Artificial intelligence is dominating tech headlines lately thanks to OpenAI’s advancing GPT models and competitors, and such models can generate text and media with alarming speed and accuracy. They may revolutionize many industries and experiences—and one Web3 metaverse app has already found a way to tap into the technology.

Oncyber, a 3D world-building platform that supports a wide array of NFTs and is used by Web3 creators and communities, revealed this week that it has developed an AI-powered tool called Magic Composer that lets users customize their environments via text commands.

It’s like using ChatGPT—and it runs on OpenAI’s GPT-3.5 model (not the brand new GPT-4)—but instead of returning well-curated information from the web, a poem or book report, or even functional blockchain code, Oncyber uses text prompts to implement real-time tweaks to its worlds.

AD

With a line of text dropped into the Magic Composer, Oncyber can automatically change the look and color of the sky, drop in an artist’s NFT works from your connected crypto wallet, customize the look and size of picture frames in the world, and more. And this is just the first version of the tool, Oncyber founder and CEO Rayan Boutaleb told Decrypt.

“In the next iteration, what we want is for people to port any hallucination or dream they have into a 3D canvas, and just directly see the result of what they might imagine and change things up,” he said. “It’s a first step towards a very big end result, which we are not that far off from.”

A screenshot of Oncyber's AI tool in its revamped 3D studio. Image: Oncyber

Embracing AI tooling wasn’t about chasing the latest buzzy tech trend, but rather finding another way to help users unlock their creativity on the platform. There’s a disconnect between the power of Oncyber’s game-like engine and the visual user interface, Boutaleb explained.

The Magic Composer provides a way for users to get what they want with relative ease. And as Oncyber layers in additional world-building features in the coming months and provides ways for creators to develop richer environments, the AI tool will hopefully cut through some of the complexity and ensure accessible ways to personalize online 3D spaces.

AD

Oncyber plans to roll out the AI tool to a select group of testers starting Monday, March 20, but Boutaleb said it won’t be long before it reaches public users—a few days later, perhaps.

More robust AI functionality is on the horizon, too. Boutaleb said that in time, he hopes to add more complex generative features, such as the ability to generate bespoke or modified 3D architecture simply by typing out a request. Over time, Oncyber’s goal is to have “less and less boundaries/limits” for users, he added.

Launched in 2021, Oncyber has called itself a metaverse platform, and shares commonalities with Web3 metaverse gaming platforms like Decentraland and The Sandbox that also let users customize their own online spaces.

But unlike those platforms, Oncyber didn’t sell NFT land plots that are required to build within those worlds, rejecting the premise of “FOMO” and speculation fueled by land scarcity. Oncyber does sell NFT-based templates, however, working with architects and designers to develop prefab spaces that can be further personalized.

There are free spaces and templates too, however, and you don’t need an NFT or a crypto wallet to create a space. Oncyber also lets users bring in owned NFTs across Ethereum, Solana, Polygon, and other networks. Some NFT collectors use Oncyber to create their own virtual art galleries, while others use the platform to host immersive online community events.

Because Oncyber—which raised $6.7 million last year—wasn’t tethered to the NFT land hype cycle that surged and then plummeted, Boutaleb said that his team has focused on steady iteration to the platform. Other recent additions include enhanced VR support plus the ability to port in external video and screen-sharing streams.

Daily Debrief Newsletter

Start every day with the top news stories right now, plus original features, a podcast, videos and more.