Microsoft just launched a $99/user/month AI mega-bundle. Salesforce is weaving AI into every tier. The playbook is familiar — bundle everything together so leaving feels impossible. But the same market panic driving these bundles is actually the best thing that's happened to small businesses in years. If you don't lock yourself into someone else's survival strategy.

My least favorite part of financial modeling isn't the analysis — it's the handoff tax. Pull numbers in Excel, manually re-explain context in the deck, then do it again for the email. It's the same data, three times, with three different tools that have no idea what each other just did. That's the thing Claude for Excel and PowerPoint is actually solving here.

So AI was supposed to give us our afternoons back. Instead, workers are spending twice as long on email and messaging, deep focus time is down, and the productivity gains are just... immediately converted into more work. The efficiency hamster wheel is real, and apparently we built a faster wheel.

We've been calling clicks a proxy for influence for 20 years while knowing, deep down, they're not. This article makes the uncomfortable case that click-through attribution captures less than 0.5% of the actual B2B buying journey. That stat alone should ruin your Wednesday.

The real signal here isn't "AI is good, actually." It's that structured role change beats ambiguous role change. High-maturity teams formalize AI ownership — someone owns prompts, pipelines, governance. Low-maturity teams use AI for one-off tasks and wonder why nothing scales. If your team is still in ad-hoc mode, the bottleneck isn't the tools. It's the org design around them.

Perplexity's enterprise play is essentially a bet that orchestration beats ownership. Instead of building one model, they route tasks to whichever model wins at each job. If that holds, the real threat isn't to Microsoft or Salesforce directly — it's to how much time employees actually spend inside those platforms. Watch your own vendor stack: anything that's primarily an interface layer for data you own is exposed. The "gap-filler" framing is diplomatic cover for what eventually becomes displacement.

The DEI conversation got politically inconvenient, so most of tech went quiet. But el Kaliouby's point isn't about fairness — it's about who captures the wealth from the biggest platform shift in a generation. Sitting that out isn't neutral. It's a choice with a price tag.

Every AI rollout I've seen has the same shape: excited executives, curious ICs, and a layer of managers in the middle quietly wondering what they're supposed to do with any of this. Turns out that awkward silence is where adoption goes to die.