I’m excited to announce that we’ve led a $33 million seed round for @yupp_ai. This is a consumer product that allows anyone to explore and compare the latest AI models for free. AI requires strong and trustworthy human data. Cryptocurrency is here for this. Modern AI systems are influenced not only by computation and algorithms, but also by human feedback. Many companies use post-training techniques such as reinforcement learning with human feedback (RLHF) and direct preference optimization (DPO) to improve their models. These techniques can reduce bias and enable higher quality, more consistent prompt responses - which is critical to accelerating progress in AI. Model evaluation is equally important, but models can only get better if you first understand what “better” means. The challenge is that companies are reluctant to share - they keep their data and training processes secret. As a result, model improvements are limited to what can be learned from closed systems or static benchmarks, which rarely reference real-world applications. These limitations make AI models difficult to evaluate. Users are also left in the dark, with little insight into how their feedback affects the model, or even whether it is actually being applied. Some leaderboards and crowdsourcing sites attempt to reveal these truths, but they often don’t allow users to audit their contributions or see any direct benefits from participating. Platforms that claim to be fair and transparent often rely more on good faith than enforceable standards. We believe cryptocurrencies can bring transparency and autonomy to the murky world of AI. Blockchains can make it easier for people to be rewarded for their contributions. They can also provide AI builders with assurances about the quality and provenance of the feedback data and evaluations they incorporate into their models. This way, users are incentivized, builders get trusted data, and everyone can audit either side of the open market. Yupp Crowdsourced Model Evaluation: Users input prompts, view multiple AI-generated responses side-by-side, and pick the best. Their choices create digitally signed “packets” of preference data that can be used for later training and evaluation of the AI. In addition to free access to the latest models, users are rewarded for the feedback they provide. Yupp’s design turns human judgment into a renewable economic resource. Data “expires” as new interactions replace it, creating a natural flywheel: more usage generates newer evaluations; newer evaluations generate better models; better models attract more usage. All participants — from users to AI model builders — can participate and see that the same transparent rules apply to everyone, ensuring a trusted neutral market. No one can hide the scoreboard, and no one can manipulate rewards or outcomes. The founders have extensive experience in both AI and crypto. Together they built consumer-scale machine learning products in the early days of Twitter. @pankaj led global consumer engineering at Google Pay and @Coinbase. @gilad was previously head of machine learning at GoogleX. The early team already has senior engineers from Google, Coinbase, and top research labs. AI requires strong, reliable evaluations based on human input at scale. Crypto is the trust machine that enables this. It enables people around the world to contribute feedback that improves models.