Michael Liu

Co-Founder of 10+ Years in Web3 Infra, AI, Energy, Cybersecurity. SFBA/HK. Views are my own.

[Michael Liu @maoconsult511]

@aleabitoreddit Computing power, storage, and motors.

[Michael Liu @maoconsult511]

@star_okx Great work! We built similar one as Fabric.

[Michael Liu @maoconsult511]

@StanfordHAI A traceable, relatively decentralized, and self iterating trust layer for AI is an urgency.

[Michael Liu @maoconsult511]

@aleabitoreddit Personally considering AGI Bot

[Michael Liu @maoconsult511]

@theRealKiyosaki 2027-29 crisis and til 2031-33 correction over.

[Michael Liu @maoconsult511]

2027-29 crisis and til 2031-33 correction over.

[Robert Kiyosaki @theRealKiyosaki]

BIGGEST BUBBLE BUST

I do not know what pin, what event will pop the biggest bubbles in histor. What ever the event, the pin is near.

It’s not IF. It’s WHEN.

When the bubbles go bust I predict gold will hit $35,000 an ounce one year after the gold bubble goes pop..

I predict silver to hit $200 an ounce a year after the bust.

I predict Bitcoin will hit $ 750,000 a coin a year after the crash.

And i predict Ethereum to be $95000 a year after crash.

What do you think prices will be a year after the next GFC?

Time to get richer?

[Michael Liu @maoconsult511]

@aleabitoreddit Agree, but follow up with a huge dump (or say correction) soon

[Michael Liu @maoconsult511]

Agree, but follow up with a huge dump (or say correction) soon

[Serenity @aleabitoreddit]

Robotics is next.

Both deal count and investment amounts are skyrocketing per pitchbook March data (source: a16z)

Good thing is: the same AI DC exposure often has cross-exposure to humanoid ramp.

Like DRAM/NAND with memory (on humanoid inference/storage) or DFB lasers with photonics (FMCW LiDAR vision/sensing).

Right now most exposure is upstream component parts… or programs within large players like $AMZN or $TSLA.

So global IPO season H2 into 2027 for pure play humanoids/robotics companies is going to be fun.

[Luffa @LuffaApp]

To continue accessing your wallet, securely import your 12-word recovery phrase into a trusted wallet that supports the relevant blockchain and compatible derivation path. For assets on the Endless network, use an Endless-compatible wallet. The Luffa Team

[Michael Liu @maoconsult511]

@TheMoonCarl Not the bottom, wait until q4

[Michael Liu @maoconsult511]

@ibubblyyou Same here! It was such a sudden...feel free to reach out via X, wish you all the best!

[Michael Liu @maoconsult511]

To everyone who believed, built, and walked this path with us — thank you. Your trust, energy, and vision gave this journey meaning far beyond what we could have imagined.

We are profoundly grateful for the community, the conversations, and the dream we shared.

The chapter closes, but the gratitude remains eternal.

With heartfelt appreciation,
Michael, CEO @LuffaApp

[Luffa @LuffaApp]

To our community,
After much thought, we've made the difficult decision to wind down Luffa.
Thank you to everyone who believed in us and helped shape this journey. Your trust meant everything.
Grateful for being part of the Luffa story.
The Luffa Team

[Michael Liu @maoconsult511]

@Gen_Hibee Thanks for your support! Godspeed!

[Michael Liu @maoconsult511]

@Gen_Hibee Thx bro! Will keep in touch! So sad we have to shut it down...

[Michael Liu @maoconsult511]

The Fabric MVP is live. Lowest-friction AI-native + crypto-native platform in the world.

Wallet → proposal → permission → execution → receipt + feedback/learning loop.

Agents & humans continuously improve behaviors.
Try: https://t.co/xWa9WdGCJh
#AgenticEconomy #AI #Agent

[Jialin @0xjialin]

Try the Luffa Fabric MVP here:

https://t.co/B3m1qH6WXz

Public testing needs no local setup. You can walk through:
wallet connection → proposal → permission check → wallet confirmation → execution receipt → feedback / learning.

If the backend is cold, wake it first:
https://t.co/KRuOVn9Eta

Use small amounts only for mainnet tests.

#Agent #agenticeconomy

[Michael Liu @maoconsult511]

Reflecting on US debt dynamics under Fed Chair Kevin Warsh: If sustainable deleveraging is a core policy priority, measures taken—whether by design or as an unintended consequence—could amplify rallies in US equities and crypto markets. A subsequent sharp correction might then contribute to fiscal adjustment through elevated revenues during the expansion phase or post-crisis recalibration.

This points to a structurally higher probability of intense boom-bust cycles ahead: periods of unrealistic exuberance followed by sobering disillusionment.

While Warsh emphasizes balance-sheet normalization and market discipline, the underlying debt overhang and growth imperatives suggest volatility will remain a defining feature rather than an anomaly. History provides repeated precedents.

Not advice. Open to differing views on the likely path.

[Michael Liu @maoconsult511]

I feel like a monkey with little knowledge about this…

[Marco @maarcoofdezz]

Dos ingenieros de Anthropic pasaron 24 minutos exponiendo cada función de Claude Code que no sabías que existía y gratis.

La mayoría de las personas pasarán de largo este contenido

[Michael Liu @maoconsult511]

AI Agents' most direct, high-ROI use cases must satisfy four criteria:

Fully contained execution on local or remote closed environments.
Complete world/model simulation.
Extremely high problem complexity with vast search spaces.
Hard to solve, yet easy to verify.

This is why I've emphasized for over a year: Programming first, followed by mathematics + Lean 4 verification.
Next: electronics, electrical, signals, mechanical, thermodynamics, RTL chip design, embedded systems, fluid dynamics—any domain where an AI Agent pairs with MATLAB toolboxes or Simulink for rapid simulation.

Each traditional engineering vertical + AI Agent + toolbox equals ~$100M value. They enable agents to drive sustained 10- 100-hour problem-solving loops on novel challenges.

The real agent economy begins in deep engineering.

[Michael Liu @maoconsult511]

Just discovered a killer workflow: Use GPT + DeepSeek V4 Pro as senior engineers to draft SOPs, CI/CD pipelines, and designs. Then throw it all at Hermes Agent's Minimax M3 for execution.

GPT finishes in ~1h, DeepSeek in ~2h. M3? Four hours in, still battling CI tests, logs flying, errors looping.

I'm not mad—I'm excited. M3's real value isn't output, it's validation. If even this "junior" agent can follow step-by-step successfully, the docs are truly clear. Stuck? It exposes hidden assumptions and unspoken expertise.

Old way: test manuals with new hires.
AI way: test with the dumbest agent.

In the agent era, the best QA engineer might be the least smart model. Like a new colleague who reliably finds every unclear spot.

M3 just became my favorite quality inspector. 🚀

[Michael Liu @maoconsult511]

The market has its own cycle, with inevitable pumps and dumps in different sectors.

[Serenity @aleabitoreddit]

IMO photonics theme + CW laser chokepoint is goated.

It's legit like markets have short term memory loss and forgot how $LITE went from $3B -> $65B+ from 2024 to now.

Because $NVDA caused EML bottlenecks, and forced architectural changes.

We're literally seeing the same thing today with CW lasers + 1.6T/CPO shifts with Nvidia signing LTAs everywhere.

Now, $AMD + other CSPs are hunting for remaining scraps with large LTAs for CW lasers + optical components.

GS Research's ~9-10x $154B optical TAM in 2028 and near $0 -> $91B CPO TAM in just 2 1/2 years.

Don't just magically disappear from a month of trading volatility.

$AAOI sitting at ~$13B, $SIVE sitting at ~$3B, and other CW laser players look strategically very valuable.

And next year I think we'll look back and say "Why didn't I learn my lesson the first time with EML from Nvidia and pick up CW laser adjacent names!"

Then there's likely gonna be some new mini trend 1-2 years from now like microled or quantum dot and we're gonna see the same thing repeat.

Think Sumitomo's projections with CW laser share + silicon photonics being majority / dominant architecture should be correct.

I'm personally just focusing on that bottleneck as you've seen with $SOI, $TSEM, $SIVE, and others.