The Delphi Podcast cover image

The Delphi Podcast

Nick Emmons: Crypto x AI Will Change The Fabric of Society and DeFi Powered AI Is the First Step

Jan 15, 2024
53:14
Snipd AI
Nick Emmons, CEO of Upshot, discusses decentralizing AI models, architecting coordination mechanisms, reimagining the crypto-AI stack, protecting model IP while enabling collective intelligence, whether crypto startups can compete with Big Tech, and predictions on AGI timelines and decentralization's impact on access.
Read more

Podcast summary created with Snipd AI

Quick takeaways

  • Decentralized AI networks can coordinate diverse models for more accurate and powerful decentralized applications.
  • The integration of AI into decentralized finance holds promise for revolutionizing financial operations and infrastructure.

Deep dives

The Evolution of Upshot: From Crowd-Sourced Valuation to AI-Powered Financial Infrastructure

Upshot initially started as a crowd-sourced appraisal platform for infrequently traded assets like NFTs. They used a mechanism design called peer prediction to incentivize accurate valuations from humans. However, they found scalability and accuracy limitations with this approach. As a result, they shifted their focus to building AI infrastructure for pricing long-tail assets and creating efficient financial infrastructure. They developed proprietary AI models that leverage various data sources, including open market data, metadata, and social sentiment data, to generate real-time price estimates for assets. Upshot is now building a decentralized AI network that allows anyone with AI models to contribute and coordinate along shared economic objectives. They are aiming to create a collective intelligence where the network's output is better than any individual model. Users can access Upshot's price predictions through various tools, lending platforms, and perpetual platforms. They are also exploring tokenized asset appraisal and AI-powered loans. Upshot's model is designed to protect IP and ensure verifiability of inference. They are leveraging modulus labs' technology to provide ZK predictors that verify the output of models without revealing proprietary information. This approach allows for a range of specialized models to be utilized in different contexts, enabling more accurate and expressive AI applications in the decentralized space.

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode

Save any
moment

Hear something you like? Tap your headphones to save it with AI-generated key takeaways

Share
& Export

Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more

AI-powered
podcast player

Listen to all your favourite podcasts with AI-powered features

Discover
highlights

Listen to the best highlights from the podcasts you love and dive into the full episode