And the network already exists because it causes profit with other uses. Adding SMSes does not increase the cost of maintenance. So yes, SMSes are still actually free.
And do what? Do you actually read the license plates of cars you pass? And even if you do would you recognize that it was the same as a random string of letters and numbers of your phone?
As for the children- don't get me wrong, if I hear a child screaming "Get away from me, you're not my daddy" or "help I'm being kidnapped" I'll intervene. Short of that- do you stare at every little kid you see to check if they match the very vague description sent to your phone? Do you know the number of false positives and wasted police effort you'd cause if you did?
Nope, the AMBER alert stuff is useless. There's a point in emergency weather notices and major traffic conditions (flash floods, closed roads from earthquakes/rockslides, tornados, a bridge has collapsed, etc). There's a use for presidential (hey, we're at war and China is launching aircraft at us, you guys on the west coast go hide). The amber stuff is just feel good uselessness.
You're perpetuating a fallacy yourself. They aren't building a system to do this- the system already exists. So those fixed costs are already paid for, and would be paid for regardless of this service because it provides other profitable services. So the marginal costs are all that matters, unless we get to the point where the bandwidth used by SMS is enough to require additional hardware to be built (which for SMS is never going to happen).
If they had another way to monetize that small amount of bandwidth used you may have an opportunity cost of using the bandwidth in this. But the fixed costs don't factor in.
Inflation has multiple causes, and existed even before fiat currencies did. Increasing the money supply is one way to cause inflation, but not the only one.
Ideas are easy. I've got dozens. Marketing is hard, it's why I'm not a millionaire. In this case I wrote the app for myself, after having an extended text conversation where a girl I really wanted to talk to was texting me at odd intervals while I was driving, forcing me to stop every two miles and pull over to respond.
Agreed, its better not to text at all. You're at least somewhat distracted when you do. But lets face it, some people won't do that. My belief is to lower the danger as much as possible for those who insist on texting, and since you keep your eyes on the road I do believe its safer.
As for homophones- voice dictation software these days operates on a sentence. 99% of the time you can differentiate between those words based on context. For the 1% you can't, you flip a coin and possibly send the wrong one. Hardly the worst autocorrect mistake you'll ever make. I'd bet on making fewer mistakes with a readback prompt than you make in normal tapping.
It launches itself when it detects an incoming text. In Android you can declare a class (a subclass of BroadcastReceiver) that will have a function on it called when the OS detects certain events (like incoming SMSes). The motion detection algorithms launch at boot.
I wrote my own hands free texting app, that automatically determines when you're driving (based on speed). It solves this in a very simple way- after you speak your response, it repeats it and asks if you're sure you want to send. If you say no, it lets you re-enter your response. No need to look at a phone at all.
Cheap plug: Text Soundly is available at the Play Store here.
No, you don't realize you suffer from it. You won't until it accumulates to the point you need corrective glasses.
Do you understand the concept of long term eye strain? Its not a matter of just getting a headache or eyestrain on a given day- there's cumulative damage from staring at a screen all day.
Tablets are horrible for reading ebooks. They smudge easily, they strain the eyes, and the battery life won't support reading for an extended period. An e-ink reader is orders of magnitude better.
And at no time did you realize that it's the completely wrong tool to use for that and that there's better and easier ways to manipulate that kind of data (if you even need it, most of those intermediate data points will never be used)? Wow, that's grade A level incompetence there.
You mean it would allow people to install applications they want on hardware they own? The horror.
You're assuming the client wants to be involved that hands on. Most don't. They want to see a prototype every now and again and that's it- they don't want to be involved in every meeting about feature prioritization etc. Remember that this isn't their job- they have other things to do to actually make money.
And while developers lack knowledge of the target domain, the customer lacks knowledge of the software domain. Involving them in every decision just makes them feel stupid, clueless, and frequently they make the wrong decision. Its not as cut and dried as you'd like.
THat's the same with any methodology- if they want a change you scope it out and give them a number. Agile is no better or worse at it. If anything the numbers tend to be a bit higher because of the tenet of not working on anything that isn't needed immediately, whereas when you design up front you tend to design in some flexibility for likely future changes/usecases.