SpaceX's valuation reflects not just current financials but Elon Musk's unique ability to sell the market on long-dated call options (Starship, orbital data centers, Mars) that hi…

Robotics is in the first inning; by 2060 there could be a 1:1 ratio of useful robots to humans. Sunday's pseudo-humanoid home robot represents a pragmatic path, and Benchmark view…

A float of just 4% plus the opening of options plus retail mania creates an artificial gamma squeeze to the upside, but the lockup release in six months will restore price discove…

First AI model regulation based on capabilities under the Export Restriction Act with no judicial review. Anthropic painted a target on its back talking about the 'god model'. IPO…

A founder in his personal portfolio using Fable says the model is so powerful that Anthropic moves from a path-to-$100B ARR to $150-200B. Anecdotal but a demand data point with Fa…

A portfolio founder using Fable believes Anthropic will exceed $100B ARR this year, reaching $150-200B, because the model is so powerful it will drive massive incremental spend.

Mistral benefits from European sovereignty demands; even if its models trail US frontier, enterprises will prefer a second-best model they control over a best-in-class model that…

The market's SaaS filter has been discriminating: Palo Alto, CrowdStrike, Cloudflare, Datadog, and Palantir all trade above 15x NTM revenue. They have a usage-based component, an…

A meaningful part of the profit comes from TurboTax, now perceived as easily replicable by AI coding agents. A classic bad-attribute SaaS.

Figma also has Figma Make in the AI market with a usage-based component, but the market is not giving it credit because the bad outweighs the good (similar to Wix). Potential misp…

Has every bad SaaS attribute: 80% market share (only share to lose), product replicable by gen AI, wrong business model, no internal AI talent. Needs an institutional catalyst (ne…

Adobe's product experience is poor (licensing confusion, login issues) reflecting decades of value extraction; a cultural reset led by a product visionary (like Intercom's Owen) i…

Despite 15k robots deployed and $180M revenue, robotics penetration remains tiny because warehouse economics require robots to cost half of minimum wage labor, and real-world edge…

SpaceX's post-IPO stock will decline in 6 months because only 4% float trades, enabling gamma squeezes and forced buying from index inclusion that artificially inflate price befor…

Tesla's market cap exceeds fundamental auto valuation because Elon convinces markets to price in 10-year moonshots (FSD, Optimus) that he has a track record of delivering, creatin…

Elon exercised the right to buy Crusoe at $60B; estimated $4B in revenue today on the way to $6B by year-end. A prescient deal with lockup retention for the team. It solved the co…

Anthropic contract of $1.25B plus Google $700M (September activation) create roughly $24B/year in CoreWeave-like AI revenue, surpassing the combined SpaceX/Starlink run-rate. Elon…

Anthropic's IPO probability remains >50% because SpaceX's stellar reception sets a high prior; the Fable export restriction is a significant wrinkle but not a blocker given eye-wa…

Even if Mistral isn't as good as the US/China models, the sovereignty risk premium justifies the bid. They've scaled to $500M+ ARR with an FD model. The Anthropic ban reinforces t…

As countries pursue sovereign AI models, they will need to buy their own chips, creating a huge tailwind for Nvidia regardless of which model wins.

Smart acquisition because Finn/Intercom is the canonical pre-AI SaaS company that successfully pivoted to outcome-based pricing. Salesforce needs this same transformation to reign…

Intercom's successful pivot from seat-based to outcome-based pricing (per resolution) makes it a canonical example for Salesforce to reignite growth via Agentforce; the deal bring…

The stock trades at 1x revenue after a guidance cut and 20% layoffs. The burn-the-boats AI story is critical versus Lovable and Replit; the Base44 acquisition was smart, but the b…

Wix trades at 1x revenue after a failed buyback; bear case is website building becomes a subset of AI agents like Lovable/Replit, but if they create any AI leverage the stock is v…

A significant portion of Intuit's profit comes from TurboTax, which is perceived as highly replicable by AI coding agents, creating a structural threat to its core revenue.

At 1x revenue, Wix is a compelling take-private candidate; management should partner with PE (like Michael Dell/Silver Lake) to grind through the AI transition away from public ma…

Adobe's 80% market share means only share loss ahead; generative AI makes its product increasingly replicable, its seat-based model is outdated, and it lacks usage-based revenue o…

Hedge funds prefer Nvidia because it captures every AI tailwind (compute demand, inference scaling, sovereign chip buys) with dominant market share, while Adobe's cheapness is a v…

It trades at 16x earnings with an 80% share of AI's most critical compute stack. The SOX index is up 325% over five years versus the Bessemer Cloud index down 44%. The obvious tra…

Standard Bots avoids over-engineered humanoid legs, focusing on integrated hardware/software robotic arms that can be quickly trained for discrete manufacturing tasks, a more viab…

Elon sells long-dated call options to the market and tends to deliver. The robotics buildout is a slow journey (just 3M robots globally versus 1B manual workers); Optimus is the a…