The "Write Like lsusr" Challenge: A Stylistic Turing Test
Coverage of lessw-blog
In a recent community announcement, lessw-blog presents the "$500 Write like lsusr competition," a challenge inviting participants to emulate the specific voice and rhetorical style of a known contributor.
In a recent post, lessw-blog outlines the rules for a unique writing contest: the "$500 Write like lsusr competition." The premise is straightforward yet technically demanding. Participants are tasked with authoring a blog post that mimics the style of the user 'lsusr' so effectively that it becomes indistinguishable from the target's authentic work. This competition is not merely a creative writing exercise but serves as an informal evaluation of stylistic mimicry and identity verification within online communities.
The competition operates on a tight schedule, with entries due by December 28. To ensure the integrity of the "blind" judging process, entrants are required to create new Less Wrong accounts to post their submissions. The resolution mechanism relies on prediction markets, which will close on December 31, effectively crowdsourcing the verification process. This approach gamifies the detection of authorship, asking the community to bet on which content is genuine and which is an imitation.
Perhaps the most significant aspect of this competition is its stance on artificial intelligence. The organizers have drawn a clear line regarding the use of generative tools. While "AI slop"-defined here as low-effort, purely AI-generated text-is explicitly disqualified, the use of AI for research, spellchecking, and refining ideas is permitted. This distinction reflects a growing maturity in how technical communities view LLMs: as valid tools for augmentation but poor substitutes for the distinct stylistic nuances of human authorship.
This event highlights the evolving challenges in distinguishing human-authored content from machine-generated or machine-assisted text. By incentivizing high-quality mimicry, the competition acts as a localized Turing test, probing the boundaries of style transfer and the current capabilities of both human writers and their digital tools.
For those interested in the mechanics of prediction markets or the nuances of stylistic imitation, the full announcement provides detailed entry requirements and timelines.
Key Takeaways
- The competition offers a $500 prize for the best imitation of the author 'lsusr' on the Less Wrong platform.
- Entries must be submitted via new accounts by December 28, with prediction markets resolving the results by December 31.
- The rules establish a nuanced stance on AI: tools for research and editing are allowed, but purely AI-generated content is disqualified.
- The event serves as a community-driven experiment in authorship verification and style transfer.