Conversation

Notices

1. emilygorcenski (emilygorcenski@indieweb.social)'s status on Friday, 23-Jun-2023 13:56:57 JST emilygorcenski

   It turns out, actual engineering is a lot harder than software engineering, and this is something you should keep in mind as people start to claim ChatGPT can do a better job.

   https://www.theverge.com/2023/6/22/23770260/titanic-tour-titan-submersible-catastrophic-implosion-oceangate

   • emilygorcenski (emilygorcenski@indieweb.social)'s status on Friday, 23-Jun-2023 13:56:50 JST emilygorcenski

     in reply to

     When you learn about engineering failures in school you quickly understand that it’s not that people were unintelligent, it’s because they were careless. And good engineering practice builds layers to make sure that carelessness doesn’t propagate.

     The rigor of engineering is not the technical complexity of the problem space, it’s everything else around it.

     In other words, literally everything the silicon valley grindset devalues.

   • emilygorcenski (emilygorcenski@indieweb.social)'s status on Friday, 23-Jun-2023 13:56:51 JST emilygorcenski

     in reply to

     This I find troubling because it presumes that defects can be resolved through sheer skill, despite the fact that the software industry’s own research shows this not to be the case. Nevertheless, the myth persists.

     You bring this mindset into building a submarine and people are going to die.

     Machismo repeated this.
   • emilygorcenski (emilygorcenski@indieweb.social)'s status on Friday, 23-Jun-2023 13:56:52 JST emilygorcenski

     in reply to

     In Engineering school, regardless of your discipline, you’ll be taught about the time someone fucked up and killed people. Software has these studies, too, but far fewer of them, and the culture of the field is completely opposite from what we see in other fields.

     Every so often I talk about THERAC-25 and ask if people know about it. A lot of people do, and the response is usually something like, “oh yeah the software was bad.”

     The mindset is that it was lack of competence, not lack of care.

   • emilygorcenski (emilygorcenski@indieweb.social)'s status on Friday, 23-Jun-2023 13:56:53 JST emilygorcenski

     in reply to

     It’s sad, because the former billionaire CEO of OceanGate had more than enough capital to do it the right way. But today’s billionaires pull themselves up by moving fast, eschewing rigor and expertise, and discarding all of the cultural philosophies required to make an engineering team work right.

     Doing things rigorously doesn’t earn you the big gains today’s billionaires seek. This sub design was simply virtue signaling to other billionaires.

   • emilygorcenski (emilygorcenski@indieweb.social)'s status on Friday, 23-Jun-2023 13:56:54 JST emilygorcenski

     in reply to

     This is not a flex.

     This is saying that the management practices and work culture that we have developed for software development, which are perfectly suited for that space, are equally in magnitude unsuited for designing physical systems.

     Every aspect, from testing to user experience to documentation and more.

   • emilygorcenski (emilygorcenski@indieweb.social)'s status on Friday, 23-Jun-2023 13:56:55 JST emilygorcenski

     in reply to

     Nothing, no single thing I have done in the entirety of my career in software has matched the difficulty or rigor of my third year coursework in my aero program, which was less complex than what I did professionally.

     The only aspect of software development that came close was safety critical software, which unsurprisingly follows standards that emerge from electrical engineer practices.

     Modern software development lacks the rigor, depth, and complexity of physical engineering.

   • emilygorcenski (emilygorcenski@indieweb.social)'s status on Friday, 23-Jun-2023 13:56:56 JST emilygorcenski

     in reply to

     I’m not going to exhume the old ridiculous debates over the word “engineer,” but I am going to say this. I studied aeronautical and mechanical engineering and computational mathematics in university, worked as an aeronautical engineer for several years, and have also worked in software for many years.