machina
ludens.



Section: AI-networks

Date First Posted 
December 2021



This page serves as my crypto and web 3.0 portfolio. The following projects are a brief collection of bounties and graphic design collaborations that I’ve done throughout the year. 
args = argparse.Namespace(
prompts=['Hungarian Society Of Cardiology'],
image_prompts=[],
noise_prompt_seeds=[],
noise_prompt_weights=[],
initial_image=[''],
tv_weight=0.,
clip_model='ViT-B/32',
vqgan_config='vqgan_imagenet_f16_1024.yaml',
vqgan_checkpoint='vqgan_imagenet_f16_1024.ckpt',
step_size=0.05,
weight_decay=0.,
cutn=64,
cut_pow=1.,
display_freq=50,
seed=0,
)

DeepDao

>>
january,

2022

I participated in my first web3.0 bounty during january, where we had to visualize DeepDao’s personas via algorithmic ways. See our official gitbook documentation here.



“Calling DeepDAO's governance nerds and NFT artists! We love DAOs, proposals, token holders, member coalitions, voting patterns. We are keen to know who are the top DAOs, the top people in governance, who are the active and inactive people and organizations, who are the rising DAO stars and the falling angels, the healthy and the epic. Inspired generative artists, how would you transform DAO data into breathtaking art?
As one of us says: "we love the punks and apes but we're curious to see new identities to separate and define the DAO world and culture".Technical Requirements None. We're looking for your inspiring, cool, innovative art. Once we select it we'll put our coding team to work with you on making the right NFT choices. Judging Criteria The winner will be selected by DAO governors from DeepDAO, and members of our community.”


Cyberknight

>> december,
2021


The CyberKnight Poster Series was an ongoing collaboration with my other collective, temporary_nites where I developed fifteen different versions of the knight for our medieval-themed music event. All pieces of the edition were constructed with mere code, where given the right prompst to the machine learning algo, the right images emerge - but always with an unpredicted twist. The colab notebooks used for the project were based on the vqgan+clip method which then I edited further in Photoshop.