Alright, let’s talk AI.

This week’s mix hits both sides of the brain, leadership, leverage, and a little chaos. We’re talking after-hours calls (hustle or harassment?), a new map showing where AI actually delivers results, and a Cloud Employee who somehow got mistaken for a dinner date. Oh, and a few tools that’ll make your analyst, your editor, and maybe even your memory obsolete.

Let’s get into it.

No, But Seriously…

After-Hours Calls: Hustle or Harassment?

Honest question: Is a phone call from your boss after 5pm hustle culture… or harassment?

Half the workforce yells “boundaries!” The other half yells “soft!”

I’ll be real, I’ve made those calls. Sometimes 9. Sometimes 10. If it mattered, I dialed.
If it didn’t matter… sometimes I still dialed.

And yeah, a few people acted like I’d just kicked their dog.

I’ve tried to chill. Schedule emails. Drop Slack notes. Save calls for “urgent.”
But the truth is, I’ve never fully stopped.

Because I’ve always believed great companies aren’t built between 9 and 5. They’re built in those weird 9:47 p.m. moments when something clicks and someone cares enough to answer.

But maybe I’m the dinosaur here.

So tell me, when your phone rings after hours… do you answer, or do you file a complaint? Answer here.

The Drop

The Map of Where AI Actually Works

Everyone says they’re “experimenting with AI.” But zoom out, and the pattern’s obvious.

Some functions are transforming. Others are still pretending. I built a simple 2×2 to explain it: four zones, one reality check.

The Power-Up Zone – AI doesn’t replace talent, it rewires how talent creates value.

Marketing, sales, and product teams are finding a new gear—where human instinct meets machine speed. The best don’t hand work off to AI, they jam with it. They move from ideas to execution before most teams finish a brainstorm.

The wall between strategy, story, and speed is gone. In this zone, it’s all one motion—create, test, learn, repeat.

The Takeover Zone – AI isn’t helping, it’s doing the job.

This is where automation stops being incremental and starts being structural. Customer service, sales development, recruiting, and engineering are already scaling faster than humans ever could.

  • Customer Service: Most companies deflect 30–50% of tickets, but I’ve got clients hitting 57%+ fully handled by AI. That’s not hype, that’s headcount.

  • Sales Development: Inbound AI SDRs deliver near-human conversion at 70–85% lower cost per meeting. Outbound’s still messy, but it’s closing in.

  • Engineering: Half of new code is now machine-generated, 70%+ for boilerplate. The new skill isn’t syntax, it’s supervision.

  • Recruiting: Companies are quietly cutting 25–40% of recruiter headcount as AI screens, schedules, and nurtures automatically. The high-volume recruiter role is on borrowed time.

The Warm-Up Zone – Structured work, slow to trust AI, but not for long.

Finance, ops, and HR aren’t blind, they see the leverage coming. But trust still lags. Too many systems, too much risk, not enough control. The tension isn’t about if AI fits here, it’s about when it feels safe to let it in.

What happens when safety no longer outweighs speed? This zone flips overnight, and efficiency stops being a project, it becomes the baseline.

The Human Zone – Where instinct still outruns automation.

Leadership and strategy still run on intuition, the kind you earn from pattern, pain, and pressure. AI can model outcomes, but it can’t want one. It doesn’t choose a direction, it optimizes one. The question isn’t whether AI will lead. It’s whether leaders will still know how to, once everything else can think.

AI isn’t evenly distributed across the org chart. But every company will move through these four zones, some faster, some by force.

AI will power up the creative, take over the repetitive, and pressure-test the rest. But the Human Zone still matters. Because when everything else can execute, the only real advantage left is who decides what winning looks like.

The Shortcut

When Your AI Gets Hit On, You Know It’s Working

We just ran one of the funniest AI experiments I’ve seen yet. Meet Betty, our event follow-up Cloud Employee for Company XYZ.

Biggest trade show of the year, 750 leads to follow up on, and the client says: “Email only. And make her feel like a real person.”

Why not? Agentic email is new, so we gave it a shot.

Friday we kicked off. First email goes out…Result? dozens of back-and-forth conversations. 14 meetings booked. Off one send.

But here’s where it gets wild:

▸ One prospect replies, “I think I saw you at the restaurant last night.”

(Remember… Betty isn’t real.)

▸ And Betty? She just rolls with it: “Great to see you too!”

Hahaha, this was just an experiment...I don’t recommend tricking people. But…

This is where AI gets interesting. Not the shiny demos. Not the jargon bingo.

It’s when a buyer mistakes your Cloud Employee for the person they split mozzarella sticks with last night :)

If your AI can’t survive a random restaurant encounter, it’s not ready for revenue 😂😂😂

Tools I’d actually use

Quick Hits

Julius
Julius is your data team without the sighs. Ask it, “Which customers are about to bail?” and it rips through your spreadsheets, builds a chart, and emails the report before your analyst finds the login.

Taya
This looks like jewelry, but it’s actually your memory on a chain. It records your conversations, transcribes them, and turns your life into a searchable journal. Finally, something that listens better than your coworkers.

Mosaic
This takes your unedited chaos and turns it into content gold. One upload, one prompt, and it slices your podcast into reels so clean your editor might start sweating.

The news that matters

The Weekly Wire

  • The End of Outsourcing: AI is rewriting the global labor map. The offshore model, call centers, sales pods, back offices, is collapsing. The work isn’t coming back to humans, it’s coming back to algorithms. As Hemant Taneja says, globalization changed where things were made. AI is changing who does the making.

  • Salesforce Loses Its Vibe: Salesforce dropped Agentforce Vibes, but it feels more like flailing than focus. After layoffs, credit pricing, and fading traction, “vibe coding” looks like another buzzword in search of a win. The real question: can legacy SaaS survive the Autonomous Era or is it already being replaced?

  • OpenAI Goes Proactive: OpenAI’s new Pulse feed researches for you overnight and delivers a personalized daily briefing. It’s AI turning proactive, chasing your first click of the day and your morning routine.

  • The New Venture Math: The old “triple-triple-double-double” playbook is dead. AI hype and trillion-dollar funds now expect startups to jump from $1M to $100M fast. But not everyone needs to chase that pace — profit and durability still win for founders who pick their own lane.

That’s it for today.

Catch the latest Talk AI newsletter editions here, and tune in to Talk AI Weekly for conversations with some amazing guests here.

Want to support the newsletter? Check out today’s sponsor: GetSignals Logo

See you next time.

Reply

or to participate

Keep Reading

No posts found