🚀 Excited to announce WorldForge — a training-free approach for unlocking emergent 3D/4D world generation in video diffusion models! 🌍🎥 👉 Supports 10+ 3D/4D tasks 👉 Project page: https://lnkd.in/gUPjs2wi 👉 Demos & code coming soon!

This weekend, Google AI for Developers and Kaggle launched a hackathon to celebrate the release of the new image generation model, NanoBanana, also known as Gemini Flash Image Preview. I took this opportunity to revisit Google App Builder, a fantastic low-code application from Google that enables the creation of web applications. I submitted a project called Vibe Mooding 😉, designed to assist with interior decoration. The goal was to leverage Google's range of LLMs to provide interior decoration insights, combined with the image mixing and stability capabilities of the NanoBanana model, ensuring coherent and high-quality image outputs. I'll make this application available on the apps builder gallery once the hackathon concludes. The NanoBanana model has proven to be highly efficient and fast. The development environments, including Google AI Studio and Google App Builder, have undergone significant improvements, making it possible to develop the entire project by defining specifications in the morning while gardening, and building it from my phone during family shopping in the afternoon. At the end, the most challenging part was creating the VERY concise 2-minutes explanatory video XD.

CRO isn’t just about changing a button’s color. It’s the mix of: - Fast page speed. - Smooth navigation - Clear menus. - Strong visuals. - The right copy. - The right product. Let that sink in. Post idea came when i was in btw of demo day of KPH hackathon.

After a 12-hour build sprint right before the hackathon deadline, I added two new features to Synaptik Core. This walkthrough shows how it’s different from today’s RAG-based memory solutions. Focusing on durability, replayability, and verifiable governance rather than just retrieval. 🎥 Walkthrough below:

I built Alur One so you can use nodes to experiment creatively — in this demo, Hakim (our closed alpha user) use his artwork, create multi-view shots, and generate a 3D model.

👉 Start here: write a prompt ➡️ image generated ➡️ video generated ➡️ integrate this video into an interface. 👉 How it went? I've created a full prototype, your habits tracker if you will: https://sola.figma.site/ and submitted for the Contra/Figma Hackathon. #concept #interface #avantgarde #dothingsdifferently #designwithai #aixdesignxart #figmahackathon #contra #promptleddesign #promptleddesigner

Excited to Share My Latest Project! Over the past month, I’ve been working on something that truly tested my skills and passion for game development. As a student in the Gaming Tech branch, I’ve always been fascinated by the scope of creating games — and this project gave me a firsthand experience of how challenging and rewarding it can be. RoadWise3D – a simulation game designed with the purpose of teaching traffic rules and road safety in a gamified, risk-free environment. As the Game Programmer, my role was to develop a working prototype for our Project Exhibition 1. Here’s what I was able to deliver so far: - A physics-based driving model with realistic suspension and tire physics - An AI traffic system that uses way-points to create believable traffic patterns - Lane logic , Object detection & collision avoidance for AI vehicles The project is still in progress, and while there isn’t a playable version ready yet, I’m sharing a short clip showcasing the driving and AI traffic in action. This journey has been a great learning experience, and I’m looking forward to improving and building on it further.

🚦 If your session needs 6 bits of kit, 9 demos and a rocket scientist to explain it… It’s too complicated. Simple = scalable Simple = coachable Simple = effective #sgpt #grouptraining #groupcoaching #smallgrouptraining #groupfitness #grouppt #coachingcoaches #smallgrouppersonaltraining #ptmentor #personaltrainer

We're happy to see so many members of the Furhat community at ICSR25 - International Conference on Social Robotics this year. 🎉 One of the highlights today is the paper presentation on "HRI-based Interview Training using the FurHat Robot" by Adam Jones and Professor Roger Moore from The University of Sheffield. The paper compared a virtual Furhat and physical Furhat in an interview training setting, with the results indicating that participants found the physical Furhat more engaging, helpful and satisfying to talk to. Their work highlights the potential importance of using physical embodied systems in professional skill development instead of virtual systems.👏

Our autocomplete feature is now out of experimental mode! To enable it, head over to Settings → Autocomplete. Check out the video for a short demo.