10.6
Talented people can often create amazing products that work most of the time, but the success of trusted engineering feats like bridges, skyscrapers, medical devices, satellites, etc. is due to much more than talent and creativity.
For software to achieve the same level of trustworthiness, software developers must follow processes that are known to produce dependable software products. One of the core ethical principles set forth by the software engineering profession is that the software product be of the highest quality, so it is incumbent upon an ethical engineer to use dependable practices in any and all cases where dependability is important. A system failure of Angry Birds might not be a huge issue, and I can deal with my personal chromebook chroot failing from time to time, but large and widely used systems that are integrated into society have a much higher standard of dependability.
Just as the scientific process is the best path to good science, a dependable software process is the best path to a dependable product. The process works because it harnesses all of the skill and talent of the developer and constructs a well-defined and thoroughly specified goal so that the developer is aiming for the correct target. The process then runs the product through a variety of automatic and manual check systems that ensure quality and consistency, and finally provides a carefully planned system of validation tests to ultimately demonstrate that the product is as dependable as is needs to be.
Without a dependable process, the developer may get lucky and produce a dependable product, but he or she could not say for certain or argue convincingly that the product should be depended upon by society.
10.10
Though I have not yet been on the cutting edge of any software development, I would nevertheless argue that such complaints are short-sighted. First, I think it is often true that the most creative products come out of narrow constraints. When faced with limited choices, people can often think of problems in new ways rather than resort to the obvious tools that they think of first and start using right away.
Second, I will embrace the old man perspective for a moment to say that old tools often stick around for a reason. One should always be skeptical of established protocol and ubiquitous tools, but the flip side of the coin is that one shouldnot disregard such established methods for the reason that they are old. Often new products are created for the sake of convenience, not for the sake of creating more interesting, useful, or beautiful products. A carpenter can make an equally beautiful table using a hand saw, chisel, bore, and whatever other unpowered tools as another carpenter could make using power tools. It might take a bit longer to use the old tools, but the second carpenter should acknowledge that the benefits of the new tools are not substantive.
Third, assuming constraints applied by regulators actually did limit creativity, what is the benefit of making a creative product that does not work well? A painting or a song do not need to be resilient to system disturbances or resist unwanted intrusions into confidential data, so it is a bit of a false, or at least incomplete, analogy to liken a software product to a work of art in regards to how free the artist should be while creating it. This is the most beautiful app ever created, congratulations! Now get my identity back from the hackers who used it to get into my phone and steal all of my data...