Investing in Yupp
Modern AI systems are shaped not only by compute and algorithms but by human feedback. Companies use post-training techniques such as Reinforcement Learning from Human Feedback (RLHF) and Direct Preference Optimisation (DPO) to improve their models. These techniques can reduce bias and enable higher quality, more coherent responses to prompts — crucial for accelerating progress in AI. Model evaluation is similarly critical, but a model can only be made better after first deciding what “better” means.
That’s where challenges arise: Companies don’t like to share — they keep their data and training processes secret. As a result, AI model evaluation is constrained by what can be learned from closed systems or static benchmarks that are rarely informed by real-world use. This limits the improvements that can be made to AI models. Users are also left in the dark, with little insight into how their feedback shapes models or whether it’s used at all. Some leaderboards and crowdsourcing sites attempt to shed light here, but they generally don’t enable users to audit their contributions or see any direct benefit from participating. Platforms that claim to be fair and transparent often rely more on good faith than enforceable standards.
We believe crypto can bring transparency and ownership to this murky area of AI. Blockchains can make it easier for people to receive rewards for their contributions. They can also provide AI builders with assurances about the quality and provenance of the feedback data and evaluations they’re incorporating into their models. So users get incentives, builders get trustworthy data, and everyone can audit either side of the open market. This is why we led a $33 million seed round in Yupp, a consumer product that allows anyone to discover and compare the latest AI models for free.
Yupp crowdsources model evaluation: users enter prompts, see multiple AI-generated responses side-by-side, and then pick the best ones. Their choices create digitally signed “packets” of preference data that are useful for AI post-training and evaluation. In addition to users getting access to the latest models for free, they receive rewards based on the feedback that they provide.
Yupp’s design turns human judgment into a renewable economic resource. Data “expires” as newer interactions replace it, creating a natural flywheel: more usage yields fresher evaluations; fresher evaluations yield 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 credibly neutral marketplace. No one can hide the scoreboard, and no one can manipulate the rewards or results.
The founders bring deep experience in both AI and crypto. They built consumer-scale machine learning products together in the early days of Twitter. Pankaj Gupta ran global consumer engineering for Google Pay and Coinbase. Gilad Mishne was a machine learning lead at GoogleX. The early team already counts senior engineers from Google, Coinbase, and top research labs.
AI needs strong, reliable evaluation based on large-scale human input. Crypto is the trust machine that can help deliver it. By enabling people worldwide to contribute model-improving feedback, Yupp aims to become the default evaluation layer for the future of AI. We’re proud to back Yupp and look forward to helping them build the onchain feedback loop that ensures the rewards of AI innovation are shared by everyone who helps create it.
***
The views expressed here are those of the individual AH Capital Management, L.L.C. (“a16z”) personnel quoted and are not the views of a16z or its affiliates. Certain information contained in here has been obtained from third-party sources, including from portfolio companies of funds managed by a16z. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the current or enduring accuracy of the information or its appropriateness for a given situation. In addition, this content may include third-party advertisements; a16z has not reviewed such advertisements and does not endorse any advertising content contained therein.
This content is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. References to any securities or digital assets are for illustrative purposes only, and do not constitute an investment recommendation or offer to provide investment advisory services. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund and should be read in their entirety.) Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by a16z, and there can be no assurance that the investments will be profitable or that other investments made in the future will have similar characteristics or results. A list of investments made by funds managed by Andreessen Horowitz (excluding investments for which the issuer has not provided permission for a16z to disclose publicly as well as unannounced investments in publicly traded digital assets) is available at https://a16z.com/investments/.
The content speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects, and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see https://a16z.com/disclosures for additional important information.