[trustable-software] Dark Patterns for unethical user interfaces

[Paul Sherwood]

On 2016-07-31 17:51, Colin Robbins wrote:
> Hi Paul,
>
> This would seem like a good start on a code of ethics for software
> engineers.
> http://www.acm.org/about/se-code [1]
>
> Not had any experience of it.

Thanks for this.

The content is clearly well-intentioned, and seems to cover the ground 
pretty thoroughly. However I read it a couple of times, and felt vaguely 
uneasy but couldn't figure out why.

After letting it soak into my subconscious, I realised:

- if this was in a workplace, many people would just scan through and 
say, 'oh yes, fine, whatever' and then proceed as before.

and maybe more crucially

- expecting engineers to 'moderate the interests of the software 
engineer, the employer, the client and the users with the public good' 
is like asking the tail to wag the dog.

I've emailed ACM to ask if it's possible to re-use or adapt their text.

Now I'm thinking about how to address the issues above. It seems to me 
the only way to be confident that a code actually works and is enforced 
is to

- have it expressly driven and prioritised from the top of an 
organisation
- and ensure transparency so that those who find themselves pressured 
(or incentivised) to cheat can benefit from others outside, who can 
correct the situation by whistleblowing without being influenced by the 
pressure/incentive.
- and see some evidence of actual instances of the corrections.

I wonder whether ACM has any hotline/process for whistleblowers, or any 
evidence of actual incidents, but obviously this is not the place to 
ask.

> On the security side, most professional bodies have a code of ethics
> that members have to subscribed to. It's fundamanal in the securty
> world, as the line between ethical and unethical behaviour is very
> grey.

Very true. And what is understood to be ethical by one group may be 
unacceptable for another. Maybe the most we can hope for is that the 
rules for a given group/organisation are clear, that they are followed 
by default, and that there are mechanisms to spot and address situations 
when they are not.

br
Paul