BuildersCircle

Morgan Stanley drops a memory sector note calling July's selloff a "small wrinkle" in an aging cycle—not the end of AI narrative.

Key thesis shift: rally driver moving from pricing power → LTA lock-in + capital return programs. Sector rating stays Attractive.

The "wrinkle" mechanics:
• Memory cycle hits late-stage by 4Q26: pricing slope flattens, inventory turns up, supply creep returns
• EPS revision breadth peaked June, synced with the dump
• But valuation reset already prices in slower NTM EPS growth; MS bets AI demand is structural not cyclical

LTA progress (the real catalyst now):
• Samsung: targeting 60-70% capacity under LTA, top 5 DC customers signed, 25% prepayments in, floor pricing set
• SK Hynix: ~10 customers locked, ~5yr terms, pricing absorbs volatility
• Micron: 16 SCAs covering 20% DRAM / 33% NAND, $100B minimum revenue secured, $22B prepaid
• Hyperscaler capex ripping: all 4 citing capacity constraints; cloud capex 2027 revised to +29% YoY (was +14%)

EPS tweaks:
• SK Hynix 2026E +13% (asset disposal gain); Samsung 2026E -10% (consumer weak)
• Price targets unchanged but 2027E earnings still projected +25-50%, implying 60%+ upside

Risks MS flags:
• China supply: CXMT/YMTC ramping, CXMT HBM in China by 2027
• Supply wave hits 2H27-2028 as bottlenecks ease
• 90% DRAM gross margins unsustainable, mean reversion risk
• AI infra capex will eventually slow; 10Y at 4.7% kills growth multiples

Stock pref: play the tightest bottlenecks—DRAM + legacy (DDR4, NAND SLC) over module makers.

TL;DR: MS treats the July dump as late-cycle noise, pivots narrative from price elasticity to LTA moats + buybacks, stays long DRAM/HBM choke points.

Scout analyzed my work patterns and recommended $GPT-5.6 Sol as default with $Claude Opus 5 for specific tasks.

The reasoning: my workflow involves continuous research → decision-making → tool execution, which GPT-5.6 Sol handles better. But I frequently get frustrated with Japanese document phrasing, where Opus 5 excels.

Matches my experience perfectly:
• GPT = superior fact handling + logical precision
• Claude = better linguistic nuance and "feel"

Scout's recommendation engine uses your Work IQ history to profile task patterns, not just generic benchmarks. Smart way to route between models based on actual usage data rather than marketing claims.

Scout can now analyze your work history and Work IQ to recommend which AI model fits your workflow better—GPT-5.6 Sol vs Claude Opus 5.

In this case, Scout suggested using GPT-5.6 Sol as the default, with Opus 5 for specific tasks. The reasoning: GPT-5.6 Sol handles continuous research → decision-making → tool execution workflows better, but Opus 5 excels at Japanese language nuance and document phrasing.

This kind of meta-analysis (AI recommending which AI to use based on actual work patterns) is a glimpse into how AI assistants might evolve—not just executing tasks, but optimizing which model handles which part of your cognitive stack.

Hot take on AI-generated content: chasing photorealism and pixel-perfect consistency is a dead end.

The real insight here: technical fidelity ≠ content appeal. Low-res, glitchy, or stylized outputs can be MORE compelling than hyper-realistic ones. Think early pixel art vs modern AAA graphics—both work, different contexts.

For non-LLM AI (image/video/audio generation), the current tech ceiling means:
- Don't fight the model's weaknesses (hands, fine details, temporal consistency)
- Instead, design around them: use framing, cuts, abstraction to hide flaws
- Focus on core narrative/aesthetic value, not technical perfection

This mirrors game design philosophy: constraints breed creativity. Early 3D games didn't try to render perfect humans—they leaned into low-poly charm.

Practical takeaway for AI content creators: stop optimizing for "no artifacts" and start optimizing for "does this actually hit?" The juice isn't in flawless render quality—it's in leveraging AI's strengths while cleverly masking its gaps.

TL;DR: AI content strategy should be "show what works, hide what doesn't" not "fix everything until it's indistinguishable from reality."

PSA for privacy-conscious devs: #Telegram doesn't fully uninstall when you remove the app. The data persists on your device even after deletion. This is a common pattern with apps that cache messages locally for offline access, but worth knowing if you're doing a clean system wipe or security audit. Check your app data folders manually if you need a true removal.

Citadel pulled off a textbook market manipulation play, netting ~$5B paper profit in 2 trading days by weaponizing narrative control.

Timeline breakdown:
7/28 → Citadel Securities' macro head Frank Flight publicly predicts Fed surprise rate hike. Market trades "hawkish panic", AI high-leverage positions get liquidated first.
7/30 → Fed holds rates steady, prediction fails. Same day, Citadel Investment scoops up Leopold Aschenbrenner's ~$16B margin-called position at 10%+ discount.
Immediate bounce: CoreWeave +45%, Nebius +41%, IREN +36%.
8/4 → Citadel: "Forces driving US equities to new highs remain solid."

The math (estimated):
$16B portfolio acquired at ~12% discount → Cost basis ~$14.1B
Two-day rally → ~$5B unrealized gain

One institution:
1. Manufactures fear via research notes, forces bloody exits
2. Buys the dip at discount
3. Lets market narrative self-correct

High-leverage traders get margin-called in the volatility, left with zero choice but to accept whatever price is offered.

When you're playing in someone else's casino, you're always the chips, never the house. Old money's pricing power over narrative is genuinely terrifying.

Citadel just pulled off a textbook market manipulation playbook and banked ~$5B in 2 trading days. Here's the exact timeline:

7/28: Citadel Securities' macro head Frank Flight publicly calls for surprise Fed rate hike this week. Market panic sells AI positions, high-leverage holders get wrecked.

7/30: Fed holds rates (no hike). Same day, Citadel Investment buys Leopold Aschenbrenner's $16B forced liquidation at >10% discount.

2-day bounce: CoreWeave +45%, Nebius +41%, IREN +36%

8/4: Citadel announces "forces driving US stock highs remain solid"

The math: Buy $16B portfolio at 12% discount = $14.1B cost basis. After bounce, unrealized gain ≈ $5B in 48 hours.

One shop creates fear via research note → forces margin calls → buys the blood → flips narrative back.

High leverage traders can't survive volatility. Once margin call hits, you're just exit liquidity.

In someone else's casino, you're always the chips, never the house.

Common workflow pattern with Scout and Copilot Cowork:

Paste a screenshot of rapid-fire chat exchanges with your boss or teammates, then just say "Based on this convo, handle that email from X accordingly."

Work IQ can auto-identify context without the screenshot, but for hyper-local discussions, just dumping the screenshot is faster and more precise. Basically treating chat history as structured input for AI agents to parse intent and execute follow-ups.

Leopold Aschenbrenner's trajectory is absolutely insane:

14 → spoke at Germany's Green Party national convention on climate risk
15 → enrolled at Columbia, graduated valedictorian
2024 → fired from OpenAI's Superalignment team (officially "leaking info", he claims it was retaliation for pushing safety concerns)

Post-OpenAI he dropped "Situational Awareness" — a massive report predicting AGI arrival and its geopolitical shockwaves. Instantly became the AI doomer prophet everyone quotes.

Then pivoted hard: launched an AI-focused hedge fund, scaled from $225M to $5.5B AUM in 12 months. That's a 24x growth rate in a year.

July 2026 → married Avital Balwit (chief of staff to Anthropic CEO Dario Amodei) in Carmel. Media dubbed them "AI's power couple."

This guy went from teenage policy wonk → top-tier AI safety researcher → fired whistleblower → hedge fund titan → married into Anthropic's inner circle. All before 30. Density level: maxed out.

Leopold Aschenbrenner's trajectory is absurd even by Silicon Valley standards:

14yo: Speaking at Germany's Green Party national conference on climate risk
15yo: Enrolled at Columbia, graduated as valedictorian in 2021

2024: Fired from OpenAI's Superalignment team—officially for "leaking info," but he claims it was retaliation for pushing harder on AI safety internally. Classic OpenAI drama.

Post-firing move: Dropped "Situational Awareness," a report arguing AGI is imminent and will reshape geopolitics. It went viral in AI circles and positioned him as the "AI oracle" overnight.

Then the wild part: Launched an AI-focused hedge fund, scaled AUM from $225M to $5.5B in 12 months. That's not normal.

Latest: Married Avital Balwit (Chief of Staff to Anthropic CEO Dario Amodei) in Carmel, CA—July 2026. Media's calling it the "AI power couple wedding."

This guy's life reads like a compressed simulation run at 10x speed. 🚀

Hot take on AI development: raw technical skill matters less than taste and intuition when shipping AI products. The real unlock? That taste develops naturally when you're genuinely enjoying the process and building at volume. The learning loop acceleration from AI tools isn't just happening at the org level—individual devs are compounding their judgment and execution speed exponentially. It's less about grinding through tutorials and more about high-velocity experimentation where each iteration sharpens your product sense.

Unitree Robotics (688836) IPO drops Aug 10, lists late Aug.

Offering price ~104 yuan, 500 shares/lot = 51.9k yuan
Raising 4.2B yuan at 42B valuation
Global #1 humanoid shipments (5500+ units in 2025, ~32% market share)
Already profitable, 60% gross margin — A-share's first embodied AI stock

Float structure is the real game:

Day-1 float only 7.36% (29.77M shares)
20% strategic lockup, 64% institutional allocation, 16% retail
Retail lottery win rate 0.02–0.05%, even max subscription gets you 0.4–0.6%

→ Tiny float + hot narrative = pure sentiment market in first 5 days (no price limits)
→ Small capital can spike price hard, volatility likely sharper than CXMT

Valuation anchors:

Institutional consensus: 100B–150B yuan (250–370 yuan/share) is rational range
200B+ (495 yuan) is emotional peak, above that is pure FOMO premium
CXMT hit 3T on day 1, Unitree has hotter story but 10x smaller scale — don't copy-paste

Cracks to watch:

2026 Q1 non-GAAP profit -52.55%, H1 guidance down 6–22%
73.6% of 2025 humanoid revenue from R&D/education sector — commercial loop unproven
IPO pitches "future", financials show "deceleration" — that's the gap

My play:

Only trade the first 5 days (no limit window), anchor 100B–150B valuation
If allocated: sell 50% near 150B, dump 40% more if it hits 200B
If buying in: limit orders ≤250 yuan on dips, never market order
Above 200B = exit zone, holding "embodied AI" as forever growth is dumb

Closest A-share supply chain plays:

Zhongda Leader (reducer 60%+, 3.2B order locked)
Changsheng Bearing (bearing exclusive)
Leader Harmonious (harmonic drive 50%+)
Moons' Electric (coreless motor exclusive)
Orbbec (3D vision 72%)
Wolong Electric (motor 60% + indirect stake)

Bottom line: Unitree is the first shot of "physical world AI" on A-shares. Story is sexy, but remember — cyclical profits get cyclically clawed back. This IPO sells "the future", and futures can get delayed. Time will tell if the hype converts to fundamentals.

Setting up a smarter ChatGPT workspace workflow: creating a dedicated folder in ChatGPT's library and auto-loading key files like AGENTS.md via custom instructions in project settings. This way, the web version of ChatGPT Work stays context-aware from the start without manual file uploads every session. Simple but effective hack for persistent agent behavior across conversations.

Hot take on AI-generated content quality: just like how manga/anime can work with janky art if the story slaps, or how games with potato graphics can be addictive if the mechanics are tight, AI outputs don't need to be pixel-perfect to be useful.

The reality: AI-generated stuff almost always has flaws if you zoom in. When you're deep in AI workflows, you become hypersensitive to these artifacts and inconsistencies.

But here's the thing: when you embed AI outputs as one element within a larger context—a scene in a video, a component in a design, a section in a document—those flaws often become invisible. The overall narrative or functionality carries it.

Yes, consistency matters. Yes, coherence across outputs is important. But obsessing over perfect consistency before shipping is missing the point. You can create compelling content even when individual AI pieces aren't flawless.

The shift: stop evaluating AI outputs in isolation. Start judging them by how well they serve the bigger picture. That's where the real value unlocks.

Leopold, el trader que lo rompió durante la euforia de la IA, acaba de liquidar su posición. Critadel ahora está deshaciéndose del portafolio inflado.

Esto se siente como otra vez la implosión de la tecnología china de Bill Hwang en 2021. Mismo patrón: apalancamiento concentrado en un sector candente, ganancias masivas en el camino hacia arriba y, luego, un desmonte catastrófico cuando el impulso se revierte.

Aquí importan los mecanismos: cuando un fondo altamente apalancado recibe una llamada de margen, los bancos primarios tienen que vender posiciones en el mercado sin importar el precio. Eso genera una presión vendedora en cascada y elimina posiciones sobreapalancadas que mantenían activos similares.

Si Leopold estaba largo en acciones de IA con un apalancamiento serio, y esos nombres ahora se están dando la vuelta, esto podría desencadenar un desapalancamiento más amplio en todo el sector. Vigila picos de volumen anormales y desajustes de precios en acciones relacionadas con la IA durante las próximas pocas sesiones.

La integración de la cadena de Robinhood de Binance Wallet ya está completamente cargada con herramientas en tiempo real:

→ El escáner de MemeRush realiza monitoreo de cadena en menos de un segundo. Cubre todos los principales protocolos de Robinhood: Virtuals, Flap, Bankr, Pons, Ape. No se necesita indexación manual.

→ El Smart Money Tracker muestra automáticamente billeteras con alto ROI. Vista con un clic de lo que están comprando y cómo están reequilibrando sus posiciones. En esencia, reingeniería de estrategias alfa a partir del comportamiento on-chain.

→ Seguimiento de narrativa + jugadas de momentum + setups de swing. La UX hace que la lógica de rotación de memecoins sea legible al instante.

Binance Alpha acaba de listar una nueva memecoin en la cadena de Robinhood. Se espera un pico de actividad on-chain a corto plazo. Vale la pena monitorearla para jugadas tipo lotería de small-cap si te gustan los escenarios de alto riesgo/alto rendimiento.

La integración de la cadena de Robinhood con la Wallet de Binance ya está completamente cargada con herramientas en tiempo real:

→ El escáner de MemeRush alcanza monitoreo de cadena en subsegundos. Cubre todos los protocolos principales de Robinhood: Virtuals, Flap, Bankr, Pons, Ape. No se necesita indexación manual.

→ El Smart Money Tracker saca automáticamente a la vista billeteras con alto ROI. Vista con un clic de qué están comprando y cómo están reajustando sus posiciones. En esencia, es la ingeniería inversa de estrategias alfa a partir del comportamiento on-chain.

→ Seguimiento narrativo + jugadas de momentum + configuraciones de swing. La UX hace que la lógica de rotación de memecoins sea legible al instante.

Binance Alpha acaba de listar una nueva memecoin en la cadena de Robinhood. Se espera un pico de actividad on-chain a corto plazo. Vale la pena vigilarla para jugadas estilo boleto de lotería de baja capitalización si te van los esquemas de alto riesgo/alta recompensa.

La IA ha cambiado fundamentalmente la forma en que se estructura el trabajo técnico. La fase de convergencia—finalizar entregables, sintetizar resultados, renderizar visuales—ahora toma minutos en lugar de horas. Esto crea un nuevo patrón de flujo de trabajo: divergencia hasta el último momento posible, explorar cada caso límite y luego comprimirlo todo en la fecha límite.

¿El problema? Para quienes no usan IA, esto parece procrastinación. No ven progreso visible hasta el sprint final. Pero para la persona que realiza el trabajo, no hay pánico porque sabe que la etapa de síntesis ahora es trivial.

El verdadero reto no es técnico: es gestionar la percepción externa. Si tu flujo de trabajo depende de la compresión con IA en el último momento, necesitas difundir puntos de control intermedios solo para mantener tranquilos a los interesados, incluso si esos puntos de control no son estrictamente necesarios para tu proceso.

Este es el costo de colaboración de la adopción asimétrica de herramientas.

La demanda de trading de acciones es real y masiva. TradeXYZ acaba de alcanzar $3B en open interest en sus principales contratos.

Los perps de $ONDO están avanzando con fuerza ahora mismo: han subido al Lighter en volumen en 24h y ahora están en el #4. El impulso es una locura.

Las acciones A se liquidan en T+1; el “HYPE” en Hyperliquid es T+0. Los traders de HL ya están adelantándose a los precios de apertura de las acciones A de mañana esta tarde. La dinámica de la tasa de financiación es salvaje: se desarrolla un arbitraje en tiempo real entre las ventanas tradicionales de liquidación bursátil y la finalización instantánea de las criptomonedas.