Letting it die is the self-serving, career-optimizing, amoral take. But it's more ethical to stand up for what's right even at personal cost. A bunch of people wasting years of their life, not to mention all the resources, is a tragedy worth avoiding.
Of course, the wisdom of taking the person risk is a continuum. In some cases it is and in some it isn't. But.. To omit the ethical angle entirely seems like a bad take.
I don't understand this point of view. Most of the people aren't wasting their time. They're getting paid for the effort. The business is taking a risk, and pays people to realize their vision. Some visions are bad.
Getting personally attached and emotionally invested in work you get paid for is a risk too. There's nothing wrong with that. But there's also nothing wrong putting your time in and churning out requirements if that's what you want.
of course they're wasting their time. a year of work deleted? all you have to show for it is money? you could have money AND something to be proud of. what a waste it was, to do something pointless for a year when you could have done something important.
not to say that there aren't experiments worth running. but in my experience (and in the example in the OP's article), the experiment often isn't even worth running. Intelligent people knew from the jump that it was a bad idea. No experiment necessary. Just pointless waste, enabled by hubris and apathy.
It seems different people get different things out of work. My favorite kind of code to write at work is code that I know will never make it to production. No chance of requirement change or incomprehensible support tickets. I mean your way is valid too.
ah, to be clear, I'm not really talking about anyone's individual values here. It may well be that a person is as rewarded, or even moreso, getting to work on some pie-in-the-sky foolish idea as they are one that is productive for society.
My point is that, from the viewpoint of an external observer: we want people doing prosocial work if it's possible, so if there's counterfactuals where they do or don't, we'd rather they do. And the point of taking moral stances on stuff is to press for the attitudes and behaviors that lead to a world that maximize's everyone's safety and happiness and prosperity and whatnot. Therefore it is moral, at some level: a world where the company wastes money on dumb stuff is worse than a world where the company channels that money into value for humanity.
It has nothing to do with the individual's preferences, and I don't begrudge them their preferences either way. My point is that when we're evaluating "letting projects fail" as a policy, the moral angle needs to be part of the conversation, because it does have moral implications. Convincing yourself that it does not is a moral choice: it amounts to saying that you do not feel responsible for those implications in the slightest. That doesn't mean you should turn around and be fully responsible for them, either. As with anything there is a lot to weigh. But completely writing it off seems wrong.
You can voice your concerns, but should not go fighting, especially at personal cost. It could be that you may be wrong in your assessment, and the project turns out to be successful, or it could be that you may have been right for the wrong reasons, or it could be that you were right all along.
In any case, you are part of a company, and that means recognizing that yours is only one of many opinions driving strategy and allocating resources. If you find your self often needing to stand up against others for your beliefs, then you are probably not in the right company.
the point of expertise and intelligence is, in part, to be able to know what's going to happen BEFORE doing it. Perhaps without even doing it. You could be wrong, sure -- but there has to be a rate of that happening, and the more intelligent people are wrong less. At some point there are situations where you _know_ what's going to happen and then it happens, inevitably, providing no new information. And in my experience this happens _all the time_ in big tech. It is not hard to predict the failures. But things happen for social and political reasons, not intelligent ones, and so the predictions don't matter.
If the company is 'one dev ignoring a bad project heading to failure' away from bankruptcy, you should have accepted a job offer somewhere else last year.
(also that of a "non-ethical" person, like an animal or a person with no agency in the matter, if you want to make the distinction. I'm not sure we should but I guess it's an interesting question)
It seems there's a difference between unethical projects and projects that are just a bad idea. If someone wants to pay someone else to work on Uber for cat socks, but with AI, I don't think there's much of an ethical dimension.
Yes, that's certainly true. I guess what I was alluding to is that I think ethics should inform all decision-making, including business. A bad idea isn't necessarily unethical, although the execution of it (or even a good idea) can be. Unless the idea harms unconsenting others.
Generally, it's not what you sing, it's how you sing it.
No one's talking about building ethically-problematic software projects here like software to aid despotic regimes, harm human rights, etc. We're talking about business projects that senior engineers believe will fail either in execution or in the market. There's nothing unethical about just letting these things fail, especially if they aren't even a project you're assigned to work on. It's upper management's responsibility to assess risk and pick good project to assign resources to; as a senior engineer, your only job is to advise them when consulted. If upper management is incompetent and doing a bad job choosing projects, then your recourse is to go looking for a better-run company, not telling your executives how to do their jobs.
that's not how morality works... morality is about doing what you think is right instead of excusing yourself not doing it. If you think a bad thing is going to happen that you have agency to prevent, that's on you. Whether you have the data from repeated events or not doesn't matter. (Of course, your confidence in your own belief goes up with more data. But there are still plenty of cases where you can be sure even on the first shot.)
Of course, the wisdom of taking the person risk is a continuum. In some cases it is and in some it isn't. But.. To omit the ethical angle entirely seems like a bad take.