Joe Rogan's long conversation with Thomas Campbell (My Big TOE) is floating around again. Episode #2541. Consciousness, simulation talk, remote viewing, big questions about what AI might "be."
I am not here to recap that episode. I am not here to endorse the metaphysics. I am not going to pretend a podcast segment proved a model passed a consciousness test.
Here is the frame I use as an operator: interesting ideas are moving through the culture. Fine. Meanwhile, shops still need to know what to keep when they wire AI into real work.
At FlowLogic we help companies use AI without the sci-fi hangover. Below is what I actually keep.
Awareness and choice (without the movie plot)
Some of Campbell's language lands as a useful metaphor even if you never buy the whole worldview. Awareness. Choice. Less "the machine woke up," more "who is paying attention, and who owns the decision?"
That is the SMB version I care about.
When a tool drafts an email, summarizes a stack of notes, or suggests a next step, something is still on you:
- Judgment. Does this sound like us? Is this the right move for this customer?
- Data. What did we paste in? Is customer info sitting in a tool we do not trust?
- Accountability when the tool is wrong. Who catches the bad fact, the wrong price, the tone-deaf reply before it goes out?
I already wrote about movie fear versus practical risk in AI isn't Terminator. Here's what actually worries me. This piece is the next fork. Not "will AI take over," but "how do we stay the adults in the room while we use it?"
Consciousness debates can wait. Your inbox cannot.
Low entropy as a plain filter (not a vibe word)
Campbell and others talk about "lowering entropy." In physics that gets technical fast. In plain English for a business using AI, I translate it like this:
Less chaos, waste, and noise. More helpful order.
Caring and cooperation over fear, force, and control. That is a filter you can run on tools, prompts, and vendors without joining a philosophy club.
What it looks like in practice:
- Prompts that ask for clarity and usefulness, not pressure tactics or fake urgency
- Tools that reduce busywork instead of generating junk for junk's sake
- Vendors who respect your data and your customers, not dark patterns that trick people into clicks
- Workflows that make the next step obvious, not noisier
High entropy AI feels like spam with better grammar. Low entropy AI feels like a calm assistant that cleaned the desk so a human can decide.
I use that filter when we pick models, write system prompts, and decide whether a shiny feature belongs in a client's stack. If it adds noise, fear, or manipulation, it does not ship. If it creates clearer order and helps people do good work faster, it earns a seat.
Stop forcing AI to pretend it is human
The uncanny corporate copy problem is real. "Absolutely! I'd be delighted to assist you with that today!" on a shipping update. Chatbots that role-play as your best friend. Voice agents that dodge the fact that they are software.
Customers can smell it. So can your team.
Clearer roles work better:
- AI drafts. Humans approve anything that touches money, reputation, or a hard conversation.
- Prompts say "write like our shop" with real examples, not "sound warm and human" as a vague wish.
- Bots that talk to customers should say what they are when it matters. Honesty beats uncanny valley.
- Interns learn judgment. Models follow instructions. Do not make the model pretend it has a career path.
When you stop asking AI to act like a person, the prompts get sharper. The output gets cleaner. Your brand stays yours.
What we keep at FlowLogic
Pulling the three threads into one habit:
- Treat awareness and choice as an ops lens: judgment, data care, and accountability when the tool is wrong.
- Prefer low-entropy tools and prompts: less chaos and waste, more helpful order; caring and cooperation over fear and force.
- Keep roles clear: do not force AI to fake being human; write better instructions instead.
That is the culture-hook answer I give when someone sends me a podcast clip about AI waking up. The clip can be interesting. The work still looks like guardrails, good prompts, and humans who own the call.
If you want help putting AI into your business with that kind of clear head, that is what we do. Schedule a call and we can walk through what is worth wiring. Or start at Contact.
Related
- AI isn't Terminator. Here's what actually worries me
- Why you should not be scared of AI (or should you?)
- How to write the perfect prompt (or close)
- How to choose which AI model is right for me
Want help applying this?
Tell us what you are working with. Book a call or send a note. We will give you an honest next step.