The most contrarian call is [15] "ShitAI", which asserts that OpenAI, Anthropic, and similar players are retreating from scaling not due to technical limits but because their economic models rely on regulatory capture and hype—comparing them to Lehman pre-collapse. This outright rejection of the dominant AI bull thesis, branding leaders as “ShitAI,” stands in stark contrast to even cautious optimists.
Today’s AI skepticism ([15], [29], [43]) escalates prior bear arguments from *MBI Deep Dives* (Aug 31) and *Epoch AI* (Aug 24) that the strongest critiques come from within the AI community itself—now manifesting in forced unwinds (Leo), valuation resets, and elite safety discourse ([33]). Similarly, oil shortage warnings ([22], [35]) follow and intensify *HFI Research*’s Aug 19 thesis that the market is underpricing structural deficits, now validated by price action and inventory data.
The post highlights practical AI applications in niche domains like AI-powered basketball coaching and tooling comparisons (Jira vs. Linear), emphasizing efficiency gains for platform teams. Key takeaway: AI’s highest ROI may lie in specialized, domain-specific workflows rather than broad enterprise deployment—relevant for investors assessing productivity tool adoption in tech (watch LNKD, PLTR for signals on workflow AI integration).
Huawei is unlikely to close the AI chip gap with Nvidia this decade due to inferior per-chip performance and constrained production from export controls, remaining under 4% of Nvidia’s 2026 compute output. US GDP growth has been understated by ~0.3 percentage points over the past year due to missed valuation of Nvidia’s domestically designed, overseas-manufactured chips. GPT-6 Astra sets new benchmarks, topping the Epoch Capabilities Index and solving 2/68 Erdős problems—first on FrontierMath Erdős—while data center power capacity doubles every 10 months (now 950 MW at SpaceXAI’s Colossus 2).
Photosynth (4379 JP) is a Japanese biotech firm advancing a novel photosynthetic cell therapy platform with potential applications in chronic ischemia and degenerative diseases, supported by early clinical traction and partnerships with major research institutions. The core thesis hinges on successful Phase II data readouts in 2027, which could re-rate the stock; monitor 4Q26 preclinical results as a key catalyst. Actionable view: Consider initiating a speculative long position ahead of data, given current market cap of ¥85B implies limited downside if trials fail but >3x upside if successful.
MetaMaterials’ Series A preferred stock ($MMTLP) was part of a misleading corporate structure that funneled retail investors into acquiring worthless shares of Nextbridge Hydrocarbons (NBH), an insolvent oil and gas spin-off, via a reverse merger and flawed distribution mechanism. The key event: $MMTLP stopped trading on December 9, 2022—despite widespread retail belief it would trade until the 12th—because shares purchased after December 8 couldn’t settle before cancellation, rendering post-8th trades impossible under T+2; this was clearly stated in the FINRA notice but ignored or misunderstood. Retail losses likely exceed $100 million, fueled by social media-driven misinformation claiming an imminent short squeeze and false narratives later debunked by an SEC order confirming Torchlight Energy (the originator) misrepresented its intentions and orchestrated the squeeze narrative—actionable takeaway: avoid $MMTLP/$NBH entirely, recognize this as a case study in retail manipulation via opaque structures and coordinated disinformation.
The post argues that the key to long-term investment success lies not in choosing between a great business (the "horse") or a great manager (the "jockey"), but in identifying businesses where capital allocation is critical—i.e., cash-rich, mature companies with limited reinvestment opportunities—and ensuring a durable governance structure ("the stable") that secures competent succession. Actionable insight: favor companies with strong, institutionalized capital allocation frameworks—like those run by Thorndike’s "Outsiders"—and prioritize governance durability over individual brilliance, as managerial tenure is finite and irreplaceable without proper succession mechanisms.
The Alpha Engineer highlights increased institutional accumulation in semiconductor equities, particularly ASML, following the recent pullback, citing favorable long-term positioning ahead of the AI capex rebound. Based on 13F transaction clustering and options flow, the report suggests a tactical opportunity with a $650 price target by Q2 2027. Action: Add ASML to tech growth sleeve on dips near $520.
The Alpha Engineer’s weekly ranking highlights high-conviction small-cap and microcap stocks, with new entries into the top 5% of the model. As of Sep 13, 2026, newly ranked top performers include tickers like VCY:CAN (Vicinity Motor Corp., Canada, $85M market cap) and AOSL (Alpha and Omega Semiconductor, US, $1.2B), both showing strong momentum and liquidity. Focus remains on microcaps ($20M–$300M) for outsized return potential, with actionable ideas emerging from global markets but filtered for tradability (> $50K avg daily volume, > $0.50 price).
Adobe (ADBE) and Oracle (ORCL) both reported volatile earnings reactions, with initial market moves reversing by day’s end. Adobe shows decelerating RPO and ARR growth despite a lower valuation, raising concerns about monetizing its freemium push, while Oracle’s strong capex returns face skepticism over sustainability once cloud supply normalizes. Macro-wise, Stanley Druckenmiller warned the AI-driven earnings surge may be transient, echoing caution that AI adoption could hit bottlenecks or saturation, implying selectivity is key—bullish on agentic AI’s long-term runway but bearish on current winner expectations.
Sung Capital reiterates its thesis on a small Japanese software company (not named) with a significant divergence between cash flow and reported earnings, driven by a high-growth segment masked by slower legacy operations. Following a strong Q32026 beat, the firm maintains conviction in the compounder potential, supported by a 5-year FCF model implying substantial multiple expansion—actionable for investors seeking undervalued, cash-generative tech exposure in Japan.
The Hermit highlights $APG and $FTC as companies benefiting from AI and space infrastructure trends but now priced for perfection, implying elevated vulnerability to any operational or macro shortfall. While not issuing a formal short recommendation, the author expects mean reversion ahead and flags $FTC with a low 3.0/10 grade, signaling overvaluation relative to fundamentals. Actionable view: consider trimming or avoiding exposure ahead of potential de-rating.
Nvidia has evolved from a chipmaker into the architect of AI "factories," designing full-stack $100B data centers that dominate energy-constrained environments by maximizing tokens/watt—the key ROI metric. Its ecosystem dominance (via partnerships, financing, and tech integration) creates self-reinforcing network effects, favoring Nvidia (NVDA) and its neocloud partners like CoreWeave (private), while pressuring alternatives like AMD; monitor tokens/watt efficiency and GW-scale deployment capacity as critical value drivers.
This week’s Latticework digest curates key insights on AI’s evolving infrastructure (e.g., Marvell CEO on AI bottlenecks, a16z on rethinking AI investing), macro risks (Gundlach on long bonds), and valuation discipline (Damodaran on scaling vs. profitability), with thematic undercurrents around rising compute costs and physical AI. Notable tickers mentioned include Nvidia, XPeng, and Stripe (via Kai AI agent usage). Actionable idea: Monitor AI’s hardware and energy constraints as potential inflection points for investment in infrastructure plays like Marvell or Base Power.
SED Energy (ENH) and Ventura Offshore (VTURA) are merging in an all-share deal (5.5:1 indicative ratio), creating a $1bn EV offshore player with exposure to tender rigs, deepwater drilling, and seismic. The combined entity trades at ~4x EV/EBITDA—well below US-listed peers at 7-8x—and generates ~$200m annual FCF despite conservative assumptions, enabling dividend yields of 19-36% (NOK 1.4–2.6/share) post-refinancing and deleveraging to ~1x net debt/EBITDA. Key catalysts: NYSE dual-listing for improved liquidity, removal of Ventura’s 10% bond overhang, and accretive bolt-on M&A enabled by scale; the deal is accretive to SED shareholders unless Catarina remains uncontracted.
The author argues that traditional 60/40 portfolios are broken due to bonds’ persistent negative real yields and loss of inverse correlation with equities, rendering them ineffective hedges. However, he suggests a re-evaluation of bonds in the current regime, hinting at their potential renewed role as safe-haven assets if macro conditions shift toward disinflation and risk-off sentiment. No specific ticker is given, but the implication is to monitor long-duration Treasuries (e.g., TLT) for tactical allocation if flight-to-safety dynamics return.
The post argues that the AI narrative has shifted from boundless optimism to existential fear due to the collapse of the "brute force" model of AI development, as OpenAI, Anthropic, and others retreat from aggressive scaling—paralleling the 2008 Lehman Brothers unraveling. The author dubs them "ShitAI" for relying on regulatory capture and hype over sustainable economics, with implications for data center demand and AI infrastructure stocks. No specific ticker is recommended, but the piece signals growing skepticism toward large AI capex plays and closed-model vendors.
Pre-Fed jitters are mounting, with S&P500 breadth deteriorating, liquidity tightening, and correlations rising—bearish technicals across growth, semis, and credit. Equal-weighted performance and corporate bond weakness signal caution; a bullish resolution requires improved liquidity and breadth, but for now, risk-off is the dominant theme. Action: Favor value over growth, monitor pre-FOMC positioning, and watch for a breakout in breadth as a re-risking signal.
Iran escalation and rising oil (CL1 > $100) are fueling inflation and pushing 10Y yields to 5%, lifting September hike odds to 86.5%, creating a headwind for commodities and precious metals—yet we see this as a tactical pullback ("Warsh’s gift") to buy, not a trend reversal. Copper (COPPER COMDTY) corrected on stalled White House tariffs, but structural demand remains; we use the dip to add, targeting $5/lb as a floor. Credit markets show strain in AI/data center buildout—$125B in debt, much off-balance-sheet, with demand cooling (cover ratios <2x), favoring private credit and high-yield; DeepSeek’s $75B pre-IPO valuation pressures inflated AI multiples. Yen broke 155—intervention level—suggesting trend shift: keep selling USD/JPY rallies.
This post is a promotional teaser for 1035 Capital’s paid subscription services, highlighting access to 240+ proprietary charts, small cap research, and the BullFrog AI (BFRG) stock feature. No concrete market thesis or actionable data point is provided in the free update—core insights and performance data are gated behind the paywall. The only named ticker is BFRG, but no specific buy/sell recommendation or analytical view is offered.
Legacy Education (LGCY) operates a chain of for-profit career colleges in California focused on allied health and nursing programs, targeting non-traditional students seeking direct pathways to employment. The company’s model hinges on regulatory compliance, clinical placement access, and student outcomes—key drivers of its ability to attract Title IV funding and deliver on employment promises. While LGCY is expanding via acquisitions, program additions, and campus growth, its success depends on execution across a complex, highly regulated operating environment with thin margins for error.
The post argues that the AI rally is exhibiting classic bubble dynamics, where deteriorating company quality and backdoor listings (e.g., reverse mergers) signal late-cycle excess—akin to past market manias. Obscure firms with dubious histories are resurfacing with AI-themed narratives to exploit investor enthusiasm, a warning sign of froth. While no specific short call is made, the implication is to avoid speculative micro-caps chasing the AI theme and remain wary of "backdoor" public companies pushing hype over fundamentals.
The post highlights three key investment insights from recent reads: Boka Group’s success stems from a codified culture of hospitality—a durable, if intangible, competitive advantage that allows scalability in a low-moat industry; J.P. Morgan’s longevity underscores the cyclical nature of financial excess and regulation, favoring firms with fortress balance sheets and risk discipline through cycles; and Bellingcat’s open-source intelligence model reinforces the investor edge in synthesizing public data into a verified mosaic, challenging consensus with independent analysis. While no direct ticker is recommended, the implicit view favors businesses with strong cultures, resilient balance sheets, and the ability to thrive across cycles—particularly relevant amid today’s AI-driven capital spending boom, which mirrors the 1873 railroad bubble.
Josh Young argues that oil is headed for a massive shortage due to chronic underinvestment in supply, driving prices significantly higher. He sees leveraged exposure in undervalued E&P equities, specifically highlighting VRIC as a high-conviction play on rising oil prices. Recent price strength in oil validates his near-term thesis.
Core Scientific ($CORZ) has delivered on key execution milestones—completing nearly all CoreWeave build-outs and securing AMD as a second tenant with 377MW direct lease (15-year, triple-net, 2.5% escalators), plus conditional credit-backed leases totaling 1.1GW—yet the stock fell ~40% to $17.93, driven by broader AI selloffs and ERCOT-related delays in Texas. Despite $14B+ in contracted revenue and a strengthened balance sheet via $3.3B secured notes and liquidity enhancements, the market appears to price in excessive pessimism; valuation gap exists if leases are valued independently and near-term execution risks are discounted. Action: Analyze implied valuation of unpriced capacity and assess catalyst potential from ERCOT resolution and 2027–2028 ramp.
Li and Mulvey (2026) propose a regime framework where "optimal" assets are identified in hindsight, boosting industry rotation Sharpe ratios from 0.56 to ~1.0 (1995–2025) by rotating into semiconductors offensively and retreating to tobacco/T-bills defensively when drawdown risk spikes—key insight: clean crash protection via regime labeling matters more than sector selection. Separately, Bhambhwani (2026) shows Aave’s stablecoin borrowing rates track the 10-year Treasury yield (beta ~1.1 per SD move), implying DeFi yields are tethered to traditional rates; meanwhile, Liu (2026) finds nested daily price ranges signal turning points without tuning, especially in individual stocks, suggesting range geometry adds predictive value missed by return-only models.
The podcast highlights a disciplined quality investing framework emphasizing durable moats, rational decision-making via mental models (e.g., inversion, margin of safety), and long-term compounding over hype-driven trends like AI. While not naming specific tickers, the analyst stresses avoiding overvaluation risks in AI-related stocks by focusing on businesses with proven reinvestment capacity and pricing power—actionable lens: favor quality firms with legacy moats and strong free cash flow yields (>4%) trading below 20x earnings.
The core thesis distinguishes between "owners" — investors who focus on durable business fundamentals like returns on capital, pricing power, and competitive positioning — and "renters" — short-term traders driven by price momentum and sentiment. Sellmann argues that the current market rewards renting behavior, but long-term outperformance comes from owning businesses you can describe concisely across eight key parameters (e.g., product, customer, margin structure) and holding them with a 10-year horizon, as Buffett did with the Washington Post. Actionable insight: Test your understanding by articulating any investment in two clear, jargon-free sentences that capture the full business model — if you can’t, you don’t truly own it mentally, regardless of position size.
The post draws a philosophical parallel between *Blade Runner* and modern questions of identity and authenticity, implying growing societal unease with artificiality—relevant to tech and AI-driven markets. No actionable ticker or data point provided; purely thematic commentary.
The post argues that climate chaos is accelerating systemic failures in infrastructure, supply chains, and governance, necessitating decentralized, community-based adaptation as top-down systems prove inadequate. Investors should anticipate stranded assets in centralized utilities and real estate in high-risk zones, while positioning for growth in distributed energy (e.g., solar + storage), water resilience tech, and local food systems—sectors aligned with on-the-ground survival infrastructure. No specific ticker cited, but implied focus on resilient, localized physical assets.
The post argues that in both investing and policy, an overreliance on flawed or reified metrics—such as body counts in war or token prices in AI—leads to poor outcomes, citing Goodhart’s Law: when a metric becomes a target, it ceases to be meaningful. It questions the AI investment boom’s sustainability, highlighting rising data center costs, falling token prices, and unproven returns, while drawing parallels to dot-com era excesses; actionable takeaway: scrutinize AI-related valuations (e.g., OpenAI, Anthropic) where revenue growth relies on unrealistic market expansion amid commoditizing models.
Younger generations prioritize experiences over material wealth, driving durable demand for travel and IRL (in real life) events amid declining trust in traditional systems—evidenced by gold rising to $4,000/oz as a "thermometer" of systemic uncertainty. This shift supports investments in experiential sectors and hard assets like gold, BTC, and insurance (e.g., WR Berkeley), while favoring business models with minimal operational friction and maximum autonomy.
Pre-Fed jitters are mounting, with S&P500 breadth deteriorating, liquidity tightening, and correlations rising—bearish technicals across growth, semis, and corporate bonds. Key risk: a hawkish Fed pivot; only a dovish outcome or stabilization in bond markets could resolve bullishly. Monitor SPX (especially equal-weighted) and HYG for near-term risk signals.
The Lancang-Mekong Cooperation (LMC) has driven significant trade growth, with China’s trade with Cambodia, Laos, Myanmar, Thailand, and Vietnam reaching $500.9 billion in 2025—up 150% over a decade—but the benefits are uneven, favoring China through infrastructure dominance and import surpluses. Market implication: Consider underweighting commodity-dependent ASEAN exporters (e.g., KTC TBSP.F) exposed to Chinese demand volatility; monitor Thai and Vietnamese logistics plays (e.g., THCOM.BK) for LMC-related infrastructure tailwinds.
The post captures the surging intensity of AI safety discourse, with key figures like Sam Altman and Elon Musk signaling concern, while highlighting a growing grassroots movement—evidenced by high-demand events like the Berkeley meetup (Sept 14) featuring Redwood Research and OpenAI policy insiders. No direct market call or ticker is made; the actionable insight is qualitative: elite networks are converging on AI risk, suggesting heightened regulatory and societal attention ahead—monitor entities like Redwood Research and Eleos AI for early signals.
The post argues that Sabah’s 1963 entry into Malaysia was based on a negotiated “founding bargain” with specific constitutional safeguards—covering autonomy, representation, and resource control—that have not been fully implemented, despite the formal establishment of Malaysia on 16 September 1963. Citing British records, the Malaysia Agreement (MA63), and the Inter-Governmental Committee Report, it asserts that Sabah never achieved full international independence on 31 August 1963 and questions whether the promised federation was ever delivered, particularly after Singapore’s 1965 exit weakened Borneo’s political balance. While no direct market view or ticker is provided, the analysis implies long-term governance and federal-state risk in East Malaysia that could affect infrastructure, natural resource projects (e.g., oil & gas, forestry), and policy stability for firms operating in Sabah and Sarawak.
The post argues that despite visible oil inventories showing only modest draws (3-4M bpd, partly from SPR releases), the market is underestimating supply tightness due to unaccounted-for depletion of *non-visible* inventories—especially in China—amid Middle East escalation. With Chinese product stocks drawing (gasoline down 2.2M bbl, diesel 1.7M bbl in August per Goldman) and refiners poised to boost runs, latent demand reemergence will tighten global balances further, forcing higher prices to induce demand destruction; upside pressure on crude (ticker: CL=F) is inevitable unless geopolitical tensions ease materially, which the author deems unlikely given Iran’s entrenched strategic position.
Federer’s longevity stemmed not from relentless grind but from disciplined restraint—protecting personal priorities to sustain peak performance, a metaphor for strategic pullbacks in markets. In energy, October crude (CLV2026) approached its 107 target after a head-and-shoulders breakout, while November RBOB (RBX2026) neared 3.3967; we remain long Brent and wheat (KEZ2026), exited corn (ZCZ2026) at 548.75, and watch copper (HGZ2026) for potential failure at the neckline.
The post argues that while forgetting past pain can be a psychological advantage, the most traumatic experiences—like personal loss or 9/11—are both impossible and wrong to forget; their lasting impact gains meaning when shared openly, transforming pain into resilience. The core market-irrelevant insight is the power of collective memory and emotional catharsis in driving human perseverance, symbolized by rebuilding at Ground Zero and a friend’s decision to embrace joy after tragedy—actionable only as a reminder of intangible drivers of long-term societal and personal recovery. No ticker or data point.
GPT-6 Astra represents a significant leap in AI capability, particularly for ambitious, multi-step tasks involving 3D modeling, scientific reasoning, computer use, and agent coordination—outperforming prior models in benchmarks like FrontierMath Tier 4 (98%) and ARC-AGI 3 (99.9%). While not a quantum leap in general coding or conversational fluency, it excels in structured, high-complexity domains; OpenAI touts its alignment and safety improvements, though claims of AGI are premature and potentially misleading due to benchmark overstatement (e.g., ExploitBench). Actionable view: Use Astra alongside Fable 5.1 (ticker: MSFT for OpenAI exposure) for high-stakes R&D, design, and automation workflows, but remain skeptical of marketing hype—real-world edge comes from task-specific experimentation.
The post argues that five free, self-study resources—covering finance certifications, CFA formulas, financial analysis, macroeconomics fundamentals, and a Warren Buffett compendium—deliver more practical investment knowledge than a traditional MBA. No specific market view or ticker is offered; the actionable takeaway is to leverage these educational PDFs to build analytical edge at low cost.
Li and Mulvey (2026) propose a regime-switching model that defines market states by the optimal hindsight asset choice, boosting industry rotation Sharpe from 0.56 to ~1.0 and cutting max drawdown from 57% to 22% (1995–2025) via a hierarchical overlay favoring tobacco/T-bills in risk-off regimes; the edge stems from clean crash protection, not sector selection. Separately, Bhambhwani (2026) shows Aave’s stablecoin borrowing rates rise ~1.1 ppt per 1 SD increase in 10-year Treasury yields, implying DeFi yields are tethered to traditional rates—bearish for stablecoin yield seekers if terminals keep rising.
The post uses a mix of satire and insight to argue that standing out from the crowd—whether in illicit ice cream vending, AI-driven writing, or professional hockey—is essential for impact and success. The core market takeaway: in saturated markets, differentiation trumps conformity, and investors should back businesses or individuals who defy pack behavior to capture asymmetric rewards. No specific ticker, but the implicit view favors disruptive, non-consensus strategies in media, tech, and consumer sectors.
The author reviews progress on key investment themes for 2026 outlined earlier in the year, maintaining conviction in core macro trends like energy normalization and tech-driven disruption, while acknowledging underemphasis on geopolitical risks and labor dynamics. No specific tickers are called out, but continued overweight exposure to energy infrastructure and AI-enabling technologies is implied as central to the portfolio strategy. Key data point: energy sector volatility remains below historical norms despite structural supply constraints, suggesting a pending repricing opportunity.
Situational Awareness, Leo Aschenbrenner’s AI-focused hedge fund, lost ~$35B in July (67% drawdown) due to a highly concentrated, leveraged AI stock portfolio that was forcibly unwound, with Citadel absorbing much of the book; despite this, Aschenbrenner has returned weeks later trading options in AMD, SK Hynix, CoreWeave, Bloom Energy, and the Roundhill Memory ETF—reengaging similar high-risk exposures. The episode underscores extreme fragility in crowded AI trades amid deteriorating macro conditions (high yields, sticky inflation, slowing growth), raising red flags about repetition of blowup risk, especially given reliance on leverage and lack of apparent risk management evolution. No actionable recommendation, but strong caution on levered AI infrastructure bets and skepticism toward media-anointed “genius” narratives absent durable strategy.