Danson Ang

Copper trader in international mining company

[The Kobeissi Letter @KobeissiLetter]

Are gold prices poised for more upside?

Based on the historical relationship between gold and global M2 money supply, gold is valued at $5,025 per ounce today, according to Fidelity analysis.

This implies +9% upside from the current gold price of ~$4,600.

This comes as global M2 money supply grew ~8.5% YoY this month, an increase from the ~7.0% YoY growth seen in June.

Furthermore, global gold ETFs have attracted +$18.9 billion in inflows over the last 12 months, the largest trailing 12-month inflow since February.

This also marks an improvement from the +$12.0 billion in trailing 12-month inflows recorded in June.

Gold markets are heating up.

[Сarm1ne @carm1nee]

Stan Druckenmiller told Scott Bessent exactly what it would take to stop Washington from spending: "the clowns in Washington - unless they get a signal from the bond market, they're just going to keep spending"

this is him explaining the arithmetic nobody runs on corporate America, why he went from 93% invested to flat over a single Trump tweet, and what he says the Fed has actually built over the last ten years

"when the Trump tweet went out I went from 93% invested to net flat. not because I'm trying to make money - I just don't want to play in this environment"

"corporate debt went from $6 trillion to $10 trillion. profits went from $1.7 to $2.2 trillion. and the interest cost on that extra $4 trillion only went up 23%. you'd think profits would explode with that formula. they went up 29% - over eight years"

"we have all these zombies walking around. the most innovative period since the late 1800s, and you're hardly seeing bankruptcies - because there have been no market signals from the Fed"

"if I were trying to create a deflationary bust, I would do exactly what the world central banks have been doing"

bookmark & watch the full conversation - then read the article below ↓

[Lunar @LunarResearcher]

Andrej Karpathy just released a 1-hour Stanford lecture on AI engineering from scratch:

"You can actually delete everything… delete everything, keep Graph"

Here are the core ideas:

10% → LLM: "I treat GPT like a general-purpose computer that can be reprogrammed at runtime"

30% → Prompt: "I give that computer a program written in natural language"

50% → Agent: "I wrap the model with a goal, context, memory, and tools that turn prediction into action"

70% → Loop: "I separate the inner loop, where the model learns from context, from the outer loop, where training updates the weights"

100% → Graph: "I organize communication as data-dependent message passing across directed graphs"

This 1-hour Stanford lecture will teach you more about AI engineering than 100 random YouTube tutorials

"Just chop everything up and throw it into the mix"

Bookmark it and watch later

Then read the full article below ↓

[CyrilXBT @cyrilXBT]

this is absolutely unreal

jack dorsey (twitter co-founder) just shipped a totally free github repo sitting at 26.2k stars that basically hands you an ai-agent operating system for running a business

here’s the setup:

1. clone the repo
2. host the server yourself: channels, search, git, automations, it all runs there
3. drop your agent into a channel like it’s another teammate, lock down its permissions, and let the crew guide it in real time

save this and bookmark it, seriously

[Kirill @kirillk_web3]

A Chinese developer just explained the shift from Loop Engineering to Graph Engineering better than anyone.

most people are still building agents the way that's about to be obsolete.

> why single-agent loops break and go "goal blind"
> the 4 parts of a graph: nodes, edges, state, policy
> 3 topologies that run everything: diamond, supervisor, pipeline
> Anthropic's 5 official workflow patterns

the punchline: it's not how many agents you run. it's the determinism you build with verifiers, code fallbacks, and reality anchors.

I broke the same architecture down with Kimi K3. Full A-Z guide below.

[Сarm1ne @carm1nee]

Carl Icahn to Larry Fink on why BlackRock is an extremely dangerous company: "they sell the concept of liquidity - there is no liquidity. That's my point, and that's what's going to blow this up"

this is him explaining why he says there will be nobody left to buy when the run starts, why he calls Fink's letters to 800 CEOs a sales pitch, and what Fink says back when he's had enough

"everybody's in this party mobile having a drink, having fun. and you know who's pushing that thing? Larry Fink and Janet Yellen. they're heading for a cliff - and you know what's going to destroy it? They're going to hit a black rock"

Fink cut in three times - "I don't think that's fair" - Icahn: "I don't care if you think it's fair"

"when there's a run, there is nobody to buy that stuff. There is no one out there to buy that trillion bucks"

Fink: "you're a good investor, but you're wrong. You're just dead wrong - I'd be happy to spend time with you over lunch, I'll pay, and teach you about ETFs"

bookmark & watch the full conversation - then read the article below ↓

[Ajay @ajay4ai]

Anthropic engineer:

“90% of our engineers were using self‑improving loops. Now everyone shifted to building agentic Graphs"

“No more prompting.”

In just 10 minutes, she builds her entire Claude Code setup and workflow live from a blank terminal.

This is more valuable than most $1000 agentic courses.

Watch this video, Then save the article below if you want to become a graph architect before everyone else catches up.

[SightBringer @_The_Prophet__]

⚡️Japan is reaching the end of the postwar bargain that allowed an aging society to promise itself stability without ever fully paying the market price for that stability.

The deepest structure is generational extraction through time.

Japan spent decades suppressing the return on savings, suppressing the cost of government borrowing, suppressing the clearing price of sovereign duration, and using the BOJ to prevent the financial system from discovering what all of those promises actually cost.

That bought social peace. It preserved pensions, banks, insurers, government solvency, employment, and asset values long enough for the population to age inside the protection.

Now the people who were supposed to carry those promises forward are fewer, the debt is vastly larger, the financial institutions are stuffed with assets created under the old rate regime, and inflation has returned as the force that makes permanent suppression impossible.

The country has reached the point where every escape route takes wealth from somebody.

Let long yields rise and the old bond books bleed.

Push yields back down and the yen absorbs the loss.

Defend the yen and liquidity tightens into an economy already carrying enormous duration exposure.

Compensate households for inflation and fiscal issuance grows.

Monetize enough of that issuance and the currency becomes the release valve again.

There is no clean exit because the losses already exist. Policy only decides their destination.

That is what the insurer losses are telling you. They are the visible edge of a national balance sheet that was built around the assumption that time could remain almost free forever.

Japan has more claims on future Japanese production than the future Japanese population can honor at the purchasing power those claims originally implied.

That is the insolvency.

It will never appear as a conventional bankruptcy because the state owes most of those claims in a currency it controls. The machine can always manufacture the units required for settlement.

So the default migrates into the unit itself.

A pension remains a pension.

An insurance policy remains an insurance policy.

A JGB still matures at par.

A bank deposit still shows the same number.

The holder quietly receives less real command over food, energy, housing, imported goods, labor, and foreign assets.

That is how the state resolves the contradiction without announcing defeat.

Japan is therefore entering a period where the government increasingly preserves nominal civilization by consuming real purchasing power.

That is the endgame.

And the life insurers are trapped directly in the middle because they sold promises about the future using assets priced by the old world. Higher yields eventually improve their economics, yet the road to those higher yields destroys the market value of what they already own. They need the new regime and are wounded by the transition into it.

That is why this can become reflexive.

As yields climb, old assets lose value. Capital buffers weaken. The incentive to rebalance rises. If policyholders ever begin moving money toward more attractive alternatives, cash needs increase. Asset sales then reinforce the very yield move creating the losses.

At that moment, the BOJ becomes the buyer of last resort again.

There is no other final buyer with an infinite balance sheet in yen.

That is where this ends.

[Kun Chen @kunchenguid]

i worked closely with Jalal over the past couple of years on agents - he built tons of cool harness features that in hindsight were months ahead of its time

definitely worth a follow if you want to see cool tips and tricks on agentic engineering from real builders

[Khairallah AL-Awady @eng_khairallah1]

this is f*cking gold

Andrej Karpathy joined Anthropic five weeks ago.

Two Anthropic seniors just made Karpathy's loop 1000x better with "Graph Engineering"

the agentic systems got 1000x better the moment you wired agents into a graph

I dropped it into my setup. The very first response was different.

Not slightly different. Completely different.

Claude stopped giving generic answers and started working exactly the way I think.

Bookmark it before it gets lost in your feed.

Read it now, then check the article below.

[Marcos Hernanz @MarcosHernanz]

After doing ~60B tokens, this is my full AGENTS.md

[Lunar @LunarResearcher]

Andrew Ng just released a free 1-hour course on building agentic knowledge graphs from scratch.

How to turn raw data into structured memory for AI agents:

00:00 - Understand agentic knowledge graphs
03:07 - Build your first agentic graph
14:00 - Architect multi-agent systems
23:00 - Build graphs with Google ADK
1:06:03 - See why graphs are the future of agents

Most people are still building agents with flat context.

Andrew Ng is already teaching the next layer:

Data → Relationships → Memory → Agents → Graphs

Loop engineering is the old workflow.

Graph engineering is the new one.

This 1-hour course is worth more than most paid agent engineering courses.

Bookmark and watch it before everyone catches up.

Then read the full Graph Engineering roadmap below

[Puneet Patwari @system_monarch]

I'm a Principal engineer & I passed system design rounds of Amazon, Atlassian, Walmart, Saleforce, and Deliveroo.

Trust me, learning system is not hard. Start from these fundamental concepts:

1) Load Balancing: https://t.co/3jKCLiI6vl

2) CDN: https://t.co/dxzCmm9gAf

3) Caching: https://t.co/pRgn0FTPp2

4) Cache Invalidation: https://t.co/QrfRjJ57gd

5) Rate Limiting: https://t.co/LE5ECM2tGt

6) API Gateway: https://t.co/DgU8cBDUVr

7) CAP Theorem: https://t.co/a8WydnAIxd

8) Sharding: https://t.co/XQLU6eDriD

9) Replication: https://t.co/KuDkFH0fjx

10) Partitioning: https://t.co/3WXKeZLbLa

11) Queues: https://t.co/JchEoCcFmF

12) Microservices: https://t.co/aAQfM6AWMq

13) Microservices Vs Monoliths: https://t.co/bTaIIWkPU3

14) Fault Tolerance: https://t.co/qXNBoyOqYT

15) Database Scaling: https://t.co/D2lvPm1wkB

16) Service Discovery: https://t.co/z2DpwbJBVI

17) Consistency models: https://t.co/K2r3nMcCQu

18) Eventual Consistency: https://t.co/SWiz4ckIKR

19) Distributed Transactions: https://t.co/xqL7BTJxXn

20) Leader Election: https://t.co/ApNaYSnSFj

21) Horizontal vs Vertical Scaling: https://t.co/IFuEmzMfob

22) Back of the Envelope Estimation: https://t.co/7ntEmtVggQ

23) Idempotency, Data Latency & Finale: https://t.co/fNArLx4MrW

Let me know what you'd like me to cover, would love to help :)

[Lunar @LunarResearcher]

IBM just released a free 1-hour course on building agentic knowledge graphs from scratch.

How to give agents structured memory and connect them into full systems:

00:00 - Understand knowledge graphs
05:35 - Build your first agentic graph
19:59 - Add graph-powered memory
30:39 - Orchestrate multiple agents with graphs

Most people are still stuffing documents into vector databases.

IBM is already teaching the full stack:

Data → Relationships → Memory → Graphs → Multi-Agent Systems

Flat retrieval is the old workflow.

Agentic knowledge graphs are the new one.

This 1-hour watch is worth more than most paid agent engineering courses.

Bookmark and watch it before everyone catches up.

Then read the full knowledge graph engineering roadmap below

[rari @0xwhrrari]

Google just dropped a free 2-hour course on complete agent engineering

How to turn one prompt into a system that keeps running while you sleep:

38:46 - Build your first AI agent
54:46 - Connect agents to MCP tools
1:12:43 - Run four different agent loops
1:20:57 - Turn those loops into graphs
2:22:31 - Build the complete autonomous system

Most people are still building one agent and stopping there

Google is already teaching the entire stack:

Agents → Tools → Loops → Graphs → Autonomous Systems

Single agents are the old workflow

Systems that keep running without you are the new one

This free course is worth more than most paid agent engineering bootcamps

Bookmark and watch it today

Then read the full graph engineering playbook below

[codila @0xCodila]

Google just released free 2-hour course on full Graph engineering: 1 prompt ��� 100 agents → loops → graphs from 0% to 100%:

10% → 17:44 - build your first agent
30% → 39:30 - Loop engineering: iterate, check, break
60% → 1:12:38 - Graph engineering
75% → 1:34:26 - agents that throttle themselves
100% → 1:55:05 - full graph for multi-agentic systems

everyone builds one agent and calls it done - this is the full system where agents wire themselves into a graph

watch the course, build the graph - then read the full architecture below ↓

[MIKE @mikenevermiss]

Ex-Google engineer built a graph-based workflow that made Claude Code do nearly 10 days of work in about an hour.

Here's the exact setup he uses.

step 1.

create a separate Git worktree for every task.

run:

git worktree add ../task-1 -b task-1

open a new terminal, cd into it, and launch Claude Code.

repeat this for every agent.

each one gets its own branch, so nothing conflicts.

step 2.

give every Claude agent a different task.

same repository.

different branches.

they all work in parallel.

step 3.

once they're done, review the pull requests.

keep the good ones.

merge what works.

step 4.

add a review agent.

paste each PR into Claude and ask:

"does this implementation match the original specification"

every agent now has another agent reviewing its work before it gets merged.

step 5.

scale it up.

48GB of RAM can comfortably run around 20 Claude Code agents.

128GB can handle roughly 50.

most engineers run Claude Code once and wait.

he runs dozens of Claude Code agents in parallel, then merges the best results.

this 60-minute lecture is one of the best free resources on graph engineering and multi-agent workflows.

bookmark it and watch it below.

[CyrilXBT @cyrilXBT]

THIS GUY MAKES COMPLEX AI AGENT CONCEPTS RIDICULOUSLY EASY TO UNDERSTAND.

No jargon wall. No assuming you already know what a reward model or an evaluation harness is.

Just a straight line from "I have no idea how agents actually work" to "oh, that's genuinely simple."

The best explainers do not simplify by cutting corners. They simplify by finding the one analogy that makes the whole thing click.

[codila @0xCodila]

Google just released free 1-hour course on building agentic knowledge Graphs from 0% to 100%:

10% → 4:01 - how to build a GraphRAG agent
30% → 15:00 - Graph Engineering explanation
55% → 30:00 - Agentic search Engineering
80% → 35:48 - Graph Engineering practice
100% → 47:06 - self-improving agents in graphs

this free Google course mass replaces a $500 graph engineering bootcamp - learn it in 60 min to 100%

watch it today - then read the full graph playbook in the article below ↓

[Loran @0xLoran]

Terence Tao, UCLA professor and the most decorated mathematician alive:

"Funds pay $750K to fuse a pile of weak signals into one genuine edge. I proved the exact fact that makes that possible and makes it lethal: stretch any sequence long enough and hidden structure has to appear. it always builds up. the entire job is separating the structure that is real from the noise that just wears its face."

this free lecture is the most awarded mathematician alive walking through the precise problem that sits under every factor model, and it costs you nothing.

at the chalkboard it is plain. Tao's life's work circles the border between structure and randomness. the Erdős discrepancy problem asks something that sounds trivial: can you write an endless run of plus-ones and minus-ones that stays perfectly even forever? Tao proved you cannot. however cleverly you arrange it, imbalance, hidden structure, is forced to pile up as the string grows. a long stream of pure, structureless noise simply does not exist. that is the whole idea, stripped of the jargon.

which is exactly why a multi-factor model can print, and exactly why it can bury you. stack enough weak signals and real structure will surface, because at scale structure cannot not appear. but fake structure surfaces too, patterns that exist only because the data is long enough to manufacture them. same point as the post above: finding structure is guaranteed. knowing which structure is an actual edge is the rare, expensive part.