Deniz.eth

After a year of work I can finally share what me & my awesome team have been building: The *ultimate* crypto experience. Read news, participate in DAOs, research NFTs, swap tokens.. Do ANYTHING & EVERYTHING with hundreds of drag&drop services&dapps alphaday.com /1

@gregskril 😂😂 What model is this?

new favorite AI interaction

@bricexeth @AnthropicAI A huge update might be coming but I really think its a failure to scale up to meet the unforeseen popularity of their latest models. On Openrouter Opus4.6 is by far the most widely used llm in its price range, nothing comes remotely close to it.

@DenizOmer @AnthropicAI Maybe a huge update is coming

12+ incidents over the last 24 hours for Opus+Sonnet, @AnthropicAI services buckling under huge growth or something else??

@leolanza Check any of @TheAhmadOsman's post on this. GPU is the clear way to go: x.com/TheAhmadOsman/…

Is it better to use a Mac Studio with 256GB of RAM or a GPU for running a local LLM?

@Shaun__Furman @bradmillscan Could you expand on the systems and pipelines or at least give us a direction to read up on? Thanks!

QMD But I basically just use memory for conversations. Everything else is systems and pipelines. I don't want it to hallucinate from memory. I always want verified info I get it to pull out key decisions, human moments, the reason we make decisions and new canonical structures almost daily.

For OpenClaw power users who are using QMD or Obsidian ... what do you use your memory.md file for?

@hashnode Hey Hashnode, when are you going to unblock my blog posts that were erroneously marked as 'violating community guidelines'?. Unblock them so that I can migrate them to a better blog host. Thanks!

We missed you too. New Hashnode, same home.

@Sandeepg33k Hey Hashnode, when are you going to unblock my blog posts that were erroneously marked as 'violating community guidelines'?. Unblock them so that I can migrate them to a better blog host. Thanks!

@sawinyh Yep, and thanks to openclaw it's off the shelf again and being speed run through the leanfire product cycle 😅

Every developer has that one side project they're convinced would work if they just had 2 more weeks. It's been 2 years. You know which one I'm talking about.

@DennisonBertram @claudeai Looks like it's just on, my openclaw context windows switches to 1M overnight

@claudeai Do we have to switch to it? Or it's just on?

1 million context window: Now generally available for Claude Opus 4.6 and Claude Sonnet 4.6.

@sawinyh @AlphadayHQ Thanks for the recommendation, just added it to my to-read-next list!

@AlphadayHQ Layered Money clicked for me in ways most bitcoin books don't. The gold to dollars framework just lands differently when you've watched stablecoins grow

📚 Alphaday Weekend Reads — the 5 highest-rated blockchain books on Goodreads: 🔍 Tracers in the Dark — Andy Greenberg (⭐️ 4.49) The true story of how investigators cracked crypto's biggest criminal cases 🔧 Inventing Bitcoin — Yan Pritzker (⭐️ 4.40) The clearest technical explainer of how Bitcoin actually works 🧠 Mastering Bitcoin — Andreas Antonopoulos (⭐️ 4.31) The OG technical bible. If you're building in crypto, you've read this ⚖️ Blockchain Bubble or Revolution — Neel Mehta (⭐️ 4.30) Written by Harvard & Google engineers. Balanced, thorough, no hype 💰 Layered Money — Nik Bhatia (⭐️ 4.30) Connects gold → dollars → Bitcoin in a framework that actually clicks We've read all five — Mastering Bitcoin is still our all-time favorite. Which ones have you read? 👇

@JulianGoldieSEO Anyone else stuck with this warning after the update?

OpenClaw 2026.3.7 just dropped. It turned OpenClaw into a real AI operating system. Here’s the upgrade path. → Route tasks across GPT 5.4 and Gemini 3.1 Flash Lite. → Keep ACP bindings alive through restarts. → Deploy with slim Docker multi-stage builds. → Store secrets safely with secret ref. → Swap memory systems with pluggable context engines. Save this video, you’ll understand why this update is a turning point. Want the SOP? DM me. 💬

@openclaw Thank you for the most unstable release in a while...🤦

OpenClaw 2026.3.12 🦞 🎛️ dashboard v2 — slick new control UI ⚡ /fast mode for models 🔌 ollama/sglang/vllm → plugins (core goes on a diet) 🛡️ device tokens now ephemeral ⏰ cron + windows reliability fixes touch grass? nah, touch main 🌿github.com/openclaw/openc…

@RibbeFelipe @petergyang @nateliason Ask your bot to diagnose the issue and implement a solution?

@petergyang @nateliason My gateway crashes all the time, specially when calling browser tool. Any tip to make it better?

My top 5 takeaways from @nateliason on how he set up his OpenClaw bot to build a business that makes $4,000 / week: 1. Set up a 3-layer memory system first The 3 layers are a knowledge graph (facts about your projects and life), daily notes (what's actively happening), and tacit knowledge (your preferences, patterns, and security rules). 2. Give your bot its own accounts — never your personal ones. Felix has his own X/Twitter, Stripe, GitHub, and email accounts. If something breaks, the blast radius stays contained. 3. Every time your bot asks you for help, ask it: “How can I remove this bottleneck?” Each time Felix needed Nat to do something manually, Nat asked it this question. “The more I asked, the more capable he has become.” 4. Use Telegram group chats to work on multiple projects at once Create a separate group chat for each project and add your bot to it. This way, context stays clean in each chat. 5. Cron jobs are super useful for OpenClaw Felix has 6-8 cron jobs for X/Twitter alone. Without cron jobs, your bot just waits for instructions. With them, it becomes proactive. 📌 Watch my full episode now on how Nat set up his OpenClaw bot to build its own business: youtu.be/nSBKCZQkmYw

@iamliamsheridan Prompt

outbound will never be the same after Claude Sonnet 4.6. dropped today. most people will use it wrong. there's a fine line between AI that makes you faster and AI that makes you sound like everyone else. we took a different route. we use Claude to make our human outbound operators 3-5x faster - not to replace them. that's why i put together a doc of Claude prompts built specifically for outbound pipeline. designed for $10K-$250K+ B2B deals. inside you'll find prompts to help your team: → analyse accounts and map buying committees in minutes → run 10-minute research before writing a single email → generate 3-5 cold email variants per angle without sounding like a bot → draft follow-ups based on specific replies and objections → prep talking points before outbound calls → tighten copy to avoid spam triggers and protect deliverability these aren't generic chatgpt prompts. they're built to make Claude think like a seasoned outbound operator plugged into a human-led engine. the best part? paste them straight in and go. comment "PROMPTS" and i'll send it over.

@SCOOT_vibes @AlexFinn Can you elaborate on how to use discord to organise this?

@AlexFinn You could just use discord to organize all this. Bonus points if your OC creates, updates and categorizes everything itself.

Sorry this happened, it really sucks! Me and my colleague at Alphaday were targeted in the exact same way (except it was a bd call) a few months ago. I sensed something was off and dropped off the zoom immediately. My colleague had to format his pc after following the zoom update instructions.

I’m sharing this so others don’t fall into the same trap... 👉 Please block and report my old Telegram (@bby_boyeon). 👉 My new Telegram(@Boyeon0x) is now active please DM me there to reconnect🙏

My @telegram account was hacked through a highly sophisticated social engineering and malware attack. It took me a while to trace back how I got compromised and I’d like to share my experience so others can avoid the same mistake.

@CryptoPulse9 @bryan_johnson easy.. Fear and Loathing in Las Vegas

@bryan_johnson What is the best movie for @bryan_johnson to watch today during this mushroom trip experiment?

Will start around 10:15am. Last night I was asleep by 9. Woke up 5 am. Feel fantastic. 10k lux in eyes. Some protein, extra virgin olive oil, and 60 super molecules. Worked out. Finishing pre mushroom brain measurement with Kernel Flow.

@thisdudelikesAI Guide

Holy shit... Perplexity just did something wild 🤯 They dropped “Perplexity at Work” their official guide to actually getting more done with AI. Not another “productivity tips” doc this is the real framework their own teams use to: → Block distractions & reclaim focus → Scale yourself like a 5-person team → Turn AI from noise into results It’s clean, practical, and honestly the most useful thing I’ve read on using AI for work not just chat prompts. Comment “Guide” and I’ll send you the full PDF (it’s 100% free from Perplexity)

@Adilson_Ai Steal

HOLY SH*T… This AI Agent does everything 🤯 Built in n8n : 🔁 Clones viral TikToks ✍️ Rewrites w/ GPT-4o 🎥 Auto creates avatar videos 🎬 Adds captions & edits 📤 Posts to 9 platforms (TikTok, IG, YT, X…) 🔁 Like+RT ✅ Reply “Steal” 🤝 Follow me & I’ll DM you workflow FREE.

In this polymarket isn't the chance of ETH hitting $5,000 actually the sum of everything above it plus the chance of hitting $5,000? Ie. the chance of hitting $5,000 = chance of hitting $17k + $14k + $10k + $8k ... all the way to $5k? Basically, chance of ETH hitting $5k is far above 52%?

Ethereum now has a 52% chance of hitting $5,000 in 2025 on Polymarket $5K ETH feels EXTREMELY cheap for EOY Not going to lie, this may be the easiest bet of the year