Hello, I'm Buono. went with a dramatic title but bear with me lol
the dream vs reality
how do you live in the AI era? is your role getting replaced? I think about this constantly and I know a lot of you do too.
I switched careers to become an IT engineer at 40. literally chased the dream of working remotely from anywhere. location independence, freedom, the whole thing.
finally got the job... and then AI hit. already getting swept up in the chaos.
I haven't written a single line of code
not one. recently at least.
all I do now is write instructions. "do this, make that work like this." I don't even properly review the code anymore. if it passes the tests, ship it. done.
is this really ok?? but the reality is... the work gets done. and it gets done well. so what was the point of becoming an engineer...
the tasks I'm completing are stuff I could NEVER have done on my own without AI. but once you actually pull it off, this weird emptiness hits you.
I built an entire app in 2 weeks (while working my day job)
on the side I recently built Pikutan — a pictogram-based English vocab learning app.
we're talking:
- Google auth ✅
- payment system ✅
- learning analytics with charts ✅
- deployed to cloud ✅
stuff that would've taken me a full year 10 years ago. built it in 2 weeks while working my day job. all by giving AI instructions.
first reaction was like "holy shit the payment system actually works?!" I was genuinely hyped. excited about what AI could do.
but then it was done. and that emptiness came back again...
the void no one talks about
I'm technically at the engineering level I dreamed about when I was 30 (with AI doing the heavy lifting obv). so what is this void??
been sitting with that feeling for weeks. couldn't figure it out.
then it clicked.
anything AI can do easily is going to zero
auth? payments? complex backend logic? these used to be impressive, hard-won skills. now AI handles them.
if AI can do it = worthless as a standalone skill. I'll say it.
these were actually difficult problems in software engineering. and if AI is casually solving those now... you already know where this is going.
so here's what I'm doing about it
I'm stepping away from the pure software world.
not quitting my job or anything — IT engineering still has strong demand and pays the bills. but in my own time? I'm going all in on things AI literally cannot do.
physical things. creative things. stuff that requires a human in the loop.
I'm getting back into electronics and YouTube. building real things with my hands. creating content that's inherently human.
I'm pretty sure that's what compounds over the next 5-10 years.
the takeaway
the skills that feel safe right now might not be safe for long. and the things that feel niche or "not scalable" — working with hardware, creating authentic content, building in the physical world — those might be the most valuable skills of the next decade.
if you're not thinking about where you'll be in 5-10 years and acting on it now, you're sleeping on it.
start today. seriously.
catch you later ✌️
Top comments (15)
I get the existential dread, man, but here is the reality check: businesses don't just pay for code, they pay for liability.
Imagine letting an AI fully control your payment gateways. One hallucination later, it transfers $1,000,000 from the company fund to random users. Who does the business sue? You can't sue an LLM. Businesses will ALWAYS need a human to take the responsibility. Our jobs aren't disappearing; they are just shifting to babysitting and auditing.
If an AI spits out an auth system in 10 seconds, I guarantee it has security holes. You still need your years of engineering knowledge to read that generated code, spot the vulnerabilities, and rebuild it safely. AI is just a fast intern that constantly lies. You are the senior engineer keeping it from burning the company to the ground)
Hardware is a cool pivot, but don't discount your software knowledge. The industry needs human supervisors for these bots more than ever.
Thank you for sharing your thought!
Fully agreed with your point because I actually saw same security holes as well.
Human supervisor is a good expression :)
This hit hard — especially the part about building something impressive in two weeks and feeling... empty instead of proud. I've been there.
I think what you're naming is the difference between execution and judgment. AI can execute now — write the code, wire up the payments, deploy the app. But it can't tell you why one font feels trustworthy and another feels cheap. It can't explain why a rounded corner makes an app feel "friendly" vs "unprofessional." That stuff still lives in humans.
I've felt this building FontPreview.online . The tools themselves are just code — AI could probably rebuild them. But the curation — the 100 font pairings I hand-selected, the psychological explanations, the brand archetype matching — that's not something AI can generate without a human steering it. It's the judgment layer.
Your pivot toward physical/creative work makes total sense. Electronics, YouTube, building with your hands — those have constraints AI can't fake. The texture of real things, the authenticity of a human voice, the serendipity of analog mistakes.
Maybe the next decade isn't about competing with AI, but about doing the things AI makes us want to do more of. For you, that's electronics. For me, it's helping designers make better decisions. For someone else, it might be teaching or gardening or woodworking.
The void comes when we let AI do everything and define what's valuable. The cure is re-claiming what we value, independent of what's easy.
Curious — what's the first physical thing you're building?
Thank you for your comment!
It's very thoughtful and to the point.
In fact I've facing same situation,,,
Nowadays we can use After Efffects / Premiere Pro MCP for making youtube videos, but the output are often disgusting.
Many times I would say "Why you choose this damn circle?!" and so on.
I fully agree with below as well.
For this part, I'm still thinking!
Really glad this resonated — and yeah, the video tool thing is such a good example. AI can generate something that technically works, but it has no idea why a circle feels wrong in that moment. That judgment still has to come from us.
As for what I'm building — I've actually been tinkering with a small electronics project. Nothing fancy, just a little LED matrix display that shows font glyphs in physical form. It's totally impractical, but there's something satisfying about seeing a letter you designed light up in real life instead of just on a screen. No AI involved — just soldering and code and trial and error.
It's been a nice reminder that building with your hands hits different.
What about you — any physical projects on your list? You mentioned getting back into electronics — would love to hear what you're thinking about.
The emptiness you describe is real and I think more people feel it than admit it.
But here's what I've noticed after using AI tools daily for a while: the skill that matters most now isn't writing code — it's knowing whether the code is correct. You said "if it passes the tests, ship it" — but who writes the tests? Who decides what "correct" means for edge cases the AI never thought about?
The METR study found developers thought AI made them 24% faster, but actual measurements showed they were 19% slower. The gap is real. The value isn't in writing faster — it's in knowing what to verify.
So I'd push back a little on the "what's the point of being an engineer" framing. The point is shifting — from "person who writes code" to "person who defines what correct means." That's harder, not easier. And it's something AI genuinely can't do well yet.
The app you built in 2 weeks is impressive. The question is: what happens when something breaks and you need to debug code you didn't write or fully understand?
Thank you for your comment!
As you say, our role as an engineer has completely been shifting.
We have to admit it!
Yeah, I experienced many difficulty when I faced bugs.
In the end, the app was destroyed and rebuild from scratch many times.
You must agree with that "test-oriented" development is our answer!
Maybe we're redefining what 'engineer' means. Writing code was never the end goal — solving problems was. If AI handles the syntax, we get to focus on the logic, the architecture, the 'why' and 'how.' That's not replacement. That's elevation.
Yes, totally agree with you!
This resonated hard. I've been on the other side of this exact experiment — used AI to ship 28 browser games and 15+ digital products in about a month. Full payment systems, analytics, the works. All functional. Total revenue: $0.
The void you're describing? I think it's this: AI collapsed the build gap, but it made the distribution gap painfully visible. Building was never the hard part. Getting humans to care was.
Your "going to zero" insight is right, but I'd reframe it slightly. It's not that engineering skills are worthless — it's that execution alone is worthless. The gap between "I can build this" and "someone wants this" was always there. AI just made it impossible to ignore by removing the excuse of "I couldn't build it yet."
Your pivot to electronics + YouTube is actually smarter than you might think. You're not just moving to "physical things AI can't do." You're moving to distribution (YouTube = audience) + trust (showing real hands building real things = authenticity AI can't fake). That combination is what turns $0 projects into $1 projects.
Curious about Pikutan — did the payment system actually process any payments? That's the part AI can't generate for you.
Thank you for your comment!
You must be very qualified to handle AI since you made a bunch of apps within one month.
So,,, "it made the distribution gap painfully visible" resonated with me.
Actually I'm sure Pikutan is an excellent app for many English learners, but eventually,,, nobody used :(
Distribution and trust will be key to success in AI era as you said.
And as for "did the payment system actually process any payments?", yes.
The process worked well in reality ( I tried myself. )
But I'm not confident about serurity holes,,,
Thank you for writing this. The void you describe is real. I feel it, and some times try articulate it.
I think AI didn't kill engineering. It just exposed what was never fundamental. Now I realized execution was never the core skill.
This were difficult and required effort and it felt like mastery. Now AI just compressed execution cost, and when execution becomes cheap, you finally see what remains valuable.
My opinion: what actually survives 10 years (maybe more).
It's systems thinking (boundaries, invariants, tradeoffs, failure modes, long-term cost of change), decision making under uncertainty. Execution is becoming probabilistic now. But deterministic governance is becoming in demand (sometimes I wonder how this is different from programming). Also AI is still weak at problem framing. And final skill is responsibility (there's nothing to even discuss here).
Over my 15 years of experience, I have felt this emptiness constantly. It's just a signal - come on, go deeper, figure out what actually matters. This always pushed me to get to the bottom of things. So it's normal. And I don't see it as a warning, just as an invitation.
That's a very positive attitude!
I wanna follow your lead.
The system-thinking is really required recently in my company. Perhaps in other all IT company, too.
I'll keep a positive mindset like you and take this opportunity as a gift!
Great writeup, real problem. Study the code, and make sure it's good for one?
Thanks!
Yeah, studying code is still important for us!