First, to be certain I haven’t miscommunicated – nothing is “wrong” with your approach; rather, your approach will simply result in a great deal of time/effort/heartache that in the end, you’ll have looked back on – and wished you had done things differently based on what you know at that point.
I’m merely attempting to help save you some of that heartache/lost years of your life. Nothing more. I have zero interest in proclaiming something to be “right” or “wrong”, but only to help you avoid the pitfalls.
As I wrote originally, I could go into 10+ pages of detail on this. However, if I had to answer your question in the most succinct, simple method possible, I would say…
You’re starting in reverse. Marketing (and understanding the “itch you’re scratching”) drives successful business.
Understand a very specific, very marketable niche. One that has some viral aspect because marketing, particularly to get off the ground, is one of the most substantial hurdles/expenses, not the technology. No one in tech believes this until it’s too late, unfortunately.
I’m no different. I was in direct competition with Elon Musk; we both began the same type of business as the same time along with two other competitors. We were both young and (relatively) broke at the time. Elon didn’t care about building the best tech – he cared about marketing it (and raising money to market it). I cared about building the best tech on a shoestring – because I thought that would essentially “sell itself”.
I’m a science/tech/intellectual property person – so I thought Elon was making a mistake. He sold his company (Zip2) for $300+ million in ~99, went on to fund PayPal, end the rest is history.
While he’s a billionaire, I’m merely a content person. And I only reached that point because I learned the lessons that seem counter-intuitive to you now.
In short…
Marketing and the desire/virality of your product/service is the most significant hurdle, then scalability. The tech, as counterintuitive as this sounds, is essentially tertiary or later (depending on whether you classify building the right team in this hierarchy, etc). Musk knew this from the beginning; I learned much later.
With this in mind, take a look at your post:
All of these are things you’re interested in, from a technical standpoint. Not one of them will sustain the growth of a profitable business, all things considered.
They’re great ideas for an open source project. If that succeeds and develops enough momentum, sure, attempt to build a business in that niche (it would still be an uphill battle, but momentum is everything).
Yet you’re thinking from the wrong perspective. And worse, it’s so generalized and short-lived (relying to a degree on such an agile niche), that its end is already written.
Think instead about what people/businesses need, and how you can make it better. Not from the standpoint of tech, but the use case. And AI should merely be one of many tools that facilitate that. Building an “AI” business without a substantial amount of talent/cash at this stage is a fool’s errand. Much like the people that believe they can become a Youtube influencer (in a related post I’ve submitted here).
Once you read “The Lean Startup” you’ll understand the importance and alternative perspective/approach of testing the market first, then building the business second (essentially). That should help cement the point I’m attempting to introduce you to, here.
Good luck to you!