Before AI can be adopted, it must work as a business. Today it doesn’t. The economics are broken on both sides of the ledger: revenue is mispriced and costs are uncontrolled. While chipmakers may prop up the NASDAQ by selling infra to LLMs, LLMs must make dollars and sense
The railroad crash of 1873 corrected capital excess without changing the product. The AI crash will correct capital excess by forcing the product itself to change. Why the 1873 railroad analogy is attractive — and incomplete The surface analogy is strong: * massive capital inflows * speculative overbuild * infrastructure ahead of demand * financial
There are three markets for AI. They are routinely conflated, incorrectly marketed as one, and that mistake explains why AI adoption, trust and monetization are breaking down at the same time. 1) Consumers: the mass market Consumers don’t want to think better. They want life to be easier. They
AI companies are shipping features at full speed while the fundamentals that can drive profitability — adoption, trust, governance, cost and revenue — remain unresolved. Feature Impact on bottom line Multimodal inputs and outputs Text, image, voice, and mixed-media interaction improvements. Improves demos and perceived capability, but shows no evidence of increased
Is AI going to work out? Not without structural changes. What says AI must change? Data. Rates of adoption went flat almost a year ago for LLMs across the board. This is true at both at the consumer level (ChatGPT) and at the enterprise level (Microsoft, Salesforce). Why did the
Download PDF then use Share to send it to your own email for retrieval What to listen for What this means 1. AI revenue Are AI revenues discussed in concrete dollar terms, or described using words like “strong,” “early,” “encouraging,” or “meaningful”? When revenue is material, it is quantified. Continued
AI 3.0 is three weeks away. Wait. What? It's as simple as 1-2-3: AI 1.0 has never made money — and never will Ignore the hype, the demos and the endless talk about “agentic AI.” Just follow the money. * Adoption flattened nearly a year ago. * Ninety-five percent
By Alan Jacobson, Systems Architect & Analyst The Mag 7 tech stocks—Nvidia, Apple, Microsoft, Google, Amazon, Broadcom and Meta—account for 36% of the S&P’s market capitalization and a larger share of its profit growth. By sheer scale, almost all investors are already exposed to their
By Alan Jacobson, Systems Architect & Analyst Can LLMs “scale” their way to profitability by charging a flat rate of $20/month for unlimited compute? * If every man, woman and child in America pays $20/month = $81.6B/year * Capex depreciation: $50B/year * Staff: $11.25B/year * Cost of electricity
By Alan Jacobson, Systems Architect & Analyst First, AI must make dollars and sense in two ways: Before AI can be trusted, governed or widely adopted, it has to work as a business. Today it doesn’t. The economics are broken on both sides of the ledger: revenue is mispriced
By Alan Jacobson, Systems Architect & Analyst The first wave of AI adoption is over. Not because the technology failed, but because users hit its limits faster than the industry expected. Models forgot. Systems hallucinated. Outputs couldn’t be trusted. And when reliability drops, adoption at scale is imp. That
Let’s start at the very beginning. A very good place to start: There are only a handful of companies that actually provide AI at scale — the firms that own or control the data centers, power, chips and routing that make large models possible. Everyone else, no matter how famous
Generative AI materially alters our cost structure in ways that are not yet reflected in reported unit economics. Unlike traditional search, which benefits from near-zero marginal cost per query, large-scale AI inference incurs materially higher variable compute costs that scale with usage, verbosity, and reasoning depth rather than impressions or
On Tuesday, The Information reported that individual investors now have a way to invest directly in privately held, pure-play large language model companies such as OpenAI and Anthropic. Until now, public market exposure to LLMs came only through diversified companies with profitable core businesses — Google through advertising, Apple through hardware
Google is feeding the future. Apple is harvesting the past. Warren Buffett is no dummy. In 1984, Apple sold a simple idea: the computer as a “bicycle for the mind.” Steve Jobs meant something very specific by that. Humans are not especially efficient at moving — until they get on a
When you don’t have the numbers, you change the subject. If a platform has real adoption, it never shuts up about it. It leads with numbers, not narratives. Even The New Yorker is asking why AI failed to show up in everyday life. That’s what adoption looks like
NASDAQ up, Chipmakers up, LLM-tied names mixed or down. The explainer: This is not a selloff. It’s a separation. Why chipmakers are up * They sell capacity, not outcomes * Revenue tied to approved capex * Near-term visibility * Paid before value is proven Shovels get paid during the gold rush. Why LLMs
There are two ways to make money online: subscriptions and advertising. ChatGPT is failing at both: * According to Menlo Ventures, 95% of ChatGPT’s users pay nothing. * According to TheInformation, 90% of ChatGPT’s users are outside the US and Canada, but advertisers want American wallets So, if OpenAI cannot
The Mag 7 tech stocks—Nvidia, Apple, Microsoft, Google, Amazon, Broadcom and Meta—account for 36% of the S&P’s market capitalization and a larger share of its profit growth. By sheer scale, almost all investors are already exposed to their AI bets. Related: Jim Cramer Says “The
The tape is clear LLMs extended a five-day selloff, even as the NASDAQ was propped up by chipmakers. But flat adoption rates and pre-execute provisioning efficiencies point to falling demand for AI infrastructure. The past is clear From the 1873 railroad crash to the 2000 dot-com bubble, when capex races
A recent note from Jim Cramer framed the AI debate in familiar terms: power, electricity, and physical limits. One line stands out because it quietly carries the entire thesis: “If someone were to come up with a less energy-intensive way to produce compute, I would be very nervous. They haven’
By Alan Jacobson, Systems Architect & Analyst Jim Cramer doesn’t usually signal retreats. He signals rotations. So when he says “the year of magical investing is over,” he’s not forecasting a crash. He’s granting permission. * Permission to stop pretending. * Permission to stop chasing. * Permission to look at
By Alan Jacobson, Systems Architect & Analyst LLMs are not SaaS like… * Oracle databases * Microsoft’s Word+Excel+Powerpoint * Adobe’s Photoshop+InDesign+Illustrator+Acrobat Those products follow software economics. LLMs do not. For two decades, investors have believed in one rule: * Build once * Sell forever * Marginal costs fall toward