 🔗 Tech blog: kimi.com/blog/kimi-k2-6 🔗 Weights & code: huggingface.co/moonshotai/Kim…
English
