Entrepreneurial ideas

Service to validate whether you should (or shouldn't) become a software engineer. There seems to be a missing step that folks should take before attempting to become a software engineer with zero experience: determine if they would even a) enjoy being a software engineer, and b) possess the right skills/talents/mindset that would make that a plausible career to switch into. This could probably be said for all careers, but since there will never be enough software engineers, it seems especially pertinent to evaluate these things up front because otherwise you end up with a lot of people thinking it's just like any other job. Enrolling in a bootcamp or degree program is a really huge step, and it's only responsible to take that decision seriously.
Software engineering heuristics: This is probably more of a research idea than an entrepreneurial one, but hey, I don't have a page for that yet. We (software engineers, as a field) should attempt to nail down the heuristics that we use day to day across the entire domain of computer programming. Not specific heuristics within an algorithm. Rather, the way that software engineers work around the combinatorial explosion that would destroy any attempt to program by some explicit set of steps. For example, debugging a problem involves running the program being debugged in one's mind, but we don't run it exhaustively — there is some heuristic by which we determine when to "skip over" the execution of some function and assume that it does it's job correctly, rather than continuing to dive deeper. What is that heuristic? How does it work? There are many similar heuristics, that, when combined, allow us to function successfully as a software engineer, and when we fail to pick up on these we fail to program.
Learning products that adapt to, or at least pre-empt, the learner. I've thought a long while now about doing more tutoring, or maybe even doing some full on teaching. I tutored my brothers for a bit in math and actually quite enjoyed it. The point of this tutoring would be to better understand where learners go wrong. Surely there must be patterns in the ways that learners misunderstand things, or situations where they understand things but apply them in the wrong context. Once I am able to make a "map" of these errors, I could then develop dynamic educational materials that are able to detect when a learner is making a mistake and send them down a side path that helps them understand the errors in their thinking before moving on to more advanced subjects. I believe that this could be a big boost to students who do not have the best at-home educational support, which is often how these errors get corrected in more affluent households.
Automated testing tools that aren't painful to use. At work I've been using cypress to create automated integration tests for our applications. The actual tests aren't too bad, but as a whole the process is incredibly complicated and prone to failure because of all of the external concerns that need to be addressed like how your application runs and is configured, any dependent applications and infrastructure, setting up test data, etc. etc. Perhaps this area is doomed to forever be complicated because of the sheer variety of how applications function, but something tells me there must be a way to do it such that it isn't so painful. The main thing is that I don't think I should need to write code in a real programming language. Maybe some custom DSL, but I shouldn't be dropping down into javascript to make this happen. Non-programmers should be able to set them up, and somewhat quickly.
Alternative Amazon search interface/algorithm. Since Amazon's isn't that good.
Employment pipeline for the trades. It seems like there's a lot of demand for home services and other trades, but the issue that many of those business owners have is that there aren't many qualified people signing up to do those jobs. There would likely be some kind of educational component to this, plus certification?
Storyworth, but "crowdsourced". Storyworth is a product that collects stories from a particular person (through email I think) over a long period of time. After enough content has been generated, it is organized and edited into a nice printed book. The issue with this process is that it puts the onus on the receiver of the gift to write, as opposed to the gift buyer. Even then, putting all of the effort to write all of the content in a book on one person at all is a lot to ask! So the idea is to distribute the effort among a group of people who might all write stories about the gift receiver who would then receive the book at the end of the process. Another idea is that a group might all write on some other topic (e.g. perhaps a project retrospective at a work place) and then each member of the group gets a copy at the end.
Glass measuring cups with etched measurement markers vs. the paint that comes on them normally that washes off over time. Related: Shot glass with marks representing one drink based on % alcohol of whatever liquor is being poured into it.
A tool to visualize, explore, and build knowledge. Maybe this is just a mind mapping tool, but it feels like an area that could use some more specific tooling than that. Also a way to share and collaboratively build these maps of reality. Maybe this should be more specific to exploring one's own beliefs and where they come from.
Stop picking your face trainer. AI-enabled tool that watches you and dings you annoyingly when you put your hand on your face. I pick my face too much and I think it's one of the only reason I have much acne these days. I've done all the right diet stuff — cut out certain foods, etc. I'm pretty consistent with skincare routine — moisturizing, using benzoyl peroxide, using differin (adapalene) gel, and using sunblock. I've cut out a lot of sodium lauryl sulfate products which can irritate skin. Right now I'm trying a low-tech route to try to get myself to stop picking that involves wearing latex gloves during the workday when I'm sitting at my computer, since that's when I do almost all of my face picking.
Map of friendly weirdo twitter / other various groups. Should be based on interactions (likes + replies) not on follows, since a lot of people have switched to using lists, which are often private. More ambitiously, a way for this group + others to bootstrap away from twitter as a platform — perhaps some kind of forum or chat software, but my gut tells me it would have to conform to some level of weirdness in order to get traction. What kind of weird features would be a draw?