My father-in-law was in town. Somewhere between coffee and catching up, AI came up. He asked the question a lot of people ask when the news cycle gets loud: do I think AI is going to take over humans like Terminator?
I laughed, then I answered him straight.
I have plenty of fears about AI. A sentient machine turning an eye on humans and wiping us out is not one of them.
What I do fear looks a lot less like a movie trailer and a lot more like the evening news. Foreign military forces using AI against power grids and other critical infrastructure. Drone warfare getting smarter, faster, and harder to defend against. Those sit in the real-fear bucket for me.
This post is the split I wish more people made. What we should not freeze over, and what we should take seriously.
What I'm not scared of (and why)
I am not waiting for Skynet. I do not lose sleep over a model suddenly deciding humans are the problem and launching the nukes.
That is not me dismissing people who worry about long-term risks. Serious researchers argue about alignment, misuse, and what happens if these systems keep scaling. Fair conversation. I am saying the Hollywood version (conscious killer AI with a grudge) is not how I spend my worry budget day to day.
What I see in the work I do is different. Models that are powerful, useful, and sometimes weirdly literal. Tools that draft, sort, summarize, and speed up real jobs. They still need humans for judgment, taste, and "should we even do this?" They do not wake up angry. They follow instructions, train on data, and fail in boring ways when the prompt is vague or the workflow is sloppy.
If you want the business-side version of "scared of AI" (looking dumb, picking the wrong tools, freezing on hype), I already wrote that: Why you should not be scared of AI (or should you?). This piece is the other fork. Movie takeover versus practical and geopolitical risk.
What I am scared of
Foreign military AI against the grid (and similar targets)
This is the one I named at the dinner table. Not robots hunting people in the street. Systems pointed at infrastructure.
Power grids, water, communications, logistics. Soft spots that make everyday life hard fast. I am not claiming a specific attack timeline, a specific country's plan, or classified details I do not have. I am saying the direction of travel is obvious enough for a regular person to take seriously: AI makes targeting, planning, and cyber-adjacent operations cheaper and faster for whoever already wants to do harm.
For Texas owners and families, that fear is not abstract. We already know what a bad week of outages feels like. Add an adversary who can probe and adapt at machine speed, and resilience stops being a buzzword. It becomes backups, vendors you trust, and not pretending "someone else handles security."
Drone warfare
Drones are not sci-fi. They are in active conflicts right now. AI layered on top (better targeting support, swarm coordination, faster decision loops) raises the stakes. Again, I am not inventing casualty counts or claiming secret programs. Public reporting already shows how cheap, numerous, and hard-to-stop unmanned systems can be.
That is military and geopolitical. It is also a reminder that "AI risk" is not one blob. Some risks are existential movie plots. Some are nation-state and battlefield realities that civilians feel through energy, supply chains, and insurance long before they feel a robot uprising.
Adjacent fears that still count (without diluting the point)
A few other worries sit next to the main ones for me:
- Deepfakes and fraud that make it harder to trust a voice, a video, or a "CEO email"
- Rushed automation that removes a human from money, safety, or reputation decisions
- Shops that paste sensitive customer data into random free tools because "everyone is doing AI now"
Those are not Terminator. They are Tuesday. We help companies avoid the Tuesday mistakes while they put AI to work.
Why the distinction matters if you run a business (or just live here)
If you treat every AI headline like the end of humanity, you freeze. You skip useful tools. Your competitors quietly draft faster, follow up faster, and free humans for the work that still needs a person. Related: What happens when your competitors use AI and you do not and The quiet cost of running a business without AI in the workflow.
If you treat every AI tool like harmless magic, you get naive. Bad data habits. No judgment layer. No thought about outages, vendors, or what happens when the internet hiccups.
The useful middle is boring and adult:
- Use AI where it cuts admin and speeds decisions you still own
- Keep humans on calls that touch money, safety, and reputation
- Take security and resilience seriously (backups, access, what you paste where)
- Separate sci-fi panic from risks that actually show up in ops and geopolitics
At FlowLogic we build and advise for Texas SMBs and teams that want practical AI, not theater. Separating movie fear from real risk is part of that. It is how people adopt tools without freezing, and without pretending nothing can go wrong.
Eyes open, then get to work
So: scared of a conscious killer AI hunting humans? That is not my fear. Scared of military AI aimed at grids and infrastructure, and of drone warfare evolving in public view? Yes. Those worry me. Uncertainty on timelines and exact capabilities is honest. Pretending those categories are the same as Skynet is not.
If you want help putting AI into your business with guardrails and a clear head, that is the work we do. Schedule a call and we will walk through what is worth wiring and what is noise. Or start at Contact.
Related
- Why you should not be scared of AI (or should you?)
- What happens when your competitors use AI and you do not
- The quiet cost of running a business without AI in the workflow
- 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.