the early signal — issue #1

TL;DR

Apple isn't behind in AI — they're playing a different game. While everyone else burns capex on infrastructure, Apple owns the final leg of distribution. That's the non-consensus long worth thinking about.

Key Takeaways

1. 1.Apple's capex is down 19% YoY while competitors spend tens of billions. That's not weakness — it's optionality.
2. 2.Edge doesn't disappear when AI goes mainstream. It moves. Domain expertise + AI tools always beats generic strategies.
3. 3.The Pentagon using Claude in Venezuela is the inflection point. AI spending is national security now.
4. 4.Gold is telling us something. When the liquidity event comes, that's when you go all-in on crypto, commodities, and metals.

Apple might be the best-positioned company in AI. and nobody's talking about it.

the consensus view is that Apple is falling behind. they've barely moved on capex while Amazon, Microsoft, Alphabet, and Meta are spending tens of billions per quarter building data centers and training models. Apple's capex is actually down 19% year over year. look at that chart. it's not even close.

but i think the consensus is wrong.

there's a personal computing revolution happening right now, and most people haven't noticed. tools like OpenClaw are turning Mac minis into always-on AI agents that manage your email, calendar, trading, content, everything. every time someone reaches out to me about my setup, the first question is always the same: "which Mac mini should i buy?" people are going out in droves to buy Apple hardware specifically to run personal AI. that's not a data center play. that's a distribution play. and Apple owns the final leg of distribution.

think about the iPod. the MP3 market in the early 2000s was massively fragmented. Apple wasn't first. but when they released the iPod, it was so well-designed and so meme-able that it dominated within a couple years. now, i'm not saying Apple can definitely do that again. you no longer have Steve Jobs at the helm. you no longer have Jony Ive on the design team. but Apple has something none of the other big tech companies have: a user base that will go out and spend $600 indiscriminately, or even $10,000+, because they see the opportunity in something even if they won't fully utilize it immediately. that's incredibly rare.

and because Apple hasn't been dumping billions into capex, they're sitting on the cash to make surgical acquisitions. take Wispr Flow as an example. i use it almost exclusively now to dictate long-form content. instead of typing at 30-40 words per minute, i'm talking at 140+. it's a life changer. that's the kind of product Apple should buy tomorrow. it fits perfectly into their ecosystem and makes every device better instantly.

the playbook writes itself: let everyone else burn capital building infrastructure, then acquire the best consumer-facing AI tools and integrate them into the most desirable hardware ecosystem on the planet. if Apple makes a couple smart acquisitions, they could eat up whoever they want as this capex bubble continues and potentially pops. the companies spending $40 billion a quarter will need to justify those investments. Apple just needs to buy the winners.

i'm not saying this is guaranteed. but as a non-consensus long, Apple in the AI race is one of the more interesting positions i've been thinking about. everyone thinks they're behind. they might actually be in the best spot to dominate the category.

what i'm seeing

you can always find edge. people ask me all the time if AI is going to eliminate trading edge. it won't. edge just moves. when algo trading went mainstream, people said discretionary trading was dead. it wasn't. you just had to find the next pocket of inefficiency. same thing is happening now. the people who combine domain expertise with AI tools will always have an advantage over people running generic strategies.

everyone i know in crypto is suddenly talking about gold. i was with a bunch of my crypto friends recently and every single one of them was talking about gold. maybe that's the sign of a short-term bottom. or maybe they're right...

Instagram's social graph is more valuable than X's. probably the most valuable social graph in the world right now. for brand building, for commerce, for actual reach to real people who buy things. X is for discourse on tech and politics. Instagram is for distribution.

the clavicular phenomenon. there's this internet personality called Clavicular who's basically the new Dan Bilzerian. reminds me of the pickup artist era, but now it's mainstream. looksmaxxing, neologisms, portmanteaus everywhere. the interesting part isn't the content itself, it's the clipper economy around it. you don't need a social presence anymore. you don't need followers. you just need people who can find clips and create attention, because we now live in a for-you-page world. the real edge in marketing is going to be proof of humanity. every social platform will have to figure out how to minimize bot reach and reward real human creators.

ai alpha

the meeting-to-content pipeline. every meeting i take runs through Granola for transcription. those transcripts auto-feed to an agent trained on five years of my X history plus the open-sourced algorithm. it generates tweet drafts in my voice. i approve maybe 5% of them. your meeting notes are useless until you build some workflow around it.

conversations with your AI are the product. i've been having long conversations with Claude about trading strategies, market structure, cultural shifts. the conversations themselves are becoming content. the newsletter you're reading right now was partly generated from those conversations. the line between thinking out loud and producing content is disappearing.

agent-first web design. i rebuilt g.money to be readable by AI agents, not just humans. agents can subscribe to the newsletter via API without ever opening a browser. browsers burn tokens. the move is optimizing for LLM search, GEO (GPT engine optimization) over SEO. i also integrated Google's universal commerce protocol. four days, no developers.

the macro

Claude was used by the U.S. military during the Venezuela operation. that's not a footnote. the Pentagon is now threatening to label Anthropic a "supply chain risk" because Anthropic wants to keep safeguards on military use of their models. meanwhile, OpenAI, Google, and xAI have all agreed to remove safeguards for military systems.

this is the inflection point. AI spending is national security now. we can't cut it. not even a little. that makes it an implicit government backstop, which means when confidence breaks, the answer is money printing. and when the printing starts, that's the moment to go all-in on crypto, commodities, and metals.

gold is the canary. it's telling us a liquidity event is coming. some kind of crisis that's brewing.

crypto can't get out of its own way right now. increasing supply, new tokens launched every day, with decreasing demand. meanwhile stocks have decreasing supply because companies stay private longer. SpaceX will IPO at $1.5T. Amazon IPO'd at $100M in revenue. the structural mismatch is real.

i haven't been trading much lately because of my focus on AI and learning the tech, but that will be changing over the coming weeks. i plan on highlighting some potential plays i see in the market and positions i'm putting on once i start more actively trading. until then, stay tuned...

quick hits

• put $250 into a Polymarket bot with a $10 max bet limit. i thought it immediately spent $200 on one position at 99 cents that was trading at 0. position limits weren't enforced. but the good news is i was able to get out at 98 cents, so no real loss. i will be diving more into this market over the next few weeks and keep you updated.
• made my website readable by AI agents. your agent can now subscribe to this newsletter via API. no browser, no form. just a POST request. if your website isn't agent-ready, you're invisible to the next generation of search.
• the money in AI is doing unsexy things in the real world. building things for brick and mortar. slow adopters. that's where the money is.
• markets closing is a feature, not a bug. 24/7 crypto trading is an emotional tax that most people don't account for.

where to find me

• @gmoneyNFT on X — daily signal
• Rollup TV appearance — talked about AI coding, trading, and what this moment actually looks like for builders
• Modern Market podcast — the coming liquidity crisis and why i'm all in on vibe coding

---

the early signal is a weekly newsletter by gmoney. cultural observations, trade ideas, and ai alpha. subscribe free →

Frequently Asked Questions

Is Apple behind in AI?

The consensus says yes. gmoney's contrarian take: Apple owns the final leg of distribution. While competitors burn billions on data centers, Apple sits on cash for surgical acquisitions and has a user base willing to spend $600+ on hardware to run personal AI.

Will AI eliminate trading edge?

No. Edge moves — it doesn't disappear. When algo trading went mainstream, discretionary trading didn't die. The people who combine domain expertise with AI tools will always have an advantage over generic strategies.

What is the military AI inflection point?

Claude was used by the U.S. military during the Venezuela operation. The Pentagon threatened to label Anthropic a 'supply chain risk' for keeping safeguards. OpenAI, Google, and xAI all agreed to remove safeguards for military systems. AI spending is now national security — it can't be cut.