Nvidia just reported earnings and... yeah. $68.1 billion in quarterly revenue. Up 73% year over year. Data center revenue alone hit $62.3 billion—that's up 75% from last year. They beat Wall Street estimates on every single metric.

But here's what actually matters: they shipped the first Vera Rubin samples to customers this week. That's their next-gen AI superchip—72 GPUs, 36 CPUs, 1.3 million components in a single system. It delivers 10x more performance per watt than Blackwell. Oh, and it cuts inference costs to roughly one-tenth of what the current flagship charges.

Meanwhile, combined capex from the big four hyperscalers (Alphabet, Amazon, Meta, Microsoft) could approach $700 billion this year. That's not a typo. Seven. Hundred. Billion. And more than half of Nvidia's data center revenue comes from those four companies alone.

The AI infrastructure arms race isn't slowing down. It's accelerating.

On February 9th, OpenAI started showing ads inside ChatGPT for U.S. users on Free and Go tiers. Brands like Expedia, Mercedes-Benz, Adobe, and Best Buy are already in.

The numbers: $60 CPM (that's 3x what Meta charges). $200,000 minimum to get in the door. Only enterprise brands through major agency holdcos like Omnicom, WPP, and Dentsu are in the beta.

What makes this wild: Ads can appear after your very first message. Before ChatGPT even has context about what you want. They're using your conversation topics, past chats, and ad interactions to serve them. If you have memory enabled, the AI can reference your stored memories when picking which ad to show you.

Why it matters: This is the moment AI search officially became an ad platform. And it creates a fascinating split in the market: OpenAI said yes to ads. Perplexity abandoned ads entirely last month. Anthropic says Claude won't carry ads. Three AI platforms, three completely different bets on how to make money. OpenAI is projecting roughly $1 billion in revenue from free-user monetization this year alone.

Darren Mowry, who runs Google's global startup organization across Cloud, DeepMind, and Alphabet, went on TechCrunch's Equity podcast and basically said two entire categories of AI companies have their "check engine light" on.

The two dead models:

LLM Wrappers — startups that slap a UI on top of GPT, Claude, or Gemini and call it a product. Mowry's exact take: the industry has lost patience with companies that white-label someone else's model with thin IP on top.

AI Aggregators — companies that route queries across multiple LLMs through one interface. Mowry says this is already being commoditized by Azure AI, Amazon Bedrock, and Google's own tools.

Why it matters: Here's the brutal context. 17 U.S. AI companies raised $100M+ in the first 49 days of 2026. Four days later, Mowry essentially declared many of their business models dead on arrival. Seed rounds that raised $15M in 2023 are closing at $150M now—and the justification is still just "we use GPT better than anyone else." In 2026, that's a red flag, not a pitch.

The survivors: Companies like Cursor (coding) and Harvey AI (legal) that built deep vertical moats with proprietary data. If you're building an AI tool, the lesson is clear: own something the model can't replicate.

LinkedIn quietly dropped a bombshell: non-brand, awareness-driven B2B traffic declined by up to 60% across key topics. Rankings stayed stable. Click-through rates cratered anyway.

Why? Google's AI Overviews are answering questions before anyone clicks. ChatGPT resolves queries inside the interface. People are getting answers without ever visiting the source. LinkedIn said the old model of "search, click, website" is dead. Their new framework: "Be seen, be mentioned, be considered, be chosen."

What they did about it: LinkedIn formed a cross-functional AI Search Taskforce and abandoned traditional SEO metrics entirely. They're now measuring visibility, mentions, and citations inside AI-generated responses instead of traffic and clicks.

Why it matters for you: Gartner projects traditional search engine volume will decline 25% by 2026. If you're running a business and your entire acquisition strategy depends on people clicking links from Google... you're watching it evaporate in real time. The new game is getting your brand mentioned IN the AI response. Not below it. Not linked from it. In it.

Everyone's mourning the death of AI wrapper startups. I'm celebrating.

Here's why: when VCs were throwing money at any startup that could spell "GPT," it created a flood of mediocre tools. A new AI writing assistant every week. A new "ChatGPT for [industry]" every day. None of them solving real problems.

Now the VC money is drying up for lazy AI plays. The survivors will be companies with actual domain expertise, proprietary data, and deep vertical knowledge.

And who has domain expertise? Who has deep knowledge of their industry? Solopreneurs and small business owners. The exact people who've been told they can't compete with funded startups.

The playing field just got leveled. Build something specific. Own your niche. Let the wrapper startups burn while you build something real.

LinkedIn just admitted they lost 60% of their B2B traffic to AI search. They're one of the most sophisticated marketing organizations on the planet and they had to form an emergency task force to figure out what to do.

Now imagine how screwed your average small business is.

Here's the play: Start a consultancy that helps businesses get mentioned in AI responses. Not SEO. Not content marketing. AI Visibility Optimization.

What you'd actually do:

Audit how AI platforms (ChatGPT, Perplexity, Claude, Google AI) currently describe the client's business and competitors. Restructure their website content so AI systems can easily extract and cite it—clear headings, semantic HTML, expert credentials, timestamps. Build a "mention monitoring" system that tracks when and how the brand appears in AI-generated answers. Create authoritative content specifically designed to be cited by LLMs, not clicked by humans.

Why it works: Every business that currently pays for SEO will need this within 12 months. The agencies that figure this out first will own the category. There's barely anyone doing this right now because the problem is only 6 months old at scale.

Gartner proved the demand is coming — 25% search volume decline by end of 2026. LinkedIn proved it's already here. Somebody needs to sell the solution.

That's the briefing. Now go build something.

— Michael

P.S. Nvidia just printed $68 billion in a single quarter and people are still asking if the AI boom is real. It's real. The question is whether you're building on top of it or watching from the sidelines. If you need help figuring out where AI fits in your business, book a Work-Life Balance Audit — 60 minutes, $25, and you'll leave with a clear roadmap. Reply to this email if you've got questions. I read everything.