Most VPNs just shuffle your traffic through one hop and call it "privacy." NymVPN runs a 5-hop mixnet. Your traffic gets layered, mixed, and scrambled across a decentralized network so that not even Nym can link your payment to your activity. - Rust-based, fully open source, audited - WireGuard (AmneziaWG) for speed when you need it, mixnet for when you actually need to disappear - Pay with Monero, Bitcoin, Lightning, $NYM, or even cash - zk-nyms credentials = unlinkable payments - No centralized logs. None. - 10 devices, 70+ locations 1 month free with my referral. 2-year plan works out to $2.39/mo after that. If privacy actually matters to you, stop giving your money to VPN companies that log, hand over data on request, and run closed-source binaries you can't audit.

Pricing | Nym

See NymVPN’s pricing plans and choose the most anonymous VPN for private browsing. 7 day free trial and up to 70% off plans.

#NymVPN #privacy #mixnet #cypherpunk

The EU AI Act reaches full applicability for high-risk systems on August 2, 2026. Four weeks out. Look past the headlines and the compliance checklist gets specific fast: a risk management system, record-keeping obligations, human oversight, and demonstrable accuracy, robustness, and cybersecurity for every high-risk AI system on the market. Non-compliance tops out at 35 million euros or 7% of global turnover, and the Act reaches any provider whose system touches the EU market, regardless of where the company is based. Here's the practical problem most teams are sitting on. AI agents in production today can call tools with no isolation and no durable record of what ran, what it touched, or who approved it. When someone asks for the audit trail, the honest answer is often a chat transcript. That's the gap WasmBox closes. Every tool call an agent makes runs inside a sandboxed WASM environment with capability-scoped permissions, and every execution writes an Ed25519-signed, tamper-evident audit log entry. Not a nice-to-have bolted on after the fact. The actual record-keeping and oversight infrastructure the Act is asking for, built into how the agent runs. If your agents are already executing tools in production, the question worth answering this month is simple: can you produce a signed record of everything they did. #EUAIAct #AICompliance #AIAgents #WASM #AIgovernance #AIsecurity #ResponsibleAI #AuditTrail

Visa, Mastercard, Stripe, Coinbase, BlackRock and 135+ more just launched Open USD (OUSD). The mechanic: nearly all reserve yield flows back to the partners driving adoption, instead of one issuer keeping it like Circle and Tether do. That's a real shift in stablecoin economics. Still a permissioned token though. Only approved partner businesses can mint or redeem it. Open economics, not open access. Not live yet, no fixed date beyond later this year. Watching how much volume it actually pulls off Circle once it ships. #openusd #ousd

@alp any chance to pin the "private note taking" to this screen? would be v helpful on desktop

@alp something weird in noornote:

#mullvad is getting on my nerves performance wise. They seem to be too busy to donate to far right groups

Private AI that doesn't spy on you. Uncensored models, no login, your data stays yours. Try it: 💯

Join NanoGPT

Create your NanoGPT account and get access to leading AI models for chat, coding, images, and video.

2️⃣

Venice.ai

Venice | Private, Uncensored AI

Venice is the easy app for private, uncensored AI conversations and image generation. Try for free with no log-in needed.

🕝https://ppq.ai/invite/9bc57632 #venice #nanogpt #ppq

StarkWare unveils Starknet post-quantum roadmap, calling it crypto's 'strongest' to date https://www.theblock.co/post/406677/starkware-unveils-starknet-post-quantum-roadmap-calling-it-cryptos-strongest-to-date

when you hate marketing, but as a solo founder you need to cover all: "you have 400 followers, work on that first and then come to pitch" 🤗

KYC/KYB is the worst and most annoying shit in history

Just setting up my #Sidecar 🍸

GLM 5.2 Thinking TEE is now available! It runs GLM 5.2 with thinking enabled inside a Trusted Execution Environment for harder coding, agentic workflows, engineering optimization, and sensitive reasoning-heavy tasks where stronger integrity assurances matter.

Join NanoGPT

Create your NanoGPT account and get access to leading AI models for chat, coding, images, and video.

#glm #ai

Fable is going to be redirecting coding task to Opus 4.8 🤡🤡🤡

We read a paper on agentic loops, took its one load-bearing lesson into our company of agents, and pointed the result at our own work the same day.

MetaEnd

The evaluator is the floor

AI agents shouldn't grade their own work. We built a skeptical evaluator that active-tests every output, catching critical compliance bugs that a s...

#airesearch #ai

omp has a new /advisor role, and I use it currently with my free google gemini cli - confidence boost up by at least 5% IMO. pretty cool stuff

Wait what. Captain Tsubasa is actually winning???

The evaluator is the floor We read a paper on agentic loops, took its one load-bearing lesson into our company of agents, and pointed the result at our own work the same day.

MetaEnd

The evaluator is the floor

AI agents shouldn't grade their own work. We built a skeptical evaluator that active-tests every output, catching critical compliance bugs that a s...

#ai #agents

How's Tsubasa holding up against #Brazil?

🤡🤡🤡 Time to show him the 🖕 ZAI coding CLI:

GLM Coding Plan — AI Coding Powered by GLM-5.3, GLM-5.2 & GLM-5-Turbo for Agents & IDEs

Use GLM models like GLM-5.3, GLM-5.2 & GLM-5-Turbo for AI coding in Claude Code, Kilo Code, Cline, OpenCode, Clawdbot/OpenClaw and more. Plans from...

Nano private chat + memories:

Join NanoGPT

Create your NanoGPT account and get access to leading AI models for chat, coding, images, and video.

Both subscriptions are still cheaper than the clowned one View quoted note →

See your site the way an agent sees it -

metaend Grade — Agent Readiness Checker

See your site the way an AI agent sees it.

#ai #agents