Conference Presentation, Fireside Chat, Panel
A Fireside Chat with Himanshu Tyagi, CEO & Co-Founder of Sentient Foundation ft. FINOS
Sentient Foundation Mission & Structure
- Co-founded by Iman Shutyagi (formerly an academic professor in India and Paris, with prior roles at Paris 6).
- Focuses on building OpenAGI through community contributions to prepare for a future where contributors own and monetize open-source assets.
- Aims to solve the issue where current open-source contributions yield no long-term financial value despite lasting relevance over 8–9 years.
Proposed Mechanisms for Contributor Value
- Advocates for a "creator economy" model where software contributions are treated as tangible assets rather than just activist or "bragging rights" efforts.
- Proposes using blockchain to tokenize specific contributions (e.g., high-value data, hyperparameters, secret workflows) to enable direct attribution and monetization.
- Suggests a "high-value token" system where contributors add specialized data (e.g., medical records) to cheap compute tokens, increasing the token's value which then flows back to the contributor.
Open Source Safety & Alignment
- Defines open source safety through "open auditing," requiring models to declare their constitutional alignment and stand for community verification.
- Promotes "open red teaming" where the community audits GitHub/Hugging Face repos to detect deviations from declared safety guidelines.
- Argues that community involvement, not just regulation, is the primary solution for ensuring safe, beneficial AGI in an open environment.
Global Sovereignty & Market Dynamics
- Notes that while "digital sovereignty" and national models are inevitable, a global baseline of open source is required for continued AI advancement.
- Highlights DeepSeek (China) as a surprising and successful major champion of open source, challenging the notion that only Western entities can drive the field.
- Contrasts this with OpenAI's decision to close its models to secure revenue returns on $20 billion in capital investment.
- Predicts that future open-source sustainability will require robust revenue models similar to YouTube or TikTok for creators, specifically targeting younger, digital-native developers.
Regulatory Challenges (EU AI Act & Global Policy)
- Critiques current liability frameworks (e.g., EU AI Act's $10 million training cost threshold) as easily evadable through decentralized compute coordination.
- Identifies regulation as technically difficult to define and enforce due to the complexity of AI engineering and the potential for regulatory arbitrage.
- Expresses caution that regulation must not stifle the evolution of AGI, viewing the current era as a necessary transition for the next generation of the internet.
Future Outlook
- Emphasizes that the next generation of software engineers will be children (10–15 years old) who are "open source first" by nature.
- Foresees a future where information barter systems and trade secrets regarding AI training are formalized and traded via blockchain networks.
- Maintains that open source remains the core economic engine required to keep AI competitive globally across nations.