Zenki

Just a human searching for zen

[Piotr Pomorski @PtrPomorski]

Someone put Fable 5 on the pirate bay, 3.4TB 😂

[Trung Phan @TrungTPhan]

an all-time bet on yourself

[Tiffany Fong @TiffanyFong]

the new world order

[Zenki @havemorezen]

Last night I pasted Anthropic's new Fable 5 prompting guide into Claude Code and asked one thing:

"Audit my CLAUDE.md and skill files against this. Flag anything that now hurts more than it helps."

It flagged three of my own rules:

a skills plugin that screams "YOU MUST invoke this" before every task - the guide now calls that exact pattern out as harmful. Disabled it.
my "always plan before acting" rule - newer models over-plan when you force it. Softened.
"read the full file before every edit" - made it conditional.

I wrote all three to babysit older models. They were quietly dragging down the new one.

Anthropic says this in the guide itself: old prescriptive instructions can make the model worse. Nobody reads that part because everyone's hunting for new tricks to add, not old rules to delete.

So this is my ritual now. New model drops → paste the official guide into the agent → let it audit its own leash.

Took 10 minutes. Deleted more than I added.

Try it.

https://t.co/A7KdCSDyi6

[Claude @claudeai]

Fable 5 is state-of-the-art on nearly all tested benchmarks, with exceptional performance in software engineering, knowledge work, scientific research, and vision.

The longer and more complex the task, the larger Fable 5’s lead over our other models.

[Teknium 🪽 @Teknium]

New Era of Computing

[Andrej Karpathy @karpathy]

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

[🍜 @cprkrn]

HOLY FUCKING SHIT OMG CLAUDE JUST CRACKED THIS SHIT, THANK YOU @AnthropicAI THANK YOU @DarioAmodei NAMING MY KID AFTER YOU 😍

https://t.co/gObNirRDpS

[Miles Deutscher @milesdeutscher]

TradingView x Claude Code is insane for vibe-coding custom trading indicators.

What used to take hours of coding and expert technical knowledge is now just one Claude prompt away.

Gaining a market edge has never been easier.

This is free sauce - FULL cheatsheet:

[Kakashi @kkashi_yt]

Someone sniped $Hanta at $9K MC, turned $70 into over $40,000 in profit which sold $18,100 so far.

[Zenki @havemorezen]

NVIDIA NIM is the most underrated thing in AI right now.

Why don’t people talk about it?

Free inference for models like Nemotron or MiniMax 2.7.
No API key costs. No additional local compute needed - can run on any laptop or desktop. Just call the endpoint.

Setup takes like 2 min: go to NVIDIA NIM → search for the model → generate an API key → plug it into OpenClaw or Hermes agent. It’s that quick and simple.

I run a full support agent on it - handles background tasks, monitoring, health checks, and any light research. It's not as smart as Opus 4.7 or GPT 5.4 but for structured, repeatable work where you don’t need the most intelligent model? It's more than enough. Most importantly, it's comepletely FREE!

$0/month for a 24/7 agent. That's hard to beat.

[Zenki @havemorezen]

Beats Opus 4.7?

[OpenAI @OpenAI]

Introducing GPT-5.5

A new class of intelligence for real work and powering agents, built to understand complex goals, use tools, check its work, and carry more tasks through to completion. It marks a new way of getting computer work done.

Now available in ChatGPT and Codex.

[Taiki Maeda @TaikiMaeda2]

This is the first time since MSTR's inception that they were able to buy billions of BTC at an mNAV below 2.

mNAV expansion gives Saylor more firepower to raise cash, either to pay future STRC dividends or buy BTC.

This is bullish as it introduces positive reflexivity on BTC.

[Brendan Falk @BrendanFalk]

To check if your Google Workspace has been compromised by the same tool that compromised Vercel:

1. Go to https://t.co/TpuIOW5Fwg
- This is Google Admin Console > Security > Access and Data Control > API Controls > Manage app access > Accessed Apps
2. Filter by ID = https://t.co/uqJnCqp5Ah
- This is the ID of the compromised OAuth app

If you see an app after filtering, you have potentially been compromised

[Claude @claudeai]

We've redesigned Claude Code on desktop.

You can now run multiple Claude sessions side by side from one window, with a new sidebar to manage them all.

[Jorge Castillo @JorgeCastilloPr]

Me and Claude working together

[klöss @kloss_xyz]

somewhere out there rn Sam Altman is screaming at the top of his lungs and attempting to track this mfer’s IP address

just so he can nuke him off the internet

protect Husk at all costs

[Zenki @havemorezen]

This is actually really nice. I have been wanting a Pro plan tier at around the $100 mark and I'm glad they decided to introduce this. Great timing and decision as many switch their OpenClaw models from Opus 4.6 to GPT 5.4

[OpenAI @OpenAI]

We’re updating our ChatGPT Pro and Plus subscriptions to better support the growing use of Codex.

We’re introducing a new $100/month Pro tier. This new tier offers 5x more Codex usage than Plus and is best for longer, high-effort Codex sessions.

In ChatGPT, this new Pro tier still offers access to all Pro features, including the exclusive Pro model and unlimited access to Instant and Thinking models.

To celebrate the launch, we’re increasing Codex usage for a limited time through May 31st so that Pro $100 subscribers get up to 10x usage of ChatGPT Plus on Codex to build your most ambitious ideas.