Snarkitect 🇺🇸

FINALLY!!!🥳🥳🥳 🚨@leopoldasch Situtational Awareness 13F just dropped!!!🚨 Here are his latest positions🧵

A framework to understand how value accrues across the AI stack. This is a blueprint for understanding what builds AI into its pragmatic parts: what each layer is, where it ends, and where value is accrued. So here’s how you can think about it: 1. Layer 1 - Infrastructure Before any AI model trains or any robot moves, an industrial foundation must exist. Land, energy grids, cooling systems, critical minerals, and fabrication facilities. Infrastructure is the constraint that all the other layers depend on. 2. Layer 2 - Chips Transistors that are etched onto silicon wafers using extreme ultraviolet light. This is what allows both physical and digital AI to take an input, process it, and return a predictive output. The more transistors that fit on a chip, the more computation it can perform. 3. Layer 3 - Data Both digital and physical models train on data. Digital models train on text, code, and images; physical models train on gravity, friction, depth, and sensor streams. The more accurate the data, the more accurate the output. 4. Layer 4 - Models A model is a system that learns from examples. Feed it enough examples of inputs paired with correct outputs, and it adjusts its internal structure until it can predict correct outputs on inputs it has never seen before. LLMs represent a specific class trained on text. They learn by processing billions of examples of human language, developing the ability to write, reason, summarize, and generate code. 5. Layer 5 - Execution This is what lets models take actions on behalf of users. The execution layer lets models pursue objectives through sequential action: observing the environment, reasoning about the next step, acting, and looping until the goal is reached. 6. Layer 6 - Application All of the AI Stack’s revenue originates at the application layer, then goes to the layers below. Every dollar paid for AI is paid for an outcome, a task completed, and an answer delivered. Nobody wants H100s for their own sake. They want H100s because someone, somewhere, wants to run an application. These are the different layers that make up the entire ecosystem of AI. We did a full study on the AI stack. If you want to read about it, head over to my Substack (chamath.substack.com/p/the-ai-stack)

Tomorrow morning’s onboarding packet for the new Fed Chair: Welcome aboard, Kevin. We’ve prepared a short orientation checklist for your first day. 😊 Staff ☑️ Reassure global bond markets without printing more money ☑️ Stop inflation without causing 1932 ☑️ Lower rates without collapsing the dollar ☑️ Keep rates high enough so boomers don’t riot over savings yields ☑️ Keep rates low enough so millennials can someday buy a garage with a sink in it ☑️ Explain to Congress that deficits DO matter while Congress actively creates more deficits in real time ☑️ Convince Americans the economy is “strong” while eggs require financing ☑️ Prevent regional banks from spontaneously turning into vapor ☑️ Quietly monitor commercial real estate, private credit, derivatives, sovereign debt, and whatever fresh hell is happening in Japan tonight ☑️ Smile confidently on CNBC while your internal monologue sounds like a man trapped in a submarine hearing new creaks Then around 2:37am: Bond market: “Hey Kevin… you awake?” Kevin: “No.” Bond market: “Cool. Anyway the 30-year yield is doing parkour again.” Also before heading home: • buy kitty litter • check smoke detectors • maybe pick up candles • and whatever you do… DO NOT look directly at the Treasury debt clock after midnight. May God be with you, Staff

$NRGV just cleared the 52week high and now at its highest level since 2022🎯 market still sleeping on this one for whatever reason Same setup as 2025 $EOSE, heading for double digits ✍️

$NRGV failed head & shoulders pushing higher ✍️ Just a matter of time before it puts in its own 20-30% day

@andrewdfeldman @Nasdaq Onwards and Upwards! Congratulations Cerebras fam!

Ringing the @Nasdaq bell is just the beginning.

The 5 levels of option trading: Level 1: Buying weeklies Win rate: 20% Stress: 10/10 Status: Broke Level 2: Buying monthlies Win rate: 30% Stress: 9/10 Status: Still broke Level 3: Selling puts Win rate: 80% Stress: 4/10 Status: Consistent income Level 4: The Wheel Strategy Win rate: 90% Stress: 2/10 Status: Replacing income Level 5: Selling + LEAPs combo Win rate: 90% Profit Boost: 3X Status: Financial freedom I spent 3 years stuck at Level 1-2. Now I'm Level 5. What level are you?

I noticed something when looking over charts this weekend.... Right now $QQQ is sitting at a daily RSI of 82.83 and is 14.50% above its 50 day simple moving average. I went back to see if there some similar setups and examples to this, and what the results were after. Here were some of the most recent. 1. June-July of 2024 before the "Yen Carry Trade" and the market dropped 16% in 17 days before pushing higher. 2. November-December of 2023 when Yellen basically put QE back on the table after the sweep of the summer. Market then pulled back 5% before resuming higher. 3. September of 2020- after the massive post COVID rally Q was almost 20% above it's 50 day average before experiencing a 14% pullback before pushing higher. The big things I was looking for was a extended rsi (Like 80 plus on the daily), and a large percentage away from the 50 day (blue line). History would say this is a time for concern, not euphoria. Yes, I've always said I believe the market resolves higher, but piling in here is a recipe for disaster. Not necessarily on the long term, but the short term. I think there will be a very buyable dip, but it's not here yet. I think most likely is a dip, then a new high with bearish rsi, followed by a much larger dip. I don't care what the narrative is, only what the chart shows me.

THINK BIGGER

$SPY Price indeed got to 732 and went to 734.59. 736-740should be the upcoming resistance zone, if price breaks out of this resistance zone too it’s going to be a crazier rally possibly to 785. That’s the next level. Crazy. A rejection at 736-740 is the most likely however because the market needs to cool down a bit before continuing up. I’ll take a swing short whenever the market hits the 736-740 zone and my stop loss will be a close above it. I’d be looking for a retrace to the 20 dma to go long, or at the very least a retest and hold above the 5 & 10. Have a good evening ✌️

THEMES THEMES THEMES. - Semiconductors still leading! $LRCX $NVDA $AMD - Reformed crypto miners turned data center names $HUT $RIOT $CLSK - Neo clouds/ Software Infrastructure $DOCN $CORZ $NBIS $CRWV $APLD - Rare earth and Nuclear $MP $CCJ $USAR $CRML $UAMY $UUUU - Aerospace and Defense $UFO $FTAI $YSS $RKLB $PL This market is on fire!

@Zac_Markovich $CGEH

Doing market research.

The president of the United States posting from an official government account an image of the chair of the Federal Reserve being thrown in the trash. There are no words or precedents for this.

@Han_Akamatsu Ah the $TLRY chart....

Found a great setup here. Escaping this year long falling wedge + now making a bullish crossover on macro 1W HMA timeframe. 👀 200 likes and I’m dropping this full analysis ✍🏻

@colin_gladman Worst example of a CEO

Talk about sucking all the air out of the room Cohen. A lot more dilution coming soon. $GME

@sovereignsense Thank you Sensei!

@sovereignsense yaaas! where can I try it out?

$SIVE – If you are wondering if it is too late to invest, it isn’t and here is why. Goldman Sachs just unveiled their Optical Communications industry map for the AI infrastructure buildout. The full ecosystem laid out – IC Design, Materials, Optical Components, Packaging and Testing. Guess what? $SIVE is not on the map, yet. The highlights from Goldman: TAM growing from $15B to $154B. Scale-up optical interconnects within racks and super-nodes are the real driver. NVIDIA’s GPU clusters jumping from 72 to 576 GPUs by 2027-2028, with interconnect dollar value per computing unit going from $315K to $9.4B. A 29x increase per unit. Silicon photonics penetration jumping from 6% in Q1 2024 to 46% by Q4 2028. CPO commercialization starting in 2026. The shift from copper to optical is structural and accelerating. In the laser/light source segment, Goldman maps $LITE, $COHR, $AVGO, $AAOI – the obvious ones. Sivers is nowhere to be seen. $SIVE powers Jabil’s 1.6T LRO architecture. Win Semi is their confirmed foundry partner. They sit upstream of the entire optical interconnect stack with InP lasers everyone needs. When Goldman publishes the next iteration of this map, $SIVE will most likely be listed on the US Nasdaq and big enough for institutional coverage. Which means institutions will try to shake the ground just so they can catch up. This report is extremely bullish for the entire photonics ecosystem. Being early pays. Having conviction pays even more.