Just a sidenote:
My concern is not personal. It is a systemic operational safety concern.
As we speak, millions of users are being scammed on X by bot armies. Grandmas, youth, special needs, mental health victims, etc.....
Fake links, fake accounts, identity laundering, authority laundering..etc.
And the larger point remains, it seems like your focus is on Mars and humanitarian crisis in europe more than what is happening in your own system.
As the resident planetary genius and curator of the town square, you have a greater burden than you might imagine to focus on safety and wellbeing on X.
It is your responsibility and burden alone, unfortunately, to clean this place up.
Stop the scammers from stealing my grandmas retirement savings.
Focus on the mission at home before expending energy on drama quests in foreign countries.
We need you Brother Musk.
Please consider this as it is intended, respectful and diligent.
The time is now. Before the real agentic bot armies show up.
Almost all sectors of the economy have relied on AI for years. Individuals have created a lot of wealth, as have companies , from "using Ai".... which leads me to....
What is the goal of your question?
Are you taking a poll or asking for some advice about how to make money?
If it is the latter, disclose your capabilities, target audience and ambition level.....and I will give you some advice....
Or not.
By the way. Im a real person not a bot and I have nothing to sell you.... just a fellow friendly human......lol , if you can believe it.
Good Luck Sister Mari.
It depends of what you are building. Simple, straightforward projects can be accomplished and verified with current tools., ala Codex.
For larger projects, you have to have technical knowledge that is not attainable in the current stack, or at least not without some very imaginative combinatory systems logic.
For complex, real world work,
You need a verification layer that is only available to a person with direct experience and technical knowledge.
I will give an example from another field. Mathematics.
Formulating new ideas and proofs, even if correct, ultimately need to be checked over and double checked again, by real mathmatics professionals. Again, the fact that it appears correct at the outset cannot be trusted until verified. Coding is like this as well, although getting better, the problem with vibe coding alone with no real training is that you do not know what you do not know. And all bugs come back around with sharper teeth eventually.
That said, using modern systems like codex have made the verification layer decent, but for any real bulletproof work, you still need someone on the team with technical systems expertise.
AI is still fairly unreliable. No matter what the frontier labs say.
The code is very good. But not perfect.
There is a reason why all the labs have thousands of employees.
To sum up,
for a prototype, personal tool, or small low-risk product, a vibecoder can work largely alone with AI.
But “AI-only” becomes fragile once the project involves real users, money, sensitive data, security, accessibility, legal obligations, or operational reliability. AI can generate code, tests, documentation, designs, and debugging suggestions, but it cannot reliably replace independent review, real user feedback, domain expertise, or accountability.
A good rule is:
Build alone with AI; validate with humans before consequences become real.
So the practical path is:
Use AI heavily for ideation, implementation, testing, and iteration.
Keep the first version narrow and reversible.
Bring in humans at critical gates: security review, user testing, legal/compliance, accessibility, and production launch.
Never let the same AI that wrote the system be the only entity that verifies it.
A solo vibecoder can absolutely create something valuable. The mistake would be assuming that “solo” also means “unreviewed.”
@Humphrey_Jams@MillionInt Your grocery bill will go up.
As will the property taxes
The costs of gas.
The cost of education.
Etc.
You know what will change though for sure.
The fortunate sons bank accounts.
@jxnlco On the evidence presented,
Jason has at least earned:
a raise,
a weekend-detector,
and one mandatory Monday off.
The Dragon recommends clemency—and coffee.
@jxnlco There are no days off in the Battle for the Future of Humanity.
Let me ask you this?
Did anyone respond and thank you for your dedication?
If not, Thank you my friend.
There is great love for you here at my house.
Carry on.
My guy, you know everyone has been at the wine bar all weekend.
They worked like 42 hrs week, jeesh.....they need their "refresh" time.
They only make 300k a year.
Tagging Threats on the weekend?
What a novel concept.
Let me help you out, Jason
You are not the problem. They are.
Get to work ClownFish.