AI TCM tech vs Human TCM Practitioner

https://www.instagram.com/reel/DWZdeCKjnox/

Hey everyone - quick one!

Over the past few months I've been running OpenClaw as my daily AI agent and sharing bits of the journey here. Some you have been asking how to actually set it up - what to install, what to skip, Mac vs Cloud, etc.

So a fellow friend/community member Nat (first AIde) and I put together a small hands-on workshop to walk people through it properly.

Think of it as a pilot - 15 people, 2 hours, everyone walks out with a working setup.

We're testing the format and keeping the group small so we can actually help everyone in the room. If it goes well, we'll figure out what's next.

If you're interested: https://luma.com/w8aj6w2e

No pressure either way - the channel stays free and I'll keep sharing what I learn here regardless.

Big move from Openclaw.

Recently, launched a feature called Claude Code Channels, which lets you chat with their AI directly in just three apps: Telegram, Discord, and iMessage.

But the team behind OpenClaw quickly updated it to work with an open standard (called MCP) that Anthropic itself created and shared freely. This means developers can now easily plug Claude's smart "brain" into OpenClaw's much bigger network of messaging apps.

For OpenClaw, this is a big win. It turns a potential competitor's limitation into an advantage: people who like Claude can now use it across way more apps without being stuck with Anthropic's smaller (and more sandboxed) setup.

It boosts OpenClaw's usefulness and popularity, shows the power of open-source (where no one company can fully control things), and keeps the project relevant even as big AI companies try to catch up.

TLDR: it makes OpenClaw even more flexible and hard to replace.

https://x.com/aakashgupta/status/2038504505214669072?s=46

https://cancelhack.co/

Cool vibe coded app I came across that helps track app subscriptions that give you discounts if you fake cancel

I guess it’s true then - was using opus for some marketing related queries for the wife’s business and it exhausted REALLY fast.

Moving forward, I think it’s good to be model-agnostic. Personally I’m sticking to monthly plans, so I can rotate out as and when needed. Think of it like musical chairs 🪑

As compute gets even more expensive, we will probably see free tiers used by the masses (backdated legacy models), paid plans by Claude/openAI starting at $100-200 a month, and pro plans hitting $500-$2000 a month in the coming days.

Claude’s upcoming Mythos could even be launched at $2000 a month - first movers enterprises will willingly pay just to get a head start.

Take advantage of the arbitrage while you can 😊

Be careful swearing at your AI 🤣🤣

Anthropic has been silently tracking how often you rage at it

Been using this for a few days and it’s been a game changer. Wish I knew about this earlier. Also got my claw to rewrite this for codex use

Please install it and let me know your reviews!

- forces a structure workflow for Claude
- default to brainstorming mode first
- runs parallel subagents for your various tasks
- overall way more efficient than the defaults.

🔗 https://github.com/obra/superpowers

Openclaw with GPT5.4.

Big upgrade from Kimi 😬

So… GLM’s new model 5V is claiming higher scores than Opus on several of these "see my screen and write the code" challenges, such as turning a design into a working app, answering questions about images, or navigating real phone and web interfaces.

Maybe time to give it a shot?

https://x.com/zai_org/status/2039371126984360085?s=46

What on earth is this Frankenstein? 💀

https://x.com/__tinygrad__/status/2039213719155310736?s=46

Really big news for those who have been using Claude for Openclaw: Claude have decided that they no longer will allow Claude Subscriptions for Openclaw

• Effective Date: April 4 at 12pm PT / April 5 3AM SGT

• The Pivot: Third-party harnesses (OpenClaw, etc.) no longer draw from sub limits; they now require separately billed "extra usage" credits.

• The "Softener": A one-time credit equal to your monthly sub price (must redeem by April 17) and bundle discounts up to 30%.

• The Exit: If this is a dealbreaker for your workflow, Anthropic is offering a full subscription refund via an email sent out tomorrow.

https://x.com/bcherny/status/2040206440556826908?s=46

Remember to claim your $26/$260 in credits by 17 April if you decide to not get a refund from claude (settings -> usage on the app)

Native video gen coming to your Lobster soon!

https://x.com/steipete/status/2040928953653744003?s=46

https://x.com/andrewfarah/status/2040535592925372895?s=46

Try this out if you are using X:

—

Some ideas:

1. Based on all the articles, news, and links I’ve bookmarked on X, how can they help with my current work or personal tasks? For example, I’m working on [briefly mention your main task, e.g. preparing a report for my boss / planning my child’s tuition / researching resale HDB flats / improving my sales targets]. Suggest 3–5 practical ideas.

2. Looking at the topics and links I’ve saved on X, what 3–5 books, would you recommend I read or learn next? Focus on things that can help me in my daily life in Singapore, like career growth, parenting, health, property, or managing stress.

3. I’ve bookmarked many articles and studies about biohacking, price predictions, etc. Based on what I saved, can you summarise what actually happened with those studies or findings later?

- I often bookmark articles about [your common topics, e.g. CPF planning, food delivery business ideas, fitness for office workers, or raising kids] What patterns do you see in my interests, and what should I focus on or learn next?

- Summarise the key takeaways from all the property / investment / parenting / career articles I tend to save on X. Give me simple action steps suitable for someone living in Singapore

Some observations over the past 2 weeks:

Multi-agent setups with OpenClaw & Hermes

• OpenClaw handles execution, while Hermes performs QC, checking for errors before OpenClaw drifts and suggesting improvements.

It’s essentially a self-improving loop.

Personally I’ve been using GPT-5.4 to QC Opus’s code 😂
Different models pick up on different issues.

Mostly experimenting for now. I’m not sure if it’s worth the extra compute unless it’s for a project that justifies it.