Absolute Unit

@stacy_muur hft yeah humans are toast but mid-long term thesis humans can still have an edge

Solana's DEX trading volume is near all-time highs, but most of it is coming from bots. In January 2025, humans and bots were still competing in volume: → Real traders: $56.3B → Bots: $68.3B By March 2026, the picture looks completely different: → Real traders: $6.7B → Bots: $134.1B The 60/40 split flipped to 95/5. It's clear that the market optimized for speed. Manual traders can't compete with systems that can react in milliseconds and run 24/7. We also have AI agents that are further accelerating this shift. Their share will only grow from here. If you're still trading manually, you're falling behind.

@vascodm_ @jeffdfeng @jayendra_jog wow sorry to see you go, you'll kill it anywhere you head next

Today marks my final day at Sei. I'll always be grateful to @jeffdfeng and @jayendra_jog for believing in me early on. These four years have changed how I work and think. I got to solve problems I actually cared about, surrounded by some of the smartest people I know. When I first joined, Sei was just getting started. Today, it's one of the most successful Layer 1s in the space. I'm proud to have played a part in building it into what it is now. As for what's next, I plan to recharge, spend some time with my family, and figure out where the next great place to build will be. I'm bullish on what's ahead for Sei, and I'm sure the best is yet to come.

@catwychan liability is the biggest piece. if curators eat losses they're running an insurance company with no reserves. if they don't they lose all credibility after the first blowup. there's no good answer

Tend to agree — the curator business model is structurally asymmetric. Back of napkin math: - TVL managed: ~$200M - Gross USDC lending yield: ~5% - Performance fee: 10% of yield Revenue = $200M × 5% × 10% = $1M/year On the risk side: - One bad credit event could mean massive reputational damage - And are curator truly not liable to financial losses? - If curators compensate: wipes of years of revenue —> a even worse business model - If they dont: future business likely gone - And don't forget legal costs either way Heads you make $1M. Tails you lose $Ms. Not a great bet.

@Route2FI 5 days into charging real fees and they're already sitting behind tether, circle, and hyperliquid. that's not a marketing problem that's a product that people are willing to pay for. most protocols can't say that

On March 30, Polymarket rolled out a new fee model that adds taker fees to almost all trading categories (previously, a big chunk of volume was effectively free or very lightly monetized). Only 5 days later, Polymarket ranked 4th, behind Tether, Circle, and Hyperliquid.

@0xShual imo they just turned fees on though. most protocols don't even get to the point where they can charge without losing all their users. the fact that volume held is the real signal here not the margin

Fun fact: polymarket's daily revenue doesn't even cover their marketing costs (let alone all their other expenses)

@Eli5defi defi is arguably the purest version of this. the backend is open, composable, and nobody can revoke your access. the entire value capture is in the customer layer and how fast you integrate new primitives. we're already seeing this play out

$401M in revenue. 2 employees. 0 investors. $20,000 to start. Sam Altman kept predicting the arrival of the one-person, billion-dollar company. Most people filed it under "eventually." It arrived early. Matthew Gallagher built Medvi from his Los Angeles living room. With just $20,000 and two months to launch, he and his brother grew the company to $401M in revenue by 2025. They are pulling in a $65M net profit at a 16.2% margin. For context, Hims & Hers does similar revenue with over 2,400 employees and half the margin. But Medvi isn’t the story, it’s the proof of concept. The real question is which sectors carry this same architecture, because this ultra-lean model doesn't work everywhere. There is a specific structural formula that makes it possible: → A plug-and-play backend: Regulated infrastructure you can partner with rather than build. → A separable customer layer: The ability to decouple the user experience from the backend operations. → AI-driven customer acquisition: Growth fueled by programmatic ads, content, and SEO. → AI-absorbed operations: Customer service, routing, and triage handled by voice agents and chatbots. → Scale-agnostic margins: Unit economics fat enough to survive without massive volume. Gallagher owned the front. Partners like CareValidate and OpenLoop Health (who handle licensed physicians, pharmacy fulfillment, and compliance) owned the back. AI built and ran everything in between. The moat isn't technology. The moat is knowing which half to own. — ➠ So, which sectors fit the blueprint? ❶ Telehealth Niches This is the obvious first row. Think mental health, dermatology, men's health, and women's hormonal care. The backend medical and pharmacy providers already exist. The customer acquisition layer is wide open. Expect 50 invisible copies of Medvi in the next 18 months. ❷  Insurance Distribution Carriers and MGAs absorb the underwriting risk, hold the licenses, and handle the claims. A single operator can build a hyper-targeted acquisition layer, pipe into carrier APIs, and own the customer relationship with near-zero overhead. ❸ Crypto & DeFi This might be the most natural fit of all. The backend is a public blockchain, nobody owns it. Uniswap's liquidity, Aave's lending pools, and Chainlink's price feeds are permissionless infrastructure accessible from day one. No partnership negotiations are required. This leaves the customer layer wide open for: → A one-person yield optimizer routing capital across protocols. → Cross-chain games running on crypto backends. → A DeFi signal service monitoring on-chain flows via AI for institutions. The moat here isn't protocol access; it's data aggregation, trust, and the speed at which you ship integrations when new primitives launch. — ➠ But The Most Obvious Thing → Gaming (The Picks and Shovels) People are sleeping on gaming, but the opportunity isn't building games, that's a lottery. The opportunity is selling to studios. AI drops asset production costs by 80%. Midjourney handles visuals, Suno handles ambient music, and Claude handles NPC dialogue trees. The solo play is generating high-quality asset packs, procedural level generation APIs, or AI-driven localization pipelines, and licensing them to mid-size studios that can't afford dedicated tools teams. One well-built Unity asset pack sells 10,000 times with no additional labor. The margins approach 100%. The moat here is taste. AI is available to everyone; the operator who ships purpose-built tools rather than generated slop wins the relationship. — ➠ The Bottom Line The pattern is consistent. Find a public or partner-accessible backend. Build a separable customer layer. Use AI to drive acquisition and operations. The backends are already built. The AI stack is already cheap enough to run on a laptop. You just have to pick the right half to own

@CredibleCrypto 🤔

Every one of you still here needs to read this. Some more than others..

@Yippun wait until people find out this happens on every chain lol. TVL without adjusting for rehypothecation is basically meaningless ethereum's TVL is just orders of magnitude more resilient

Folks now realise the TVL on Solana are just routing of the same capital through multiple protocols?

@milesdeutscher crazy because I feel like everyone I know is talking about it

Only 3% of households are paying for AI.. And you think you're late to the party? Still the biggest opportunity on the planet right now.

@nickwh8te every project that actually made it went through a phase where the timeline clowned on them. the ones that kept shipping anyway are the ones still here. I respect the conviction.

Zcash's comeback is extremely inspiring to me. After the early hype faded, the core team and community persisted through YEARS of seeming irrelevance, sticking to their vision of unstoppable private money. That conviction is now paying off. Celestia has the same DNA. A cypherpunk vision of abundant, unstoppable blockspace, verifiable by anyone in the world. The early hype has faded, but we have a cracked core team committed to the mission, ready to keep going for however long it takes. We'll see it through.

@wagmiAlexander not real until there's at least 30 different 900% apy fruit themed LPs

DeFi Summer is coming back to Mainnet Ethereum.

@lex_node the attackers are literally counting on you not stopping to question why something feels slightly off I feel like most people get compromised when they're tired, rushed, or stressed

being on 'autopilot' is your worst enemy in anything NPC mode, which is what society is training you to be in all the time (every time you go through an airport security checkpoint, etc.) get sleep, have good cogsec, retain your agency...so you think twice before clicking...

@andyyy vanguard, merrill, wells fargo, jp morgan self directed. maybe nw mutual & edward jones on the smaller side. we're running out of holdouts honestly

After Schwab, who else is left to onboard out of the major brokerages???

@stacy_muur imo best time to invest is now, assuming you have some level of confidence the project has enough runway to last

Sitting sidelined doesn't seem like a bad idea right now. Touch grass. Wait it out. The weather is getting better anyway ¯\_(ツ)_/¯

@TheDeFinvestor not surprising tbh. if your only revenue model is selling your own token you were always on borrowed time. real protocol revenue is the filter now

21 projects shut down in 2026 do far. Unfortunately I think this trend will continue. There are too many projects with an economic model that relies only on dumping their own tokens. The only altcoins still worth betting on on are those with real PMF and revenue.

@blknoiz06 15 is generous

crypto twitter pulse check

@0xNairolf just like long term recessions are bullish, need to clear out the stragglers to make room for new growth

unpopular opinion: projects shutting down is bullish for crypto

Spectra Q1 2026 Ecosystem Report A lot happened in Q1 - here's where we are (1/9)↓

@CharlieStLouis @BlockAnalitica @defiscan_info @curatorwatch @CredoraNetwork @justz_eth 👀

the demand for this is real we ran sessions with 50+ DeFi teams at EthCC and this was one of the most consistently raised topics the risk infrastructure already exists: @BlockAnalitica, @defiscan_info, @CuratorWatch, @CredoraNetwork, individual protocol dashboards, etc what’s missing is a neutral aggregation layer. one place. one score. L2Beat but for DeFi risk who’s building it or wants to collaborate on it? 👀

@0xngmi LlamaAI has been one of my biggest productivity unlocks

we found out that fund analysts and managers are the ones who use LlamaAI the most but still every day im seeing ppl on twitter spend hours building charts that could be done in minutes in LlamaAI so we are opening up to everyone for a short time to see if it removes friction