S11 E9 | Insights on the Future of DeSci w/Tarun Chitra (Gauntlet)

Diana Chen introduces Tarun Chitra, CEO of Gauntlet, to discuss decentralized science (DeSci) and its implications for scientific funding. Tarun critiques traditional funding models, particularly government grants, which have dominated since World War II and significantly advanced research. He highlights the role of private industrial research, like Bell Labs, in innovation but argues that government funding is better suited for exploratory research with uncertain outcomes. He critiques broad initiatives like DSi, suggesting that successful private funding requires specific goals.

The conversation addresses the evolution of science funding and the rise of questionable projects. Tarun explains how the life cycle of a scientist is influenced by the perceived relevance of their research area, impacting job prospects. He compares investing in academia to venture capital, emphasizing the importance of selecting young talent in emerging fields. Challenges in academia include the difficulty for assistant professors to secure ongoing funding, leading many to transition to tech or finance.

Tarun notes that science funding typically follows a milestone-based approach, contrasting it with D-Sci, which resembles ICO investing and may lead to misuse due to a lack of ongoing checks. D-Sci emerged as a response to decreased government funding post-Cold War, with wealthy individuals stepping in to support overlooked research areas. While private funding can accelerate research, it may limit the diversity of scientific exploration.

The discussion highlights the impact of reduced US funding for research, emphasizing that faster experimentation and less bureaucratic red tape foster innovation. There has been a notable rise in Nobel prizes awarded to non-academics, indicating a shift towards privately funded research. However, critiques of privately funded science point to its targeted nature and lack of accountability, particularly in longevity research.

Concerns about Decentralized Science Initiatives (DSI) are raised, particularly regarding their lack of clear direction and accountability. Traditional science relies on expert-driven hierarchical structures, contrasting with decentralized organizations that emphasize community voting. This fundamental difference presents challenges in reconciling the two approaches. The conversation also touches on the challenges in biology research, which is capital-intensive and has a low success rate, compounded by issues like the replication crisis.

Accountability in research is deemed crucial, with community science organizations potentially enhancing oversight. Recent scandals underscore the need for rigorous accountability measures. Speaker 1 emphasizes the importance of accountability in decentralized systems, critiquing DSI for neglecting this aspect. Misaligned incentives in DSI projects are discussed, with suggestions for milestone-based funding as a potential solution.

The conversation addresses the reputation system in science, which can be overly exclusive, making it difficult for new entrants with innovative ideas to gain trust. Funding inefficiencies are observed, particularly the concentration of resources in AI, leaving other important areas underfunded. The unique self-financing ability of crypto is highlighted, alongside concerns that over-concentration in funding could overlook significant innovations.

Speaker 1 draws parallels between the current state of crypto and the AI winter, emphasizing the need for funding to develop GPU software and architecture. They express skepticism about the vision of Web3 as a non-financial internet, arguing that the financial aspects of crypto are already sufficient. Speaker 1 explains that decentralized systems inherently require incentives and penalties, which leads to a financial appearance.

The conversation shifts to funding scientific research through a science DAO, suggesting that existing biotech funding models could be adapted. Concerns are raised about investing in projects without the ability to verify their success, alongside a recognition of the lack of immediate hope for improvements in the system but an optimism for long-term changes. The potential for AI oracles to assess experimental correctness and release funding is explored.

Looking ahead, Speaker 1 believes that contests on-chain involving AIs and open-source reasoning models will play a significant role in the future. He observes a competitive landscape where closed-source models are making notable advancements. In a lighter moment, Speaker 1 expresses concern about maintaining optimism after watching the new season of Black Mirror, to which Speaker 2 agrees. The conversation concludes with expressions of gratitude for the engaging discussion and hope for positive future developments.