-
Vitalik Buterin created Ethereum at age 19 as a general-purpose blockchain — a shared, ownerless computer in the cloud that anyone can build applications on, replacing the need for trusted third parties like Facebook or banks with transparent, auditable code enforced by thousands of independent computers.
- Unlike Bitcoin, which is designed primarily as digital money or a store of value, Ethereum allows developers to deploy “smart contracts”: pieces of code that live on the blockchain and execute exactly as written, with no owner who can later change the rules.
- This creates a system where users own their data and applications, and all activity is publicly verifiable — but the trade-off is extreme inefficiency, since every node on the network redundantly verifies every transaction.
-
Ethereum trades performance for security and decentralization.
- The network processes only about 15–20 transactions per second because every transaction is verified by tens of thousands of computers in parallel — more like sending a text message redundantly to thousands of people than running code on a single efficient server.
- This redundancy is what makes Ethereum trustworthy: no single entity controls it, no back doors exist, and anyone can audit the code and data.
- The practical consequence is that only high-value transactions can afford to use Ethereum directly — a $10,000 trade can justify $50 in fees, but trading $3 magic swords in a game cannot.
-
Key limitations of Ethereum today are throughput, latency, and privacy.
- Throughput: Limited block space means transactions compete in an auction, making block space expensive.
- Latency: Transactions take roughly 30 seconds to confirm (improving toward 10–12 seconds), making Ethereum unsuitable for real-time applications like video games.
- Privacy: The base blockchain is fully transparent — anyone can see every transaction. Zero-knowledge proofs (as used in Zcash) can restore privacy by proving rules were followed without revealing transaction details, but this adds complexity and doesn’t yet hide all metadata (e.g., that someone interacted with a contract at all).
-
Scaling solutions — sharding and rollups — aim to increase throughput while preserving decentralization.
- Sharding splits verification work across subsets of nodes (e.g., 1,000 nodes verify a transaction instead of 100,000), similar to how BitTorrent distributes files. This is technically complex and risks some centralization.
- Rollups (Layer 2) execute transactions off-chain, then post compressed data (like a “zip file” of transactions) back to Ethereum’s Layer 1. This achieves roughly a 100x improvement in efficiency — potentially raising throughput from ~20 to ~5,000 transactions per second — while still inheriting Ethereum’s security guarantees.
- Ethereum’s philosophy is to keep the base layer simple and push innovation to upper layers, unlike chains that bake features like social recovery directly into the protocol.
-
Decentralization is a cultural and technical choice that Ethereum prioritizes over raw performance.
- Ethereum’s community insists that nodes should be runnable on a laptop, maximizing the number of people who can independently verify the chain. This makes it much harder for any small group to collude and force unwanted protocol changes.
- Chains like Solana achieve higher throughput by requiring powerful hardware (closer to supercomputers), which reduces the number of potential node operators and increases centralization risk.
- Vitalik argues that verifiability must be the default behavior of participants, not just technically possible. If running a node is hard, people gradually delegate to trusted intermediaries, eroding decentralization over time — a cultural loss that’s difficult to reverse.
- New users in bull markets tend to ignore decentralization and privacy until a crisis (e.g., government action, hacks) makes those properties matter. The most secure, decentralized chains (Bitcoin, Ethereum) tend to retain the most value over time.
-
There is a free-rider problem in blockchain ecosystems: developers often fork existing chains rather than build on shared ones.
- Because blockchains are open source, it’s tempting to copy Ethereum and launch your own chain with your own token, capturing all the value for yourself rather than contributing to a shared ecosystem.
- Ethereum has partially countered this through ERC-20 tokens (letting projects issue assets on Ethereum) and Layer 2 rollups (letting projects build scalable applications while still using Ethereum’s security).
- Innovation has been fastest at Layer 2 and the application layer (permissionless — anyone can build without coordination) and slowest at Layer 1 (changing the core protocol requires broad community consensus).
-
Ethereum’s native token (ETH) is evolving into a store of value alongside its utility as “fuel” for the network.
- With EIP-1559, a portion of every transaction fee is burned (destroyed), and with the transition to proof of stake, ETH issuance is projected to go negative — making ETH potentially deflationary.
- Vitalik sees ETH the asset and Ethereum the application platform as synergistic: a stronger ecosystem makes ETH more valuable, and a more valuable ETH strengthens the ecosystem’s security.
- He doesn’t prescribe whether ETH “should” be money — he sees that as a community decision — but notes that the world will likely have many stores of value (Bitcoin, ETH, real estate, stocks, art, NFTs), and having dozens of viable options is itself a form of decentralization.
-
Ethereum tries to serve both conservative and experimental users simultaneously.
- Stability-seeking users (e.g., MakerDAO, Aave) can rely on Ethereum as a “digital rock” — a solid, predictable base layer.
- Experimental users can build on Layer 2 rollups, which offer more scalability and flexibility while still anchoring to Ethereum’s security.
- Whether Ethereum can successfully balance these competing demands — innovation vs. stability, scalability vs. decentralization — will be determined over the coming decade.
Vitalik: Ethereum, Part 1
Naval • • 46min → 4 min • #10