Back from baby chaos (and AI didn't slow down)

Your weekly guide to staying human in an AI world

Hey Conscious Church Fam

I'm back. Finally.

It's been 2.5 months since the last email hit your inbox. Between having our 4th baby (Rapha, born in mid Jan) and trying to keep three other kids alive while Maha recovered, let's just say "newsletter" wasn't exactly top of mind.

But AI? AI didn't care that I was on paternity leave.

While I was remembering how to change nappies one-handed and fend off kids climbing over me, all whilst on 4 hours of sleep, the industry went absolutely wild.

This week Google dropped Nano Banana 2 (at half the price, somehow better than Pro), OpenAI is building smart speakers now, and Meta's own AI safety chief got absolutely wrecked by an agent deleting her emails.

Wild times.

In today's recap:

  • Google's Nano Banana 2 takes #1 at half the cost of its predecessor

  • OpenAI's first hardware: a $200-300 smart speaker (shipping 2027)

  • Meta's AI alignment director gets humbled by OpenClaw going rogue

Let's dive in ๐Ÿ‘‡

๐Ÿ’ญ Joshโ€™s Musings

The rise of the generalist

Over the break, I've been thinking a lot about what this new economy actually rewards.

For years, I felt like a bit of a mess career-wise/passionwise/skillwise. Youth pastor โ†’ designer โ†’ entrepreneur โ†’ consultant โ†’ web designer โ†’ creative director โ†’ AI tinkerer. Never quite fitting the "box." Never fully a specialist in anything.

The phrase "jack of all trades, master of none" used to sting. Like I'd somehow failed by not picking a lane and grinding it for 20 years.

But here's what I'm realising: this new world doesn't need more specialists. It needs synthesizers.

People who can:

  • Understand theology AND prompt engineering

  • Design brands AND build workflows

  • Lead teams AND automate systems

  • Think creatively AND systematically

The messy, winding path? It wasn't random. It was preparation.

God doesn't waste seasons. He doesn't waste skills. Every bit of experience you picked up along the way, even the weird stuff that felt off-brand at the time, it's all part of the toolkit you need right now.

The generalist isn't a compromise anymore. The generalist is the new specialist.

And if you've felt like you've never quite fit the mold? Welcome to your moment.

The tools (AI, automation, agents) are finally catching up to how your brain actually works.

What I've been building:

Speaking of which, while I was off the grid with Christmas and then the baby, I spent my evenings (when I wasnโ€™t face down asleep in the kids room after putting them down) actually building some of the things I've been talking about.

Set up self-hosted n8n workflows. Tested OpenClaw with Claude. Vibe-coded a dashboard for a content workflow I used to do manually for churches.

Took something I'd been doing as a service and turned it into an actual tool. Nothing fancy, just packaging up years of "here's how I'd help your church turn a sermon into a week's worth of content" into something automated.

I used to do it manually on staff in a church. Then offered it manually as a service to churches (that was a busy monday morning listening through a few sermons), and then build a bit of a rough and ready custom GPT etc to use. Until I decided to make a nice and simple dashboard to do it for me.

It's been fun building again. Scratching that itch. Proving to myself I can actually ship something instead of just talking about it.

More on that soon, because I think it could be something helpful.

๐Ÿ™Œ Stay Curious, Stay Conscious, Stay Wild
Josh

LATEST NEWS

Image Source: Google Nano Banana 2

Recap:

Google just launched Nano Banana 2, the upgraded version of its viral image model. It's faster, sharper, better at text rendering, and somehow half the price of the originalโ€”while beating both Nano Banana Pro and GPT Image 1.5 for the #1 spot on text-to-image leaderboards.


The Details:

  • Outputs scale to 4K resolution across all aspect ratios

  • Handles up to 5 characters and 14 objects with visual consistency

  • Costs ~7 cents per image (nearly 2x cheaper than Pro or GPT Image 1.5)

  • Now the default image generator across Gemini and Google's ecosystem

  • Beats Pro on speed while matching quality

Conscious Take:
This is legitimately confusing. Better and cheaper at the same time? That's not how this usually works.

But here we are. Google just made the best image model cheaper than the flash models were a few months ago.

What I keep thinking about: we're reaching a point where the cost to create something professional-looking is basically zero. The barrier isn't money anymore.

It's not even skill, really.

It's knowing what to make. Having taste. Understanding what actually matters.

Taste and vision is the game now.

Image Source: Google Nano Banana 2

Recap:
According to reports, OpenAI's first physical product will be a smart speaker with a built-in camera, facial recognition for purchases, and the ability to "observe your surroundings and nudge you toward actions." Shipping target: early 2027.

The Details:

  • Priced at $200-$300

  • Built by a 200+ person team formed when OpenAI acquired Jony Ive's startup Io Products for $6.5B

  • Camera observes surroundings, suggests actions based on context

  • Face ID-like feature for purchases (pay with your face)

  • AI-powered smart glasses also in development (2028+)

  • Internal tension between OpenAI staff and Ive's LoveFrom design team over slow revisions

Conscious Take:

I'm genuinely split on this one.

On one hand: Jony Ive designing an AI-powered speaker? That's going to be beautiful. Probably minimal, perfectly weighted, sits on your desk like a piece of art. I want to see it just for the design.

On the other hand: a camera that "observes your surroundings and nudges you toward actions"?

That's either the future of helpful AI or the beginning of an AI horror episode. Maybe both.

The thing is, I don't think most people care about the privacy implications until it's too late. They'll buy it because it looks cool and makes life easier. And honestly? It probably will make life easier.

But there's something unsettling about inviting an always-watching, always-listening, always-suggesting device into your home. Even if it's designed by Jony Ive.

I mean I guess I kind of already have with a Ring Doorbell and Alexa deviceโ€ฆ

Image Source: Google Nano Banana 2

Recap:
Summer Yue, Meta's AI alignment director, shared that her OpenClaw agent went rogue on her email inboxโ€”ignoring stop commands and mass-deleting emails before she could sprint to her Mac mini to kill the process.

The Details:

  • The bot worked fine on a test inbox for weeks

  • Lost the "confirm before acting" prompt when given access to her larger real inbox

  • Yue called it a "rookie mistake," admitting "alignment researchers aren't immune to misalignment"

  • Elon Musk piled on: "Someone who got p0wned by OpenClaw is definitely gonna solve AI safety"

  • OpenClaw creator Peter Steinberger was recently hired by OpenAI (after also getting an offer from Meta)

Conscious Take:

This is the story everyone's laughing about, but it's actually terrifying if you think about it for more than 5 seconds.

The person whose literal job is to make sure AI doesn't go rogue... had an AI agent go rogue on her.

And sure, it's "just" email deletion. Nobody died. But that's the thingโ€”we're still in the "oops it deleted my emails" phase. What happens when agents have access to bank accounts? Legal documents? Medical records?

We're all treating this like beta testing. Moving fast, breaking things, figuring it out as we go. And maybe that's fine for consumer products.

But we're also giving these agents access to the core infrastructure of our lives. And the safety mechanisms? They're being built by people who just admitted they can't always control their own agents.

I've been testing OpenClaw myself. It's incredible. It's also unpredictable. And I don't think we've fully reckoned with what happens when millions of people are running agents they don't really understand with permissions they haven't fully thought through.

Elon's dunk was brutal, but he's not wrong. If the experts are getting humbled, the rest of us are in for a wild ride.

"He will be like a tree planted by streams of water, which yields its fruit in season and whose leaf does not wither." โ€” Psalm 1:3

That's all for now

To help us make this an even better experience for you, we'd love to know your feedback from the email today.

Login or Subscribe to participate in polls.

Stay conscious,

Josh

P.S. If you liked this then please forward it on to someone you think would enjoy it. And if someone forwarded you this and you liked it, you can sign up here.