miJ remarC

fan of Jim Cramer

[TheBigBerbowski @TheBigBerbowski]

I haven't been this excited about a company since I invested in SK Hynix.

I hope you enjoy the read.

$CRDO

[ericosiu @ericosiu]

In twelve months, EVERY company will be running a Company Brain.

The teams who build it this year will spend the next year compounding. Everyone else is going to play catch up.

Here's what it actually is. You connect your Slack, your GitHub, HubSpot, all your tools into one intelligence layer, then build the org chart around it: a main brain up top, a fleet commander running the agent fleet, specialist sub-agents handling execution.

The reason it works is change management basically disappears. Your team already lives in Slack. You're just adding agents to the room they're already in.

You NEED to start building yours now. In a year this will stop being an advantage and will become table stakes.

[Bill Gurley @bgurley]

When people talk about AI chips in China, they focus only on Huawei. Just as the US has Cerebras, Google TPU, Amazon Tanium, & AMD, there is a far bigger ecosystem of players, including Cambricon with a US$40B market cap. Many public already. More here:

https://t.co/IogI68b6DX

[Chamath Palihapitiya @chamath]

https://t.co/MCXWZTVvte

[Codez @0xCodez]

Anthropic Managed Agents team:

"Fable 5 is our best model for running self-improving agent systems.

Add /loops, dynamic workflows, dreaming and you are unstoppable"

in 13-minutes, Anthropic team shows how to build self-improving agent systems with Fable 5 from scratch.

Worth more than a $500 agent building course.

Live from the last Anthropic stage in Japan. Unpublished.

[rody @0x_rody]

At Anthropic's event, Metaview engineer:

"We stopped fixing our prompts. The system reviews its own output and rewrites its own instructions now."

In 16 minutes, he shows the Claude Code loop running in production on thousands of reviews, not in a demo.

Watch the talk, then grab the full loop setup below👇

[Dario Amodei @DarioAmodei]

Today I'm publishing a new essay, Policy on the AI Exponential. AI is progressing extremely fast—much faster than the policy process was built to handle. The essay lays out where I think the technology is now, and the action needed to close the gap: https://t.co/Lh6PWae178

[Jukan @jukan05]

SemiAnalysis occasionally publishes non-semiconductor analysis, and I really enjoyed reading this one. I know almost nothing about robotics, but I found this kind of content really fascinating.

[Aswath Damodaran @AswathDamodaran]

I valued SpaceX for its IPO a few weeks ago, with minimal information and a promise to revisit the valuation, when the prospectus was made public. The prospectus is public, the offering price has been set and my update is up and running. https://t.co/zRjpD1C0wv

[Anatoli Kopadze @AnatoliKopadze]

Anthropic engineer:

"You can build 5 assistants in one afternoon. Each one handles a task you've been doing manually every single day"

in 45 minutes he shows exactly how to do it from scratch, step by step

most people are still doing this manually

watch the session, then save the guide below

[Shubham Saboo @Saboo_Shubham_]

https://t.co/w7v8Vq1JNN

[dylan ツ @demian_ai]

The next AI factory may run out of ways to talk to itself before it runs out of GPUs.

This app is full of « photonics this, CPO that, $POET, $SIVE and the gang ».

So I wrote a piece that actually maps why this bottleneck feels different, where it bites first, and how the thesis shifts from chips/HBM/power to the back-end fabric.

You should stop picturing a data center as a room full of chips and start picturing it as a communication system.

Every GPU in a training cluster is constantly asking other GPUs what they know. Gradients, activations, experts, cache state, routing decisions, all of it has to move. At small scale, copper can carry the conversation. At AI factory scale, copper starts becoming a wall.

The signal has to cross board traces, connectors, cables, heat, loss, reflection, crosstalk, and distance. The fix is always more power, more retimers, more cooling, more cable bulk.

At some point the system is spending too much energy keeping electricity alive long enough to become useful.

That is why photonics matters.

Most coverage still frames this as “optics TAM growing.” It misses the supply chain relocation and the new constraints on light sources, integration, and test that actually determine who scales.

The names positioned at the light source + photonic integration layer for next-gen ELS/CPO?

$POET’s Optical Interposer and $SIVE’s high-power DFB laser arrays are right in the critical path, but there are many many others.

AI’s photonics bottleneck: why moving data is now harder than making chips:

https://t.co/OguHp3QzFt

[Jukan @jukan05]

China already has a NAND market share comparable to Micron and SanDisk....

[zerohedge @zerohedge]

Snubbed By SpaceX, Jefferies Now Helping Traders Short World's Biggest IPO https://t.co/qxRgBHGQ8m

[Freda Duan @FredaDuan]

IPO Supply vs. Demand for @SpaceX

Supply / lockup schedule -

Potential 9% unlock on the second trading day after 2Q26 earnings. That is roughly 2x the IPO float.

Demand / index buying schedule -

T+5: passive/index buying could equal ~7–10% of the float.
Total T+5 to T+15: passive/index demand could equal ~17–25% of the float.

Analysis (done by AI) -

Day 0–15: Thin float + passive buyers + price-insensitive demand likely create a sharp supply-demand squeeze.
Day 15–70: Air pocket. Index buying is mostly done.
Day 70–180: Digestion phase. More shares unlock, but a higher float may also force additional index buying.
Day 366: The 51% unlock is the biggest overhang. Actual selling could be much smaller if Musk does not sell.
Upside kicker: S&P 500 inclusion. If SpaceX becomes eligible after the float expands, S&P inclusion could create a second wave of passive buying.

[The Wall Street Journal @WSJ]

Exclusive: AI market-research platform AlphaSense is being valued at $7.5 billion in a fresh round of funding https://t.co/XGzb0EaEt5

[Nainsi Dwivedi @NainsiDwiv50980]

Andrej Karpathy just explained the future of software engineering without directly saying it.

The best AI engineers are no longer “prompting.”

They’re building systems around the agents.

Karpathy’s biggest insight wasn’t:
“Claude can code.”

It was:
LLMs become dramatically better when you force them into disciplined workflows.

That’s why "CLAUDE.md" files are suddenly everywhere.

Not because they’re prompts.

Because they behave like an operating system for the agent.

Karpathy called out the exact problems with AI coding:

- models assume instead of asking
- they overengineer simple tasks
- they hide confusion
- they rewrite unrelated code
- they optimize for completion, not correctness

So developers started encoding rules directly into the workflow:

→ Think before coding
→ Simplicity first
→ Surgical edits only
→ Goal-driven execution

And the results are wild.

People are now running multiple Claude Code agents in parallel like engineering teams:

• one agent researching
• one debugging
• one writing tests
• one optimizing code
• one validating outputs

Not “AI assistance.”

Actual orchestration.

And this part from Karpathy changes everything:

“Don’t tell the model what to do. Give it success criteria and let it loop.”

That is the shift.

From:
“write this function”

To:
“here’s the goal, constraints, tests, and verification system — now iterate until correct.”

The craziest part?

This already feels like a phase shift in engineering.

A lot of developers quietly went from:
80% manual coding → to 80% agent-driven coding in just months.

Not because AI became perfect.

Because the leverage became impossible to ignore.

We’re entering an era where the highest leverage engineers won’t necessarily be the best coders.

They’ll be the people who build the best systems around AI agents.

[Photon Capital @PhotonCap]

Everyone draws CPO as one optical supply chain. InP, laser, optical engine. That picture can't tell you when CPO actually ships.

CPO is really three coupled budgets: energy (pJ/bit), heat (junction temp), reach (propagation plus coupling loss). Break one and the rest fall.

The schedule isn't held by InP. It's held by power and heat. Rack power went 40kW to 600kW in three years, and 800V HVDC plus liquid cooling have to open before optics bolts on. Retimed pluggable runs 10 to 20 pJ/bit; CPO targets around 5. NVIDIA Spectrum-X Photonics SN6800 hits 409.6 Tb/s, shipping 2H 2026.

Copper isn't dying from bandwidth. It's dying from energy. And it earns the most right before it dies. The real question isn't whether copper dies, it's when the cliff hits.

[Sac @Saccc_c]

这哥们靠vibe coding直播公开build产品，月收入超2万美金

最大成本花在了token上，3个Claude Max+1个200刀Codex。每天同时开六七个终端在youtube直播使劲造

当初刷到以为是整活的，没想到哥们玩真的😢

[Mr. Buzzoni @polydao]

THIS 2-HOUR STANFORD LECTURE WILL SHOW YOU HOW CLAUDE OPUS 4.8 ACTUALLY WORKS

> pretraining. SFT. RLHF. tokenization. evaluation

> the same pipeline that powers Claude Opus 4.8 - explained from first principles by the people who research it

now read the article to see how to use those skills in practice👇