How we built the Waifu Vending Machine
In this technical retrospective (2019), @gwern shares practical architectural lessons learned from running, scaling, or migrating production software under lean constraints.
Core Principles for Independent Builders:
Defaulting to simpler architectures (such as monolithic frameworks, SQLite, or managed VPS) prevents early engineering bankruptcy.
Evaluating true compute and egress costs reveals that high cloud bills can often be slashed by 80%+ through judicious re-platforming.
Code you write at 2 AM as a solo engineer must be easy to debug six months later without complex microservice dependencies.
Community Debate Themes:
HN engineers debated the tradeoffs between cloud convenience and raw server economics, database indexing strategies, and resisting premature architectural complexity.
CATEGORY Frugal Engineering →
HISTORICAL ERA Cloud & Mobile Era (2014–2019)
COMMUNITY SCORE ★ 257 points
TOTAL DISCUSSIONS 154 comments
The dataset this is built from (https://www.gwern.net/Danbooru2018) is, simply put, copyright-infringing on a gross scale. The vast majority of the images uploaded to 'boorus' completely lack a compatible license or the artist's express consent. The redistribution via torrent of 2.5 TB of some 3 million images only compounds this problem. None of this is ameliorated by the $20 'generosity' of the dataset creator.
As a result, every single artist whose work was included in that dataset has a clear, meaningful claim that each and every 'waifu' sold ($20, if customized, or $5 if random) by Sizigi Studios is an infringing derivative work. Coupled with at least one of the project authors' ready admissions -- in this very comment section -- of scraping image sites himself, I would say that this team is playing with fire. Even in the case that an algorithm's output is somehow found to be 'creative' rather than mechanistic, AND this specific application is found to be in all cases substantially transformative, there's STILL the original massive 2.5 TB of copyright infringement up front to deal with.
All an enterprising lawyer would need to begin is to search the BigQuery metadata for the 'artist' and 'copyright' tags on these images. Note of course that the 'copyright' tag is widely misused on boorus and similar image repositories to refer to the inspiring franchise; 'trademark' would be much more accurate descriptor.
EDIT: I do not mean to suggest that litigation from the use of the dataset in this ML (as opposed to the original, clearly infringing, download & redistribution) would in any way be an easy, one-sided case --- only that this scenario would represent nearly the worst possible test case imaginable for determining the future legality of ML, short of directly antagonizing the RIAA or MPAA.
I am trying out the app at https://waifulabs.com/ and the art style is kind of one-note. Most of the expressions are the same and the face shape skews towards loli.
I am more into "disgusted anime girl that looks at you like you're trash" type and I couldn't find a waifu (even with their refinement steps).
Really impressed that this is even possible though!
The tech is interesting, but the concept and naming is pretty creepy. Waifu, from "wife", is anime slang for female cartoon characters that people get romantically attracted to.
https://www.dictionary.com/e/fictional-characters/waifu/
I'm not sure how you're shipping your posters, but here's a tip:
When shipping posters, use triangle tubes, not circular tubes, it saves you money.
[0] https://www.linkedin.com/feed/update/urn:li:activity:6549769659112054784/
I would love to replicate this. It looks like the dataset is open source. https://www.gwern.net/Danbooru2018
I don't have a sense for hardware requirements though. Does anyone have a good idea of how much time and money it would take to train such a model?