After almost twenty years on the platform, EFF is logging off of X. This isn’t a decision we made lightly, but it might be overdue. 🧵(1/5)

HOW IS NOBODY TALKING ABOUT THIS!?!?!?!? THEY PRINTED 5 BILLION OF THEIR OWN TOKENS THEN WITHDREW IT AS USDC!?!?!?!??!

🚨 BREAKING: The FBI has successfully extracted deleted Signal messages from a suspect's iPhone via notification storage, the place where all your notifications are stored for up to one month. Notification storage stores data from all messaging apps, it's a big flaw in iOS. But there's a way to turn it off...

Some thoughts on Mythos, cybersecurity, and the next 12 months ahead. Anthropic made an incredible announcement this week. Their new model is so good at hacking that they think it's irresponsible to release it to the public. This was followed by both panic and skepticism. Anyone who has data or money worth protecting is panicking and wants to make sure their software is defended before the attackers get them. There's also a large number of skeptics, including very senior cybersecurity professionals. Where am I at? I run a cybersecurity company that works with highly targeted companies; our customers like Coinbase or Uniswap are targeted because of the prize. Hackers get to steal funds if they're successful. The other prize that hackers go for is data. In today's era, data has suddenly become liquid because AI labs want more and more data to feed this insatiable super-intelligent machine that's being built. Three recent hacks illustrate where this is going. 1. A Chinese government agency was hacked, and 10 petabytes worth of data is being offered for sale by hackers. 2. Mercor, a Silicon Valley data company, was hacked, and hackers stole ~5 terabytes worth of data. It's speculated that Mercor paid the hackers and took it off the market. 3. Drift, a crypto trading platform, was compromised by North Korean threat actors and lost $285 million. If you're following cybersecurity, you'll see this trend is only going to increase this year. We've seen the increase in capability early last year and have been building an autonomous bug hunter named Apex. Around January, we realized that code security as we know it is on track to be solved. There's still a tremendous amount of work left, but we see a clear path ahead of us where the problems in front of us are engineering problems. Apex has already found and saved security bugs, if exploited, would have resulted in losses over $100B in value. Where do I stand? Anthropic isn't bluffing: the capability with the new models are increasing at an alarming rate. If you are someone with money or data to protect, this needs to be your #1 priority this year. We're entering a time of intense turbulence.

$23.5B in prediction market notional traded in March 2026. In a single month. @Kalshi drove $13B. @Polymarket followed with $10.5B. And now @HyperliquidX’s HIP-4 could eclipse both, living up to its name: The House of All Finance. Here’s our comparative analysis. ⤵ — ➠ Kalshi - The Regulated Exchange Kalshi is what you get when prediction markets go through the U.S. regulatory process and come out the other side. ▸ You sign up, verify your identity (KYC required), deposit real USD, and trade yes/no contracts on events like elections, Fed rate decisions, or sports outcomes. ▸ Each contract pays $1 if the event happens, and $0 if it doesn’t. You buy at the current implied probability, then sell before expiry or hold to settlement. ▸ Kalshi resolves outcomes using predefined reference sources. Fast. Centralized. Reliable. ▸ Fees are straightforward: roughly $0.0175–$0.02 per contract near 50¢ probability, dropping toward zero at the extremes. ▸ The constraint is real: Kalshi can’t list markets on war, death, or anything the CFTC considers off-limits. Its market catalog is narrower by design. ▸ What it does well: $13B notional in March 2026. The largest regulated volume. Institutional trust. Fiat on-ramps. If compliance matters to you, there’s no alternative. — ➠ Polymarket - The Crypto-Native Market Polymarket runs on Polygon but plans to migrate to its own layer soon. Anyone with USDC (soon Polymarket USD) and a crypto wallet can trade, as long as they’re not in blocked countries. ▸ Connect a wallet and buy YES/NO shares on any listed market. Share price = implied probability (60¢ = 60% chance). ▸ No KYC for global users. No geographic restrictions outside limited U.S. access. ▸ Anyone can create a market. That’s the power, and also the risk. ▸ Settlement goes through UMA’s Optimistic Oracle, the Markets Team, and Chainlink Oracle: a proposer submits the resolution, there’s a dispute window, and if contested it goes to a community vote. It works most of the time, but can take hours to days when outcomes are genuinely ambiguous. ▸ The fee structure is more complex. Fees are parabolic: near-zero at extreme prices ($0.01/$0.99), but steep in the mid-range. ▸ At 50% probability, the round-trip cost hits 1.80% on Crypto markets and 1.50% on Sports. Geopolitics is the only zero-fee category. Finance markets have the highest maker rebate at 50%. ▸ What it does well: $10.5B in March 2026 volume. Broadest market variety on earth. Permissionless. Transparent on-chain trades. The go-to platform for global, crypto-native event trading. — ➠ HIP-4 - The Composable Primitive HIP-4 is Hyperliquid’s native outcome contract system, built directly into the same trading engine that handles $178B+ in monthly perpetual volume. As of April 2026, it’s still on testnet, but the architecture is worth understanding now. ▸ Contracts trade between 0 and 1. Price = implied probability. At expiry, they settle to 1 (event happens) or 0 (it doesn’t). ▸ You post 100% of your maximum possible loss upfront. No leverage. No margin calls. No liquidation risk, ever. ▸ Settlement is on-chain and deterministic, using curated reference data, resolving in ~0.2 seconds on HyperL1. No dispute window. Here is the part that changes the math on everything else: ▸ Every position you hold on Hyperliquid: spot, perpetuals, and now outcome contracts shares one unified margin account. ▸ The system automatically recognizes offsetting risk. If you hold a short ETH perp and a YES outcome contract on “ETH rallies above $X,” those positions reduce each other’s margin requirements automatically. ▸ Your outcome position isn’t dead capital. It functions as collateral across your broader portfolio. ▸ Every other prediction market silos your capital. Your $500 in Polymarket is invisible to the rest of your trades. On HIP-4, that same position stays alive and productive inside your full account. ▸ Fees follow Hyperliquid’s volume-tiered structure: very low at base (0.045% taker / 0.015% maker), compressing further for active traders. ▸ Fees apply only on close or settlement, not on open. Builders who deploy permissionless markets in Phase 2 earn up to 50% of the fee revenue generated. ▸ Phase 2 permissionless deployment requires staking 1,000,000 $HYPE tokens as a bond. That creates three simultaneous effects: a supply sink for HYPE, aligned builder incentives, and a filter against low-quality market creation, without central gatekeeping. — ➠ Bottom Line Kalshi wins for regulated, fiat, U.S.-friendly trading with strong liquidity in politics/sports. It is the most “institutional-grade” but constrained by rules. Polymarket dominates crypto-native volume and variety globally. It offers the most open market creation today but faces oracle delays and fee friction in high-uncertainty events. HIP-4 is the most innovative structurally. It turns prediction markets into a composable primitive inside the world’s largest perp DEX. The unified margin + no-liquidation design solves the biggest pain point of traditional prediction platforms (dead capital). Once on mainnet with permissionless markets, it could pull high-volume DeFi traders who already trade billions in perps on Hyperliquid.