Maybe it's time to create a Hippocratic Oath for developers to publicly commit to?
A future Path developer could then refuse to implement an unethical "feature" by pointing out that the company had hired them with the full knowledge that the oath had been undertaken.
I don't think the company would pink slip the developer, as they would probably want to avoid any attention being drawn to the unethical "feature" in a tribunal or other legal setting.
People want validation. Line-level employees want praise from coworkers and bosses. Executives want praise from their peers, investors, industry, and press. Concepts like ethics, "right," or even this-is-good-for-thie-world isn't a concern when faced with "X will increase my social status and happiness with my peer brogrammers." What's X? It's anything possible, regardless of legal, right, wrong, or ethical.
A nontrivial number of companies use unethical methods (spam, false invites, false installs, phone and email address book capture, fake attractive profiles) to increase their vanity metrics. Employees see those methods as either: "this is bad, but it's sooooo good for us — look at all the lame n00bs who fall for our tricks" or "this is bad, and I'm ashamed to work here."
The ones who feel shame would take the Hippogrammer Oath. Those who revel in manipulating others and standing on their broken bodies will rake in all the profits while the good guys just sit around and "play nice."
Even the tech darlings of today used spammy methods to grow their initial user base. How do you grow your userbase to ten million when you're growing at a constant 5,000 per day? Obviously you want to "go viral." How does one just on a whim "go viral?" You can either become a meme, a social phenomenon, or spam and manipulate unsuspecting people. Spam is less work than creativity.
http://blogs.msdn.com/b/oldnewthing/archive/2006/11/01/92244...
I often find myself saying, "I bet somebody got a really nice bonus for that feature."
"That feature" is something aggressively user-hostile,...
Things are going to look different to me now when I'm on a windows machine.
I remember having to put my job on the line a few times for refusing to program / setup something awful.
One of the worst was when I was asked to combine all divisions email lists and send out a marketing email selling some overpriced book, this was against the Privacy Act (AU), against our privacy policy and highly unethical to boot, refused and was given a written warning.
It was a contract web development company I was working for, about ten years ago. One of our clients wanted some SEO work done, and my supervisor had recently been reading a lot about SEO. He started out reading white-hat stuff, but by this point he was delving into some black-hat research, and he essentially asked me to program a message-board spam-bot. I just told him straight up that I believed that would be unethical and I refused to do it, knowing full well that simply refusing to do assigned work could cost me job.
Thankfully, not only did this not cost me my job, it caused my supervisor to re-evaluate his own position and he decided to go back into completely white-hat SEO. And in the end, he actually thanked me for refusing to do that work.
I feel like lucked out on that one.
I think it is important that we should realize this kind of mentality won't work. You might see bad people making money, but eventually it will be no good for them. Either they don't sustain, or the money is no good for them, or they can't sleep with all that money under their pillow. You will see lot of examples of this from history.
I have seen people who are ethical and right also make a lot of profits. May be not in the short term, but in the long term. The idea is, you don't go behind money, instead you do what you do best, and money will come behind you.
At one time, it was cooperative multitasking and memory management. Programs were supposed to behave themselves and get out of one anothers' way. Except that, due to bugs or malice, some didn't. We called this world "DOS" (or pre OSX Macs).
Microsoft still attempts to allow vendors to install programs whereever the hell they want, and to, pretty please, not overwrite other program's infrastructure or system-level DLLs. Yeah. Right.
In the Linux world, we've solved this problem, if done right, though distro-managed, well, distributions. Any program can be included if it meets qualifications (generally limited to licensing requirements), and a sponsor steps up. Once included, the package gets the benefits of being included in the package lists, distributed over archive mirrors, and included in bugtracking and support systems. However it's also got to play along with the requirements of Debian Policy as to how it behaves on a system.
The proper way to address the issues of app privileges is to control privileges centrally on the device and grant them to specific apps. If a user doesn't wish to give an app, say, addressbook access, then they can deny it (or feed it a bogus addressbook). The app vendor can decide what they're going to do at this point, but what they can't do is override the user's explicitly stated limits.
This is silly. Stop trying to add grandeur to writing some code at X,Y startup/company.
People don't die or get harmed when some social-messaging application spams someone. Code is a way to implement an idea. Most applications exist to make money. If this a shock to you, read the user agreement before installing/upgrading, uninstall the application, or realize that in social networking, your personal data what the company uses to make a profit.
Ethics in computer-science-related fields are important and I think we do need a set of rules we can dogmatically follow like the Hippocratic Oath. Of course, the HO is different in that failing to follow can cause physical harm. However, the world is progressing quickly and more and more information is hosted online -- personal information.
I think it's our jobs to make sure we don't promote poor practice and un-ethical behaviour.
Seriously. I hope that you'll see this in time!
>>>> I think we do need a set of rules we can dogmatically follow like the Hippocratic Oath.
So you don't actually have to employ your own brain and your own moral judgement, because somebody already did it for you and wrote this nice set of rules, that you swore by Apollo to faithfully execute, without thinking, not unlike that box of wires and silicon chips you are paid to play with? Nice arrangement, I suppose.
>>>> I think it's our jobs to make sure we don't promote poor practice and un-ethical behaviour.
And you need to sign an explicit oath to do that?
Also, I think you're sort of merging the HO with the Code of Ethics for engineers in Canada -- two very different documents and I suggest you give them a read. And no, no one still thinks they're swearing to Apollo.
Why you need an oath for that - shouldn't it be always the default behavior?
>>>> and to act in the interest of the public.
"Interest of the public" is a very dangerous thing. I can remember a lot of very bad things that were done "in the interest of the public". You can make almost anything pass as "in the interest of the public" if you want to. Murder? Millions were murdered "in the interest of the public", because they were of the wrong ethnicity, class, physical features or just in the wrong place in the wrong time. Robbery? Millions were stripped of their property and reduced to utter poverty because it was claimed it is "interest of the public" to do so. And so on, and so forth.
I would rather steer clear of anything that has "interest of the public" written on it, at least until it's very clear what is underneath. Too many things that were underneath such writing proved to be a disaster.
>>> no one still thinks they're swearing to Apollo.
Swearing to a document composed by a faceless bureaucracy is no better. If you have code of ethics, live by it, if you do not - get one. What Apollo or his modern equivalent, the almighty bureaucracy, has to do with it?
bookoutlines.pbworks.com/w/page/14422685/Predictably%20Irrational
One more variation: Nina, On, and Ariely conducted a similar experiment. But, one group was asked to write down 10 books they had read in high school, and the other group was asked to try to recall and write down the 10 Commandments. When cheating was not possible, the average score was 3.1 When cheating was possible, the book group reported a score of 4.1 (33% cheating) When cheating was possible, the 10 Commandments group scored 3.1 (0% cheating) And most of the subjects couldn't even recall all of the commandments! Even those who could only remember 1 or 2 commandments were nearly as honest. "This indicated that it was not the Commandments themselves that encouraged honesty, but the mere contemplation of a moral benchmark of some kind." Perhaps we can have people sign secular statements--similar to a professional oath--to remind us of our commitment to honesty. So Ariely had students sign a statement on the answer sheet: "I understand that this study falls under the MIT honor system." Those who signed didn't cheat. Those who didn't see the statement showed 84% cheating. "The effect of signing a statement about an honor code is particularly amazing because MIT doesn't even have an honor code."
So I assume you never use any open source code in any of your programming projects, since you are fundamentally opposed to adopting any ideas that are not exclusively your own?
No, it's our responsibility as decent people. I don't need to sign some online pledge to keep myself from pushing people in front of trains. If I was the sort to harm others, why would I care about some meaningless online campaign?
The Oath is there to remind you to act in the best interest of the user. There are no formalities and although it seems common sense to people like you and me, others might not see it so clearly.
Neither do doctors really need the Oath of Hippocrates to stop themselves from harming people.
Well, yes. But there is a reason that every profession that has tackled this problem has used a system of oaths and certification. Engineers, Doctors and Lawyers are the canonical examples.
You need something that is given and can be taken away for bad behaviour in order to change behaviour at this level. Damn human brains.
Anyway, the issue at hand is bad corporate behavior, not bad programmers. I don't see why we need to start licking our chops about the prospect of forming a blacklist against individual programmers.
So go ahead and try and stop bad corporate behaviour, everyone else can use a proven system so that programmers can easily say "no" when asked to do something unethical and not be fired for it.
I'm often confused by how often programmers completely reinvent the wheel when faced with social problems. The idea of looking to other similar industries never comes up, even if the problem is exactly the same.
So what if they lose their license? They can still write code and do harm.
Yes, indeed. As it stands right now, the reality of the engineer's license is such that it doesn't fit very well the software world. The vast majority of companies couldn't give less of a damn whether you are licensed or not. However, it depends.
Regulations might eventually come in place to force software producers to hire only licensed engineers if the nature of their business is prone to put the public in danger. And as technology grows ever deeper into our lives, the danger that consumer apps can cause on the public is ever growing as well. For instance, breaching a user's privacy can be enough info to grant an ill-intended operator access to the user's e-mail through social engineering, from which it is then often trivial to gain access to that user's bank informations. You don't need that much imagination to figure out a scenario where a user's life can be turned to shit by some software abuse.
Given that this risk is ever growing, the possibility of a code of ethic on software business is plausible. Say in X months, the government of country Y decides that companies hoping to run a social network available on their territories must hire licensed software-engineers, and have them all sign-off any code that is presented to the public. That software engineer they'd hire would have to put their license and career in jeopardy if they were to implement some evil feature.
Before Québec's bridge, engineers didn't need a license to build infrastructure. The parallel between the current situation and the past isn't too hard to make.
Of course they get harmed: their time is wasted, and perhaps their concentration disturbed. This is a small harm to each victim, no doubt about it, but if you write code that makes your social-messaging application spam people then you're delivering that small harm to a large number of people. If your code wastes 10 seconds each, just once, for a million people, that's about three person-months of aggregate time you've stolen that will never come back.
It's quite true that "in social networking, your personal data what the company uses to make a profit". But if it were near-universal practice for new software engineers to swear a solemn oath not to use their powers for evil, who knows? perhaps some other business model for social networking might have had a chance to succeed.
I'm not justifying either approach, by the way, just observing what I regard as a strange disconnect.
What happens if an app for job seekers texts your boss? What if a casual hookup site texts your new girlfriend--even though you signed up a year before meeting her? What happens if your app calls an old person and they crack a hip trying to answer a phone--when everyone that knows them personally knows not to call until they're awake and their caretaker is in?
These may sound far-fetched, and we all mostly don't pretend we're as disciplined as structural engineers, but "we do it for the money lulz" is a shitty and stupid argument.
I'm okay with this. I'd rather be calculating than have my head in the sand about the business models of social networking what-have-you applications.
>What if a casual hookup site texts your new girlfriend--even though you signed up a year before meeting her?
While I don't and won't have to experience this, your imagined relationship suffers more from lack of trust and honesty than "some dumb app does some dumb, annoying thing."
>"we do it for the money lulz" is a shitty and stupid argument.
Don't Straw Man me. If my code was going to be used for something I perceive as evil, I'd leave the job.
Our industry doesn't need yet another pointless, embarrassing ethics/integrity campaign when the people writing the code don't care.
When was the last one?
I disagree. This case reminds me of Geni. You would put a relative's email address in to invite them, and they'd then receive a torrent of spammy "updates", until they registered to unsubscribe.
My less tech-savvy father added many relatives from his address book to Geni. Lots of hate from deranged relatives, and some less technically-inclined relatives are probably still being spammed, 6 years later-it made family gatherings awkward for a while. There are real-world, harmful consequences to this kind of scummy, unethical tactic.
http://www.telegraph.co.uk/news/worldnews/europe/russia/8284...
//It should be noted that no ethically-trained software engineer would ever consent to write a DestroyBaghdad procedure. Basic professional ethics would instead require him to write a DestroyCity procedure, to which Baghdad could be given as a parameter.
http://www.codinghorror.com/blog/2007/05/your-favorite-progr...
I think the system should be licensing and involve losing that license if you commit an ethics violation.
There would be unlicensed developers of course, but connecting the incentive to not do unethical things with the incentive to be part of the elite class in your profession has worked pretty damn well for Engineers, Doctors and Lawyers.
... and has pretty much screwed over the rest of society, at least in the latter two cases. The legal and medical cartels have done incredible harm to their customers over the years.
See http://mises.org/freemarket_detail.aspx?control=51 (law) and http://mises.org/daily/4276 (medicine) for details.
Even if there was some bad "allopathy" back in the day, there is more bad eclictics and homeopathy right now. And to practice medicine you have to understand scientific method, especially falsifability.
And I also have to say that the "free market" idea isn't falsifable. "Let it to free market" rarely works.
They assume rational actors (people making decisions based on their own self interest). That's been falsified (when applied to humans).
Most variations of the efficient market hypothesis have been disproved as well, for the same reasons:
Humans have cognitive biases and other types of irrational behaviour.
But anyone linking to mises.org is probably a follower of the church of the free market. And they generally strongly disagree with the idea that humans have cognitive biases (because their faith requires it not to be true).
I'm glad someone else laughed at the pro-homeopathy / conspiracy theory around the history of snake oil salesmen content on there.
> on their own self interest).
That's untrue of some schools of economics that advocate free markets, e.g. Austrian.
> But anyone linking to mises.org is probably a follower of
> the church of the free market. And they generally
> strongly disagree with the idea that humans have
> cognitive biases (because their faith requires it not to
> be true).
That's an ... interesting ... claim. Care to justify it?
I took the term "free market 'idea'" to be specifically talking about those for which it's true. That seemed to be the point, and the site linked to was Austrian. Both articles make the assumptions in question about the ability to self-regulate that assumes rational actors. So yes, my statement was not true of all schools, but it seemed like those types of Austrians were not in the scope of the discussion.
> That's an ... interesting ... claim. Care to justify it?
Subjective opinion. I read economics news and neuroscience news because it's interesting. Comment threads, especially here, frequently have two types of subjects that start the vocal libertarians arguing and proclaiming: government regulation and the phrase "humans are irrational".
Example: P.E. certified people should have no problem creating weapons systems for a nation-state at war. Does that make it ethical? Depends on who writes the history books afterward.
Example: P.E. certified people might refuse to participate in experimental, unorthodox methods. But especially in software these often become the runaway successes.
In other words _you_ have to own _your_ personal ethics. You won't be able to point and say "I was just following orders!" The pointy-haired boss who gave the orders isn't going to be able to exonerate you of the guilt. Often he doesn't even congratulate you for "doing the right thing." Maybe he'll fire you or give you a bonus – or join you in prison! – but my point is: it's orthogonal to your personal ethics.
Ethics may sometimes appear to conflict with rapid progress. That doesn't necessarily imply an existential crisis, just a lack of forethought. So many ethical problems arise due to overflowing ignorance / lack of forethought combined with a sudden rash of malice (when it comes time to pay the piper). Ethics are a way of expressing realities about the world that conflict with the general Adam Smithian "enlightened self-interest." I view ethics as meta-enlightened self interest – like how Apple is more than just industry-leading, they carved new niches where no one thought to go.
Engineers (software engineers or otherwise) have untangled things much more complicated than this. It's only overwhelming if it blows up in your face.
Path seems like a classic case of all of the above.
If you believe that standing up for your strongly-held beliefs will get you fired, you should look for a new job _now_. Sure, that incurs the trouble and uncertainty of a job switch, and possibly a pay cut (though perhaps less of that than you think). But if it means that you don't have to be ashamed of what you do all day --- it's generally worth it.
Furthermore, whistleblowers are often unemployed for extended periods of time, due to corporations not wanting to hire them as they could be a liability.
> Who would knowingly submit themselves to a doctor, knowing that they might give you a secondary, curable disease, just to ensure they got paid?
We need them to change their ways, not disappear.
The patent language says it's "a commitment from Twitter to our employees that patents can only be used for defensive purposes." Extending this more broadly would say "a commitment to our employees that the code they write can only be used for non-spamming purposes."
The problem, of course, is that it's pretty easy to tell whether a patent is being used defensively or offensively. Defining spam (or more difficult yet, privacy) is a bit more slippery.
Path is young (and building a service, not a device/OS) and it can be abandoned for something similar since all they're keeping is data. Once you buy a gadget, that's an investment on your part which will make a lot of people hesitant to give it up and the culture it that surrounds it.
And I wouldn't bet that Apple will be able to survive scandal after scandal and still survive unscathed. Cook is no Jobs.
If a developer refused to implement a feature due to their ethics then the company would do the following:
* Move engineer to different project
* Set unrealistic goals/deadlines/expectations
* After engineer fails, voice concern about performance
* Set up performance review and improvement plan
* After causing engineer to fail a second time due to unrealistic expectations, fire them due to poor performance
Even if that engineer writes a blog post, enough has happened between his initial refusal and termination as to make conclusive proof impossible. The discussion will be a he-said-she-said affair as his former employer makes a counter-blog post explaining the engineer's poor performance.
A lawsuit is similarly out of the question as most companies have sufficient funds to cause delays in court, thereby causing you to spend all your money on attorney fees and bleeding you dry.
It's not like hospitals haven't contended with this exact thing for a very long time. The wills of surgeons/doctors and their hospital administrators do not always match up.
"The Turing Oath" is on Github (https://github.com/maxmackie/Turing-Oath/blob/master/README....) and I recommend people contribute and we grow this to become something people recognize.
But I'm not sure if Turing, who is not well known for having had anything to do with privacy is the right person for this oath.
Though admittedly, no technology was involved in the whole matter.
Also, Turing’s wartime exploits involved a breach of privacy in the service of a good cause.
Actually, come to think of it, if viewed in that way, he's the perfect name for an ethics oath regarding privacy. I hadn't considered it that way.
For developers who have undertaken the Oath, the challenge would then be to write the best code, so that it sees widespread adoption. This might potentially make it harder for companies like Path to engage in activities which break the Oath.
Once Congress puts a bill forward to deal with this issue, it will be the beginning of the end for this type of behavior. I'm sure Obama will get behind it as well, as it will help the computer market immensely.
You don't need an official Oath or for the company to know or base their employment on such an Oath. Either way, the bottleneck is the employee drawing attention to the company doing something unethical; no Oath has to get in the way of that. There's already a sub-thread on top-secret/weapons/armaments jobs. What about political ethics? And religious? Marriage / gay rights? Porn? "Sexism?"
Perhaps the guy checked a box that said "Please notify all of my contacts via text message".
Then all the messages went into queue that was delayed a bit.
Then the phone companies converted text messages to voice calls.
What about software for missile guidance? Is that okay by this oath?
Or do we as a community value not texting people at 6AM more than we value not killing people?
If it does, it would never get mainstream acceptance; if it doesn't, but does cover the topic of this post, it will be ethically absurd.
I don't see how this is a straw man; when building a professional code, one has to choose what actions to allow or disallow, and this seems like a topic that would obviously come up. What do you think about this scenario misrepresents the idea of a developer's professional code?
I've never had to actually face the ethical dilemma of developing weapons, but what if the development improved precision on a missile? If we can ignore the question as to whether a missile is ethical or not, developing a better guidance system for a missile will help limit collateral damage, but could increase the "comfort-level" of using the weapon for those who decide such things, therefore increasing overall death/destruction. Utilitarianism is hard, because taking all factors into account is impossible. Kind of like machine learning.
I will never scoff at someone who turns down work for ethical objections, but some people are more pragmatic than others.
Both of Williams examples are really hard to wrap your head around if you accept the situations as presented. They are similar to a Sophies Choice [1]
[0]: http://plato.stanford.edu/entries/williams-bernard/#Day [1]: http://en.wikipedia.org/wiki/Sophie%27s_Choice_(novel)
As a programmer, I don't want a bunch of charlatans in SF to give my career an unsavoury reputation because of these antics.
Some folk tried to create a pacifist version of the GPL[1].
Others are using the RMPL (RobotGroup-Multiplo-Pacifist-License)[2] - basically a MIT license, but with a restriction that bans military projects.
[1] http://arstechnica.com/uncategorized/2006/08/7511/
[2] http://multiplo.com.ar/soft/Mbq/Minibloq.Lic.v1.0.en.pdf
I generally agree with this, which is why I find the idea of a "developer's code" somewhat ridiculous.
FTFY