Hacker News Reader

Comment by quasse | Hacker News Reader

Yes. This might be unfamiliar to software engineers, but there's a reason mechanical engineering and architecture (there might be more, but these are the ones I'm familiar with) schools put students through at least a semester of ethics classes.

We are literally the last line of defense in a lot of industries. From the first line of the ASME code of ethics:

> Engineers shall hold paramount the safety, health and welfare of the public in the performance of their professional duties.

If you can't do that, you should be in another field. Maybe software engineering at Meta would be a better fit.

View on HN

I went to a Science and Engineering school, and everyone had to take ethics: ME's, ChemE's, CompEng, Civs, etc.

This is such an important part of engineering as a profession.

This is an absolute joke in almost all fields of engineering. Milage may vary, but it's basically the wild west. I can't speak for PE's, but obviously there's at least a little more regulation there.

However a huge number of engineers critical to safety would not have one.

This is a noble goal but unrealistic and unfair at scale. Also it belays just a wildly unrealistic understanding of how risk distribution works in corporations. It’s as if in your mind one engineer makes the singular call and sees the entire picture of the project and its risks. Where as more likely a downstream worker who very likely had an incomplete picture will be blamed.

Also the personal insult is not needed. It just makes you out to be an ass rather than someone making a point.

Consider your romantic notions might not be realistic.

I didn't read it as a "personal insult".

It's true that most software engineers never work on safety-critical systems. The nature of the products they work on, or the nature of their contributions to those products, limits the possible impact of flaws and mistakes.

Once you work on a product with safety-critical features and/or become responsible for one, there really is a marked psychological change that takes place. I've lived through that change in my own career, and many comments (including some I've written, I'm sure) now strike me as callous and trivializing important matters to an extent that grates.

It's important to hold engineers accountable for their work in proportion to their impact, which is typically very significant. If you build machines that can kill, stakes should exist for you as well.

To the contrary, there is usually _several_ engineers who see the ethical/safety problems and talk about them with colleagues. They just don't report it to the right authorities.

>It’s as if in your mind one engineer makes the singular call and sees the entire picture of the project and its risks.

To a certain extent this is true. To extend the OP's example, a mechanical or civil engineer who stamps a building design takes responsibility for that design being safe enough for public use. The same could apply to the MCAS design, or the quality engineer who signs off on how the plane was put together. These are not "romatic notions"; the latter is already common practice in aerospace. Do you think if a QE stamped that the door bolts were installed should not bear responsibility for the bolts being missing?

In those cases where the engineer is supposed to be aware of the risks, there should be processes in place to account for, and mitigate, those risks. For the example of MCAS, the software was categorized as hazardous in the hazard analysis document. As such, Boeing processes required it to have redundant inputs as a default. It did not (the redundant sensor was sold an optional equipment); as such, I think the engineer who signed off on the hazard analysis may bear some responsibility. If, on the other hand, there were no business processes in place to require redundancy, that seems like a CEO responsibility.

I don't see a personal insult, but maybe SWEs should take the ethics of their work more personally. Maybe the shape of our epidemic of social and psychological dysfunction would be different (if not smaller).

When I worked at a retailer, we were told to push buy-now-pay-later schemes on our customers (on top of the corporate credit card, and insurance plans, and service plans). Most of my coworkers simply decided that they weren't going to - in part, because such schemes are nakedly predatory, but also because it was yet another dumb metric we knew we'd be held to if it were successful. I don't know of anyone who was fired or docked pay or denied a promotion because of it, and it's not worth luring someone into financial ruin over a TV and an entry-level sales position. Which leads me to another thought: if you're so low in the chain that you have no say in how and what you work on, is it a job that's worth sacrificing your morals for? Maybe take advantage of that increased earnings potential from job-hopping and get out of dodge.

> This is a noble goal but unrealistic and unfair at scale. It is in the current organisation of software development because it doesn't allow that.

In the building sector you can't build without your design being signed of by an engineering office and the engineer and its office take the responsability for it, both at the criminal and civil level.

If the same rule existed for the software industry, the organisation would adapt itself around it.

Nobody asks the engineer to see the entire picture. (Very few, senior, engineers see the entire picture - to the shock of many newbies. But that's for another discussion). The problem is when the engineer DOES see the problem and doesn't question the project, and stays on, and doesn't figure out a way to get it questioned, and in the end doesn't report anything. THAT is an ethics problem. And yes, in some companies it will cost you your job. That's where the other jobs come in - because there ARE other jobs.

Reply on news.ycombinator.com