Summary : Heurist AMA (by. Exabits. 8th_may_ 2024)
English is not my native language, so there may be missing or inadequate parts. With your understanding, I’ll summarize the main points.
<delving into “The Role of Web3 in Ensuring Transparency and Addressing Bias in AI Data”>
1. About Heurist Team:
-Heurist team comes from companies like Amazon, Oracle, Nvidia, Anthropic
-Started the project in December 2022, currently running an incentivized testnet
-Plans node sale and token launch (TGE) in Q3 2023
2. Heurist Token Launch & Mining Incentive Structure:
-Heu token launch planned for Q3
-Considering node sale format for token issuance and mining rewards
-Currently building initial community via airdrop of NFT governance tokens
-Establishing inference verification and anti-cheating mechanisms
-Expanding partnerships to increase network demand
-Building new user-friendly branding/website
3. Heurist Node Sale Plan:
-Plan to sell CPU node licenses in addition to GPU nodes
-CPU nodes will contribute to verifying inference results and network security
-CPU nodes will have relatively lower purchase cost compared to GPU nodes
-Estimated to be one-time license purchase model (no mention of recurring costs)
-Resolves GPU shortage issues and allows more users to join the network
(Note: In Discord, JW said “a lot of node sales these days are value extraction. we’re building a system that makes such nodes really useful. If we cannot deliver a sound prototype, we won’t launch the node sale. I’m 100% sure about a node sale”)
4. Regarding Heurist NFT:
-Issued 500 NFTs as initial governance tokens and airdropped to community
-Plans to launch full-fledged NFT-based governance tokens in the future
5. Heurist’s Differentiators from Competitors:
-Based on open-source models instead of centralized models
-Cost-efficient — 60–95% lower cost compared to AWS etc.
-Incentive structure to share revenue with model developers
6. How is Heurist different from ChatGPT?
What innovation beyond ChatGPT does Heurist fuel?
-Unlike ChatGPT, Heurist provides various AI applications like image generation, translation
-Instead of a single model, runs open-source AI models on distributed GPU network
-Plans to continually build more AI products and an ecosystem
7. Specific Technical Challenges for the Project:
-Complexity of coordinating large distributed GPU network
-Building a stable and scalable system
-Developing a trustworthy proof of inference protocol
8. Partnership between Exhibits and Heurist:
-Exhibits provides enterprise-grade GPU computing power to Heurist network
-Exhibits’ GPU resources are being provided to Heurist GPU miners
9. What does the next 5 years look like for Web3 and AI?:
-Web3 AI going mainstream, unicorn companies emerging, community-owned firms/tokens
-Startups gaining competitiveness over cloud providers on Web3 AI basis
-Regulatory environment still evolving
10. How big is decentralized AI market?:
-Even capturing 1–5% of total AI market would exceed current crypto market cap
11. How does Web3 relate to data transparency in the context of AI?:
-Decentralized nature allows crowdsourcing data collection for diversity
-Ownership concept ensures individual data ownership and compensation for utilization
-Incentive mechanisms reward data contribution with cryptocurrencies
12. How can individuals/communities contribute to Web3 AI transparency and fairness?:
-Voice data ownership claims, leverage social media actively
-Contribute by participating in Web3 projects, decentralizing personal data
13. Strategies for building Web3 AI ecosystem:
-Technically: Distributed GPU/CPU networks, proof of inference, establishing trust
-Economically: Cryptocurrency-based incentive structures and revenue sharing models
-Governance: NFT-based community decision-making frameworks
-Mainstream adoption: Education, partnerships, marketing efforts
-Regulatory environment: Data ownership, algorithm transparency, etc.
14. Long-term Vision for Web3 AI:
-Decentralization avoids AI intelligence monopolization, enables individual/community ownership
-Emergence of diverse AI agents and personalized AI services
-All participants generating revenue in an interconnected Web3 AI economic ecosystem
In summary, there are optimistic outlooks on Web3 AI enhancing data transparency, diversity, fairness and going mainstream, along with opinions on regulatory environment, technical challenges, and market capture.
“It helps you understand about Heurist”
Have a good day.
