Genie 3: A New Frontier for World Models08–5–2025This announcement of the new state-of-the-art in world generation came with entirely new pradigms in which to push into and discover the limits of what this model could synthesize and simulate. The shift from modeling worlds in 3D to the streaming of pixels in realtime is not neccessarily an intuitive one, and one that required looking at this technology from a new and fresh perspective.
No longer are you simply consuming / reacting to generated media, you’re interacting with it in real time, your actions and intentions are reflected on the world you’ve built around yourself. 


See Blog Post
Watch Full Launch Video


Gemini Robotics03–12–2025For the better part of 2 years, me and small team of creatives have been operating on the cutting edge of generative AI, taking shape in ways such as smart glasses, universal assistants, realtime chatbots, media generation stations, etc. For the first time, we’ve expanded into the physical realm through a medley of robotic form factors.
We designed an arrangement of Human-Robot Interactions to exemplify how Gemini 2.0 can serve a “brain” for any number of robotic functions. This embodied intelligence was showcased through a series of launch videos and blog posts, of which we handled the creative direction and production. 

See Blog Post
Watch Full Launch Video



Project Astra I/O 202505–21–2025Exactly one year after our first announcement regarding Project Astra, our test bed for research developments surrounding the future of universal assistants, we were tasked with demoing the plethora of advancements our researchers have accomplished in just one short year. 
Action Intelligence, our vision for agents with the ability to take action and complete multi-step tasks on a user’s behalf, was the main point of contact through which we explored these new interaction paradigms. Through Interface Control, Call Assistance, Intelligent Personilization, we devised a narrative throughline that would showcase these feats of strength. 

See Blog Post
Watch Full Launch Video





Exploring London with Project Astra12–07–2024Building upon the foundational work established in our initial Project Astra demonstrations, this London-based showcase represented a pivotal moment in our exploration of real-world AI assistant capabilities. 

Working directly alongside the Astra research team at DeepMind's London headquarters, the creative process involved divising the sequences through experiential documentation of stress tests against the experimental model, ensuring that each demonstrated capability felt natural and contextually relevant to real-world use cases, while also being 100% authentic to the model’s capabilities at the time.

See Blog Post
Watch Full Launch Video


Project Astra I/O 202405–13–2024 This project was my second time working with DeepMind in a creative and collaborative capacity. In a multi-team, multi-continent sprint, we were tasked with demonstrating Google’s view on realtime, multi-modal assistants that can see and reason about the world around you.

This experimental research project was multi-pronged both in its creation, but also in the tech that went into it. We were most excited by exploring this universal agent in more than the familair form factor of smartphones, in this case, hands-free AR glasses (Slide 2). 

Utility Based Knowledge Assistant
Premiered at Google I/O by Demis Hassabis 
Designed Demos for Smartphone, Glasses, and Booths
 

Project Gemini12–06–2023
This project served as the platform to launch the first public showcase of Google’s new “Any to Any” generative AI model. There was a mixture of speculative design and high-level creative direction that went into this collective effort aimed at ushering in the aptly named “Gemini Era.” 

The centerpiece of the launch consisted of a 5 minute “table-top” style video showcasing various feats of strength found within the model. This video was concepted, shot, and directed by myself and handful of amazing thinkers at the Creative Lab. It spoke to our vision of what we wanted these ever-nearing interactions with the tech of the future to not only look and sound like, but to feel like as well. 

“Google admits AI viral video was edited to look better” - BBC
“Google's best Gemini demo was faked” - TechCrunch
“Google has a new AI . . at least one demo wasn’t real“ - Verge

©2025Contact for inquiries  ->  sam@lwtnlabs.com