Your response strikes me as typical of programmers in that they don't recognize how their work can affect a great deal more people than almost all of the examples you cite. With the possible exception of mishandling food, none of the other examples come close to affecting the same order of magnitude of people as programmers can.
The recent VW emissions scandal is a perfect example: VW's proprietary software was used in around 11 million VW cars worldwide (that VW admits to) from model years 2009-2015. Comparable proprietary software was used in more cars of other makes and model years. VW's software apparently turned some VW cars into cars that never should have been sold. Other makes and models of cars are also showing bad signs of polluting too much and not being in line with regulations. The full scope of the damage has not been accounted for. Only centralized food processors working on very highly used ingredients have the potential for that kind of adverse impact.
This creates a situation that kills us slowly instead of quickly by polluting our air in ways our (admittedly inadequate) regulation framework was designed to disallow. Proprietary software cheated those tests by behaving radically differently in regular driving than in testing mode. These cars should all be taken back by their manufacturers at full cost to the manufacturer, giving the current owner a complete refund of whatever they paid for the car, and the manufacturer's higher-ups should pay with criminal penalties and huge fines because this is a serious environmental matter. Programmers know their software is widely used (some programmers even value the wide reuse of their code) but rarely do programmers brag that their software treats people ethically and well.
Being "aware of their moral compass" is too low a standard and something programmers have typically balked at besides. As Brad Kuhn points out, software freedom doesn't kill people, security through obscurity kills people, yet programmers today still debate the value of software freedom for its own sake instead preferring to either work on proprietary software outright, or choosing to value a non-free software-allowing right-wing corporate reaction to free software known as "open source". Read just about any /. thread today and you'll find plenty of technically literate people who balk at introducing ethics into the discussion, or try to explain away giving us all the means of helping ourselves via software freedom. Our best chance of finding and fixing the cheating car code is to require copylefted free software for all vehicles and make transfer of the complete corresponding source code and build instructions for said software with ownership of the vehicle. But we choose not to do our best motivated in part by those who would rather not enter into a moral discussion because they place business desires above how people ought to treat other people.
One easy way to help fix this is helping those who help us. Today the Linux kernel is used in a lot of products that end up in people's homes, listening and watching them all the time via cameras and mics controlled with proprietary software. It's hardly a stretch to imagine that non-technical customers are being spied on without their knowledge or consent. It's bad enough that Linus Torvalds' fork of the Linux kernel allows proprietary software (as opposed to GNU Linux-libre which does not), but GPL violations are rampant. We can help the Software Freedom Conservancy by funding their efforts to pursue GPL violations, and I hope you'll do so. We owe the entirety of free software routers to comparable efforts, freeing code from Linksys which we can apparently reuse in many other routers. That freed software and its derivatives makes routers more trustworthy, improvable, more long-lasting, and worth paying for. We should not fuel the destruction of civil liberties by making it so easy to put Linux into devices and then not care if those devices are used to spy on people.