Mauvis Ledford and his AI-powered dream coat

At 2am in Singapore, while most of the city sleeps, Mauvis Ledford scrolls through his Discord community, watching what people create. There’s “Doesn’t Care” — a community moderator turned artist who produces surreal, Dalí-esque pieces that stop Ledford in his tracks. Tonight, he may come across new visual territories.

“That’s what keeps me going: Watching someone with no formal art background suddenly discover they can manifest visions that have been locked in their head,” says the chief executive officer of Sogni. “Every time I think I understand Sogni’s potential, someone like Doesn’t Care pushes it somewhere unexpected.”

This drive to democratise creativity likely stems from Ledford’s unique background, which includes a bachelor’s degree in Art combined with over 20 years of experience in technology.          

The spark that changed everything

The genesis of Sogni originated from his former colleague at CoinMarketCap and now Sogni’s Chief Product Officer, Alejandro Ramos. “He saw how everyone was becoming dependent on expensive artificial intelligence (AI) subscriptions and created Sogni V1, a native Mac app that made 100-plus open-source models actually usable with a clean interface. No command lines, no technical barriers,” he recalls.

When Ramos showed it to Ledford, he was blown away by the accessibility but frustrated by the speed. Even a high-end MacBook could only generate one image per minute — the same as paying for Midjourney or DALL-E. “That’s when I had the insight: ‘What if we could wire my Macs to your Macs, and render together?’ And so, Sogni Supernet, our Graphics Processing Unit (GPU) network, was born.”

This simple idea proved to be more effective than they had imagined. One and a half years later, in April 2024, with a team of 14, they created a full ecosystem: the Sogni software suite, a dual-tier GPU Supernet with both high-speed Nvidia and eco-friendly, cost-optimised Mac networks, and a software development kit (SDK) so anyone can build generative AI apps on their infrastructure.

Play

To scale up, Sogni secured US$2 million ($2.57 million) in pre-seed funding in January 2025. A few months later, it added another US$1.5 million in a seed round led by the Tezos Foundation, which is long associated with supporting artists and creatives.

That brought the total funding to US$3.5 million, just as the company launched its live, decentralised network, moving from the test phase into full operation. 

Breaking down the walls

“The AI renaissance is happening right now, but a handful of tech giants is gatekeeping it,” Ledford asserts. “You pay them US$20 to US$30 monthly to access their models, funding their research and development, yet the moment you stop paying, you’re locked out completely. Meanwhile, open-source AI models are incredibly powerful. They are only two to six months behind the proprietary ones, but they sit unreachable for most creators.”

Sogni’s solution addresses all three obstacles simultaneously. They wrap powerful open-source models in polished, intuitive apps that anyone can use. At the same time, it builds a decentralised GPU network where everyday people with gaming PCs or Macs can share their idle computing power and earn income. 

He further simplifies the benefits, explaining: “This creates a virtuous cycle (where) artists get access to AI rendering at half the cost of centralised platforms, GPU owners earn up to US$350 monthly for their idle hardware, and the entire ecosystem remains open and unrestricted.”

His determination to democratise creative technology has historical precedent. Andy Warhol, “without question”, the artist that Ledford would work with if he had the chance. Ledford calls Warhol’s Factory an early collective of artists, musicians and misfits that blurred authorship and showed how technology could democratise creation.

What especially resonates with Ledford is Warhol’s openness: He embraced mechanical processes like silk screening instead of concealing them. “He didn’t hide his methods; he celebrated them. He’d immediately understand that AI isn’t ‘cheating’ any more than his photo-silkscreens were.”

That outlook comes alive in Sogni’s own community, where curiosity often pushes the tools in directions no one anticipated.

When accidents ignite innovation

One user discovered they could use ControlNets, an advanced feature in Sogni Studio, to photograph crumpled pieces of paper and transform them into organic architectural designs. The random folds and shadows became the blueprint for flowing, biomimetic buildings that no one would have conceptualised through conventional methods.

“It completely shifted my perspective on creative discovery,” he shares, adding, “I sometimes worry that Sogni Studio is too complex for the average user, with ControlNets, pose detection, depth maps, and edge detection. But then, someone curious enough to explore creates something that makes you rethink what’s possible.”

Such experiments inspire Ledford, but they also expose the pressure on established roles. “The reality is that clients increasingly care about outcomes of quality, speed and price — not the process. A design agency that’s thrived for decades without AI will struggle against lean teams delivering comparable work faster and cheaper.”

“Solo artists using AI can now handle workloads that once required five people. They’re also freed from repetitive tasks to focus on what matters: consciousness, storytelling, emotional resonance,” he elaborates. “I recognise this change is painful for many. Mid-level creative roles are genuinely at risk. But there’s also an opportunity. (After all) Every technological shift creates new possibilities.”

The balanced view reflects Ledford’s commitment to building technology that serves people rather than displacing them. Rather than dismiss these concerns, he channels them into practical action. Sogni’s most ambitious, some might even say audacious, undertaking yet has him working on transforming how the next generation of creators learns AI.

The university gamble

Sogni has begun hosting workshops and hackathons at Nanyang Technological University, and they’re expanding to universities across the US and Europe.

The aspiration goes beyond workshops. “We want to provide entire creative departments with free access to our decentralised physical infrastructure network (DePin) and open-source models. No expensive enterprise contracts, no complex GPU procurement. Universities can even contribute their existing computer labs to the network during off-hours, earning credits that offset or eliminate costs,” he asserts.

“The audacious part? We aim to establish a global network of university nodes where students can utilise AI for more than just academic purposes. They’re training custom models on their cultural heritage, building apps on our SDK for their local communities, and earning real income while learning.”

By 2030, Ledford would like to see a headline like: “Open-source AI networks process more creative work than Big Tech Combined; millions earn a living through decentralised GPU sharing.”

But more than market dominance, he wants proof that critical infrastructure doesn’t have to be monopolised. He would like “a community-owned network can outcompete corporate giants not by being cheaper, but by being aligned with users’ interests”.

Ledford affirms: “The lesson is to build for reality, not ideology. We shipped, learned, and iterated while others were still writing whitepapers. Many Web3 projects fail because they prioritise decentralisation over solving real problems. We started with a simple pain point — artists locked out of tools — and used decentralisation as the solution, not the goal.”

The North Star continues to prove that our DePIN actually works, demonstrating that “a rowdy group of Artists, Model Creators, and GPU Workers can contribute to a decentralised infrastructure that outperforms centralised systems on every metric that matters: speed, cost, privacy, and accessibility”.

“Power to the people isn’t just a slogan. It’s an engineering challenge we’re solving one render at a time.”