Who is Irresponsible?

  • Comments posted to this topic are about the item Who is Irresponsible?

  • Technology runs ahead of what CTOs and architects can keep up with.  By the time they have assessed it, the cat is out of the bag.  It is very hard for a founder, CTO or architect to be on the front foot.

    Beyond policies, developing mechanical safeguards is time-consuming as well.  Then there is the method of publishing any policies you have.  An email just doesn't cut it.  The briefing process I have experienced are

    1. Senior tech staff briefed in a special meeting of seniors.  This is a forewarning, so we are prepared for questions from juniors
    2. All hands tech staff briefing
    3. Team session to take questions
    4. 1:2:1s

    Only then does the final draft of the policy get circulated to everyone.  You will have been in at least 3 sessions before receiving it in writing.

    I think Software Engineers get mixed messages, and those messages are not always timely.  Then there's the whole "It is easier to get forgiveness than permissions" thing.  Again,  there's the turning of the blind eye, provided a successful delivery is made.

    I'm not sure how much code in businesses is truly sensitive.  I could understand control systems for new products, but not business code.

    The cynic in me says that people are more worried about their code turning up on TheDailyWTF or the Register's Who Me pages.

     

  • I don't have an X (formerly known as Twitter) account, so I can't login to read the replies.

    However, it seems to me like the founder himself is maybe trolling:

    ".. No wonder 90% of American startups fail if they cut corners like this .."

    My first question would be whether or not the company has an internal policy about not using AI tools. If not, then executive management should not be surprised that his developers would be using AI in their daily workflow.

     

    "Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho

  • It is certainly hard to track and ensure everyone knows policy. Multiple meetings help. Even in Redgate, with 600 people, I see this. People set a rule, they think they've told people in Slack and/or Email and are surprised not everyone knows about it

  • Where I work, I use CoPilot hosted within our corporate Office365 subscription, which is sandboxed by the organization's identity permissions and data governance controls.

    I can see why corporate policy would forbid sending data and source code to 3rd party general AI, and IT can block access to that full stop if they want (in addition to making it their policy not to use it).

    The X post in question didn't go into detail about what exactly happened, so it's hard to have a discussion about whether firing the employee was justified. If the company founder wants to forbid all usage of AI generated code, then he's painting himself into a corner going forward. I hope he's providing the development team with a managed and approved option.

    "Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho

Viewing 5 posts - 1 through 5 (of 5 total)

You must be logged in to reply to this topic. Login to reply