🚨Perplexity is paying Snap $400 million to integrate its AI search engine into Snapchat, exposing how AI startups—even those with superior technology—must spend hundreds of millions just to access distribution channels. This marks a fundamental shift in the AI economy where access to users costs more than building the tech itself.

Someone asked me last week: "What's the best AI tool for my business?"

My response: "What does your business actually do, step by step?"

Blank stare.

This is where 90% of "AI implementation" efforts die. Not from bad tools. From trying to automate workflows that aren't documented or understood.

Personal AI tools only work when you understand your workflow first.

The mistake people make is jumping straight to tools. They hear about Cursor or Bolt or some new AI thing and think "I should use that" before asking "What am I trying to improve?"

The companies winning right now are doing the opposite sequence...

They're mapping workflows first. Documenting every step of how work actually happens. Where data lives. Who does what. What the handoffs look like.

Then they look at that map and ask "Where would AI create the most leverage?"

Usually it's not where you'd guess.

One company spent weeks planning an AI customer service bot. After mapping their workflow, they realized the real bottleneck was sales reps manually pulling data from 4 different systems to create proposals.

They built a tool that consolidated that data. Took 2 days with Lovable. Saved 8 hours per proposal.

The "AI agent chatbot" they almost built? Would've taken months and solved a symptom, not the root cause.

Pick one recurring task. Document every single step, no matter how small. List what information you need at each step and where it comes from. Mark which steps require human judgment vs. which are pure process.

Any step that's pure process and happens the same way every time? Prime candidate for AI assistance.

Any step that requires context or judgment? That's where you stay in the loop and AI becomes a copilot.

The trap is trying to automate things that aren't standardized yet. If your process changes every time, AI can't help -- you need to stabilize the process first.

What's next: Workflow mapping tools are adding AI features that watch you work and auto-document your process. Slightly creepy, massively useful. Grain and Scribe are both testing versions of this.

BTW: The Pareto principle applies viciously here. 20% of your workflows probably create 80% of your friction. Map those first. The exotic edge cases can wait.

There is no better time to learn how to use AI to build solutions businesses want. I’m hosting a workshop that will show you, step-by-step, how to build tools and solutions using AI without having to know how to code.

If you’re interested, reply back to this email with the keyword BUILD and we’ll send you more info.

We are disclosing the dates soon. So if you’re interested make it known.

The Recap: Apple is finalizing a deal to pay Google approximately $1 billion annually for a customized version of Gemini to power a completely rebuilt Siri. The agreement reveals how Big Tech's AI infrastructure dependencies are evolving into mega-deals worth billions, fundamentally reshaping competitive dynamics in the industry.

Unpacked:

Bottom line: Apple's willingness to pay Google $1 billion annually for AI infrastructure signals that vertical integration in AI is harder than anticipated, even for the world's most valuable company. The real question isn't whether Apple can afford these payments—it's whether repeatedly outsourcing core AI capabilities to competitors creates an insurmountable strategic disadvantage as AI becomes central to device value propositions. This deal suggests we're entering an era where AI infrastructure ownership, not just access, determines long-term competitive positioning.

Pinterest CEO credits open-source AI for 'tremendous performance' with reduced costs, signaling a major platform shift away from proprietary models.

Australia's solar boom forces utilities to give customers free electricity three hours daily—showing what energy abundance actually looks like.

Lina Khan co-chairing NYC mayor-elect's transition team signals regulatory pressure on Big Tech intensifying at the city level.

Replika founder raises $20M pre-seed for 'YouTube of apps' where anyone can prompt-create mini-apps, testing no-code's viral potential.

Tinder testing AI that scans your camera roll to find matches reveals how personal data becomes the new dating currency.

Ikea launches 21 Matter-compatible smart home devices starting at $3, potentially making interoperable smart homes mass-market reality.

Cluely's founder won't share metrics four months after viral growth claims, exposing the gap between social hype and actual business.

Tesla shareholders vote again on making Musk a trillionaire through compensation, testing how much value boards will transfer to celebrity CEOs.

P.S. Reply back to this email and let us know what area of AI you are struggling with the most. We are going to be hosting some free trainings and want your input.

Best of AI™ Team

Was this email forwarded to you? Sign up here.