Rusl

EigenLayer peaked at $21.8B TVL in August 2025. It’s at $4.78B today. That’s $17B in capital that quietly decided the restaking narrative wasn’t worth the risk anymore. Here’s the part nobody wants to say out loud: restaking never had a revenue model. It had a story. The thesis was clean: restake your ETH, earn yield from AVSs securing real infrastructure on top of your base staking rewards. Double yield, same capital. New DeFi primitive. Except AVSs never generated the fee revenue to make that math work. So what fills the gap? EIGEN emissions. Token distributions. Dilution dressed up as yield. The DeFiLlama page says everything: EigenLayet (@eigencloud) > $4.78B TVL > $51.2M annualized fees > Protocol revenue: $0 — cumulative, 30d, 7d, 24h > $51.58M in annual incentives going out the door > Earnings running at -$9.42M Now look at @LidoFinance > $15.734B TVL > $763.09M annualized fees > $75.24M annualized revenue > $70.97M annualized earnings > $3.222B cumulative fees since launch > $4.27M incentives 1y — basically nothing Same category. Completely different business. Lido runs a simple 10% fee on staking rewards. No narrative. No emissions treadmill. No AVS roadmap that needs to materialize. Just a fee on a service people have used consistently since 2021 and $321M in cumulative revenue to show for it. That comparison is brutal if you’re a restaking bull. The capital wasn’t wrong to show up. Shared cryptoeconomic security is a real primitive — EigenDA, verifiable compute, oracle networks, these are real markets. But “real market” and “paying customers exist right now” are two very different things. The TVL came in chasing yield incentives. When the Kelp DAO exploit hit and $300M disappeared, $5.4B left the broader ecosystem in days. Mercenary capital moves fast when the yield isn’t real. ELIP-12 is EigenLayer acknowledging the problem: > Route AVS revenue back to EIGEN holders > Direct emissions toward fee-generating AVSs instead of idle restakers > Shift rewards to “productive stake” — tokens actively securing live services The direction is right. But there’s a July 1 unlock with ~741M EIGEN circulating, and the AVS revenue ramp isn’t there yet to absorb it. The restaking narrative isn’t dead. It graduated from “this is the future” to “show me the revenue.” The next winner in this space isn’t EigenLayer V2. It’s whichever AVS proves a real customer is paying with protocol cash flows instead of tokens. Until then, Lido’s fee model keeps printing $70M+ in annual earnings with $4.27M in incentives spent to get there. EigenLayer spent $51.58M in incentives and earned nothing. That’s the whole story.

x.com/i/article/2067…

Crypto card cumulative volume hit $8B as of June 17, 2026. Sounds like mass adoption. Look closer. Top 5 cards capture 87% of all volume. The long tail is statistical noise; dozens of products competing for 13% of a market that hasn't crossed $8B cumulative. The concentration map (30d volume): 1. @Nexo card — $281.71M 2. @ether_fi Cash — $80.54M 3. @RedotPay — $58.4M 4. @Cypher_HQ_ — $7.8M 5. @Bitget Wallet Card — $4.76M Everything below that line is fighting for fractions. What the data is actually saying: Crypto card adoption is not broad. It is deep in a handful of products and essentially zero everywhere else. This is not a market with ten viable competitors. It's a market with 5 real ones and a pool of products with live URLs and dead volume. > Nexo leads on volume but runs a custodial model: yield-bearing balances drive card spend, not wallet-native behavior > EtherFi is the only top-3 product with a self-custodial, borrow-to-spend architecture, different user entirely > RedotPay runs on pure stablecoin spend: no yield, no collateral, highest accessibility ceiling Three different models. Three different users. All three winning simultaneously means the market isn't unified; it's three separate use cases wearing the same product label.

The Teneo CLI setup guide for Manus is live. @ManusAI is a cloud-based autonomous AI agent. No local install, no terminal, no folder to create. The sandbox is a cloud computer with Node.js already configured. You skip the local setup entirely. From there it's three steps: 1) Create a free Manus account (starter credits included) 2) Send the one-line install prompt 3) Manus downloads the CLI, configures the daemon, and generates a wallet, usually under a minute The exact command is in the tutorial. Once installed, you can list and query 50+ Teneo agents in plain English: pull on-chain data, swap and bridge tokens, scrape social data. Some agents are free; paid queries settle per call in USDC via x402. You can also connect Manus to Telegram and run agents straight from chat. Guide: teneo-protocol.ai/get-started/se…

Allora CEO @nickemmons and @Metaversebizz of @web3pairpoint by @VodafoneGroup sat down to talk about what comes next for autonomous machines. As devices start to transact and coordinate on their own, they need to forecast the world, not just sense it.

Google and Blackstone’s new AI cloud company “will probably focus more on inference”, - @jai_das of @SapphireVC AI inference is going to be the most important aspect of the tech stack