I'm working on a thing now which no one believes is possible in the AI space (and to be honest I didn't either).
So I built a tech demo which proved to myself that it really does work.
Now I do the "I'm doing XXXX", get the cynical "yeah right" look, but the idea is interesting enough that people are prepared to see the 30 second demo.
The demo is pretty cool.
After the demo I've had people ask if they can invest, without me asking, and I think every person I've demoed to thinks they would either buy it or know someone who would.
So I'd caution that not all advice is correct in every situation.
From what you say, you did some technical exploration (which again, I'm all for). Then you didn't make a full product, ujust a 30-second demo. Next you went and tested the "people want this" hypothesis. That sounds very much in line with Lean Startup advice.
Great tech MUST be sold, once you have the proof of concept you must get someone great at marketing on board to help you sell it. There are plenty of great marketers who can help you with this (for a price...).
My only point is as a great technical person you start with getting it working before trying to sell it. Once it is working great marketing people will come on board and help you sell it - that is partially a matter of finding customers and partially a matter of making you admit those hard rough edges need to be fixed.
For most product ideas, the key question isn't, "Is this possible to build?" It's, "Will enough people buy it to make a sustainable business?"
That is to say, the greater risk of failure comes from lack of demand, not lack of technical competence. The greater risk should generally be addressed first.
This is compounded by the fact that it's usually much easier to test the "people will buy X" hypothesis than the "I can build X" hypothesis. Or, put differently, the cost of reducing market risk is usually lower than reducing technical risk.
Furthermore, even if we're in the rare case of working on something where technical risk is significant, we generally don't need to get a real, working product to reduce that risk sufficiently. Some technical experiments are generally sufficient to put us back in the realm of, "Yes, we can make that product."
As a technical person you are better at getting it working so spending a couple weeks getting a working prototype is worth the effort, it might be ugly but it builds trust that it can work. If the idea is radical your potential investors/customers won't believe it is possible.
The real problem is until you have enough paying customers to pay everybody you are living on borrowed money. How you best borrow the money is the only question, and the answer if different for every situation. Some ideas are so obvious when seen that you can get millions from investors. Some ideas will get the "that ain't possible" reaction until you show them the finished product when they will beg you to take their money. Some ideas look cool but nobody is willing to actually pay for them. Some ideas are cool and will generate sales - but never enough to pay for all the people required to make the product. You have to figure out where you are.
Even if you have a revolutionary idea, and even if you can build it, it's still a waste of time if nobody will buy it. No matter how revolutionary your idea, you can try marketing it. If you are eventually going to sell it via a website, just build that website. If you're eventually going to sell it in person, just sell it in person. If you need a video demo, just make the demo. Then see what people do.
There are two reasons for this. The obvious reason is that you can save years of your life building something that nobody will pay for. But the subtle reason is that finding out what people actually want will usefully inform what you build. You can spend those years building the revolutionary product that people will buy, rather than the not-on-target product that they won't.
The techniques for this are not hard, and the Lean Startup literature has plenty to get people started. E.g., the Alvarez book on Lean Customer Development: https://www.amazon.com/Lean-Customer-Development-Building-Cu...
(And this is not to say that people shouldn't obsessively build things that interest them. I like doing that. It's a great way to learn and explore. But they shouldn't pretend that's a good way to start a business, any more than amateur musical noodling in one's basement is likely to produce the next #1 pop hit.)