#331 – Balaji Srinivasan: How to Fix Government, Twitter, Science, and the FDA

At a Glance

Balaji Srinivasan, angel investor, former CTO of Coinbase, former partner at a16z, and author of The Network State, systematically outlines his radical blueprint for technologically restructuring core institutions such as government, Twitter, science, and the FDA in this podcast episode. The most significant judgment of the entire episode is that Balaji believes existing large centralized institutions (governments, social media companies, scientific publishing systems, regulatory agencies) have fallen into "technological stagnation." Their inefficiency and rent-seeking behavior cannot be resolved through internal reform; the only way out is to replace them from the outside with a decentralized, crypto-based "network state" model.

Topic Sections

1. Government: From "Geographic Monopoly" to "Opt-in Network State"

Balaji Srinivasan argues that the modern nation-state, as a geographically based monopoly entity, has governance efficiency that lags behind the times and should be supplemented or even replaced by "network states" based on digital communities.

• Core Mechanism: Balaji proposes the concept of "The Network State," a group of people with shared values and governance rules who organize through digital communities (e.g., Discord, on-chain DAOs) and gradually acquire territorial sovereignty. It is not a company, nor a purely online community, but a "digital-first, territory-second" social contract.
• Historical Context & Analogy: He compares this to the replacement of "feudal estates" by "companies." Companies attracted talent away from estates by offering better job opportunities and mobility, ultimately dismantling the feudal system. Similarly, network states will attract "citizens" from existing states by offering superior governance, lower legal costs, and greater freedom.
• Data & Extrapolation: Balaji points out that the legal costs of existing states are extremely high (e.g., incorporating a company, handling immigration issues in the US), while crypto technology (e.g., smart contracts, DAOs) can reduce governance costs by orders of magnitude. He predicts that the first successful network state might start with a "startup visa" or "digital nomad community," gradually accumulating population and assets, and eventually reaching an agreement with a small country to obtain actual territory.

2. Twitter: Should Transform from a "Company" to an "Open Protocol"

Balaji believes that as a company, Twitter has fundamental flaws in its content moderation, algorithmic recommendations, and business model. The only solution is to transform it into a decentralized, open-source protocol similar to email.

• Mechanism Breakdown: He attributes Twitter's problems to "a single entity controlling everything." A protocol-based Twitter means: users own their identity and data (via cryptographic keys); algorithms are freely chosen by users or third parties; content moderation rules are decided by communities or users themselves; the business model could be micropayments or token incentives instead of relying on advertising.
• Contrarian View: Balaji explicitly opposes mainstream views like "Twitter needs a better CEO" or "Twitter needs better content moderation policies." He argues that any centralized solution is just "swapping one dictator for another" and cannot solve the fundamental incentive problem.
• Falsification Condition: If Twitter successfully transforms into an open, interoperable protocol and users migrate en masse, Balaji's view will be validated. Conversely, if Twitter continues to exist as a company and successfully solves its problems through internal reform, his judgment could be falsified.

3. Science: Replacing "Slow Peer Review" with "Fast Trials" and "Community Validation"

Balaji is highly critical of the existing scientific publishing system, calling its "peer review" process slow, conservative, and biased, hindering innovation. He advocates for "fast trials" and "community validation" to accelerate scientific discovery.

• Mechanism Breakdown: He proposes a "Science DAO" model. Researchers can quickly publish preprints and raw data, then use crypto tokens for "prediction markets" or "community voting" to validate results. Those who successfully replicate findings or provide key insights receive token rewards. This replaces the authority of traditional journal editors and a few anonymous reviewers.
• Data Chain: Balaji cites data showing that traditional scientific publishing takes an average of 1-2 years from submission to publication, and many important discoveries (e.g., mRNA vaccine technology) were suppressed for years before publication. He argues that such delays are fatal during crises like a pandemic.
• Extrapolation & Uncertainty: He acknowledges this model could generate "noise" and "false conclusions" but believes its speed and error-correction capability far surpass the existing system. His self-acknowledged uncertainty lies in how to design token incentive mechanisms to avoid "voting collusion" or "fraudulent replication," and how to get traditional academia to accept this non-mainstream certification method.

4. FDA: Approval Process Too Slow, Should Be Replaced by "Fast Trials" and "Community Validation"

Balaji targets the FDA, arguing its drug approval process is overly conservative and slow, delaying life-saving drugs from reaching the market. He advocates replacing parts of the approval process with smaller, faster clinical trials and community-driven "real-world evidence."

• Historical Context: He reviews the FDA's history, noting its original purpose was to prevent "fake drugs" and "poisons," but it has evolved into a bureaucracy whose "safety" standards actually prevent patients from accessing potentially effective treatments.
• Mechanism Breakdown: He proposes a radical version of "accelerated approval": allow patients, under informed consent, to use new drugs that have passed initial safety testing (e.g., Phase 1 trials), and record all treatment outcomes (real-world evidence) on a blockchain. This data can be analyzed in real-time to quickly determine drug efficacy. This is analogous to the "right turn on red" principle—allowing individuals to assume risk when information is transparent.
• Contrarian View: Balaji's view directly conflicts with the consensus of mainstream medicine and regulatory agencies, which emphasize the gold standard of large-scale, double-blind, randomized controlled trials. He argues that the gold standard is too rigid for rare diseases, pandemics, and novel therapies, and its cost (patients dying while waiting) is severely underestimated.
• Falsification Condition: If this "fast trial" model leads to a large number of ineffective or harmful drugs reaching the market, causing a public health crisis, Balaji's judgment is falsified. Conversely, if it significantly accelerates the launch of effective drugs with a manageable rate of adverse events, his view is supported.

Position Moves

Judgments Worth Remembering

1. "The Network State" is the "Company" of the Digital Age (Balaji Srinivasan): Just as companies dismantled feudal estates by offering better jobs, network states will attract "citizens" from existing nation-states by offering superior governance and lower legal costs. The core mechanism is "digital-first, territory-second."

2. Twitter's Problem Isn't the CEO, It's the "Company" Itself (Balaji Srinivasan): Any centralized solution is just "swapping one dictator for another." The only way out is to transform Twitter into an open-source protocol like email, giving users ownership of their identity, data, and algorithmic choice.

3. Scientific Publishing's "Peer Review" is a Bottleneck for Innovation (Balaji Srinivasan): The traditional review process is slow, conservative, and biased. He advocates replacing it with "Science DAOs" and "prediction markets," using token incentives to rapidly validate and replicate results, shortening the publication cycle from years to weeks.

4. The FDA's "Safety" Standards Are Killing People (Balaji Srinivasan): He argues the FDA's conservative approval process delays life-saving drugs, and its cost (patients dying while waiting) is severely underestimated. He advocates for "fast trials" under "informed consent" and "real-world evidence" to accelerate approvals.

5. The "Right Turn on Red" Principle Should Apply to Drug Approval (Balaji Srinivasan): Under transparent information and informed consent, patients should be allowed to assume the risk of trying not-fully-approved therapies. This is similar to the "right turn on red" rule—allowing individuals to make a choice while ensuring safety.

6. Crypto Technology is a Tool to Reduce Governance Costs (Balaji Srinivasan): Smart contracts and DAOs can reduce legal, accounting, and governance costs by orders of magnitude, forming the economic foundation for the network state. He compares this to "the internet's reduction of communication costs."

7. The First Network State Might Start with a "Startup Visa" (Balaji Srinivasan): He predicts the prototype of a network state could be a digital community providing legal and community services for digital nomads and entrepreneurs, eventually reaching an agreement with a small country to obtain actual territory. This is a falsifiable, incremental path.

New Arguments, Data, and Insights (Continuation Part 2/2)

I. On the Practical Path and Feasibility of the "Network State"

1. Sovereign States Have Already Started "Selling" Regulatory Services

Balaji lists several cases of sovereign states making deals with private entities, proving that "sovereigns are open for business":

Core Argument: If sovereign states can make deals with companies and currencies, why can't they make deals with online communities? The network state eventually gaining diplomatic recognition is not science fiction, but an extension of existing trends.

2. Israel as a Historical Precedent for the "Network State"

Balaji cites Theodor Herzl's Der Judenstaat as a key reference:

• 1897: Herzl published the book, proposing the idea of a Jewish state
• Subsequently: Established a fund (Jewish Colonial Trust) and a semi-annual congress (World Zionist Congress)
• Key Steps: First, a "cloud consensus" (book + congress), then "landing on the ground" (purchasing land in Palestine)
• 1947: Israel officially founded

Comparative Data:

3. Feasibility of the Network State's "Population Size"

Balaji provides key data to counter the intuition that "a state must be large":

• UN Member State Distribution:
• 12 countries with population < 100,000
• 20+ countries with population between 100,000 and 1 million
• 50-60 countries with population between 1 million and 10 million
• Only 14 countries with population > 100 million

Conclusion: Many social networks already have more users than the total population of some small countries. An online community with 1 million members can already "rank" among UN member states in terms of population size.