A-Merk’s LoRa Artifacts & Beyond

A-Merk has been developing their craft for over a decade, initially producing music before adopting visual arts, driven by a lack of collaborators who could produce the music videos they envisioned. While audio work remains a part of their creative repertoire, they have since shifted their focus predominantly towards visual media.

He has embarked on an artistic journey utilizing various free tools accessible through phone, coupled with elementary blending and motion techniques in After Effects. Their journey has been nothing short of remarkable, eventually leading them to specialize in Mandelbulb3D videos. These videos, characterized by intricate 3D fly-throughs of otherworldly landscapes, are rooted in complex mathematical algorithms. This experience laid the groundwork for their eventual foray into AI art, a field in which they liken themselves to a photographer or explorer, navigating the latent spaces of AI image-generating models.

LoRa & A-Merk’s Visual Manipulation techniques

LoRa, standing for Low-Rank Adaptation, is a mathematical technique used to reduce the number of parameters in a machine learning model during training. Developed by researchers at Microsoft, LoRa allows for the efficient adaptation of pre-trained models to new data or styles using just a few examples. The primary advantages of this technique include significantly faster training times and smaller model output sizes compared to other methods. LoRa’s application extends to various domains, including the field of visual arts, where it is revolutionizing image generation and content creation.

Animation and Interactive Media

LoRa’s impact extends to animation and interactive media, where its quick adaptation capabilities facilitate creative freedom and experimentation. Designers in the gaming industry can utilize LoRa to swiftly generate diverse character designs and environmental assets, maintaining a consistent visual style across different elements. In virtual and augmented reality applications, LoRa ensures that immersive experiences remain seamless, even when the visual content adapts in real-time based on user interactions. This capability is crucial in maintaining user engagement and delivering a high-quality immersive experience in interactive media.

The artistic mission has remained consistent throughout their career: to encapsulate and share the enigmatic beauty of The Psychedelic Experience. They are captivated by the alien and wondrous nature of the mind under the influence of sacred substances, seeking to recreate this experience and the accompanying feelings of inner peace and universal love through their art. AI art has become a medium through which they search for the most intriguing and mesmerizing visual experiences.

More recently, they have delved into training Stable Diffusion with their past artworks, pushing the boundaries of AI art and creating pieces that are uniquely their own. Their engagement with Web3 technology began in 2017 and has only intensified since. Their first NFT project, “Cantoasis,” was launched on Canto but faced various challenges and did not reach the heights they had hoped for. Nonetheless, they learned invaluable lessons from this experience, strengthened by the support of the Web3 community and particularly some major assistance from Degenesys.

Presently, their focus is squarely on Friend Tech. They have committed themselves to a rigorous challenge: delivering daily NFT airdrops of unique art to their FT holders for 42 days straight and counting. The project, titled $FRIENDS ARE MONEY$, is a playful commentary on the cynicism surrounding social fi platforms, but at its heart, it’s about community and connection ($FaM$). The artwork for the airdrops revolves around monetary themes, providing an additional layer of creative challenge. The use of custom styles trained into Stable Diffusion through the LoRa methodology enables them to produce art that stands out from typical AI-generated pieces.

They have demonstrated the impact of their LoRa training with examples, showcasing how their personalized styles, derived from previous artworks, can transform a basic render from Stable Diffusion XL into something uniquely theirs. This method ensures a continuous supply of fresh and exciting art, reflecting their commitment to their community and the potential of AI in artistic creation.