The problem is silo development, poor senior analysts and
   engineers and yes, poor management. It doesn't matter how
   complicated a system is. If the upper management doesn't have a
   clear overview of needs... ...or if the senior engineers don't
   have a clear overview of their areas.... or if the developers
   are blinded from other aspects of the project and/or are
   prevented through poor communication channels to address
   potential issues before they arise.... ... there's going to be
   problems. We're NOT in new territory here. Management of large
   systems is *difficult* but not impossible. It's true that human
   bias is built into the systems. But good principles will catch
   most of them. It should catch nearly all of them. Look at the
   example they gave of the Volvo that crashed into the people. it
   was an optional module to be purchased that had people
   avoidance. Somebody decided to make that optional. Ball was
   dropped. Check boxes were checked as "ok" that weren't ok. The
   algorithms aren't the problem. it's the companies not crossing
   their t's and dotting their i's. == Me. I have certification in
   business management but that stuff is mostly a joke. It's my
   opinion and I'll happily stand behind it. == I'm sure I can find
   references of other mirroring my opinion. I'm not the first to
   have a systems-view of poor business practices smile emoticon I
   was a systems analyst for a few years but I also had total
   control from the point of receiving the business requirements to
   final product and providing support, as well as training a few
   people when I left the position... which I kinda created tongue
   emoticon [they didn't have anybody like me before I came in grin
   emoticon ] Project management involving people *is* a far more
   difficult affair, but I've seen it done well and I've seen it
   done very poorly. == Well, I'm sure I can drag up some
   references if I poke around. It's an opinion I've had since I
   was in school as a kid and noticing mismanagement by teachers
   and administration. Noticed it at every job. I've noticed it in
   personal relationships and even in running my own business.
   There's a name for it somewhere - I'll have to find it. But the
   'gist' of it is: most of the effective work within organizations
   is done across departmental lines and is not represented in the
   hierarchy or the formal reports nor in the engineering
   requirements. It doesn't REPLACE the documented systems but it
   is SUCH a crucial part of systems that well run organizations
   depend on these 'hidden networks" to function, mistakenly giving
   full credit to the documented structure but in reality it is the
   corporate ethos that sets the success or failure rate.
   Subdividing work *does* work and work well up to a point; it is
   a product of our industrial revolution past and factory work.
   But in complex systems, the lowly programmer needs some informal
   contact with finance. The marketing department needs informal
   connections with the human resources , etc. Then issues
   sometimes be resolved before they begin. A good corporate
   culture can catch errors before they are missed. This seems to
   contradict what I said about about cross t's and dotting i's.
   But it doesn't. The informal, undocumented connections _ought to
   be_ a part of the operations of a company. Part of dotting i's
   and crossing t's is allowing for these informal networks to
   form: It's what brings the hidden concerns up to the point where
   they CAN become documented issues. I'm sure I've read this
   elsewhere as well someplace. I guess it's just become such a
   part of "how I think" about these things that I really don't
   know where it could've come from anymore. But a little thinking
   and I'm sure I could come across some reference materials that
   will let you go beyond some guy's comment (me grin emoticon ) on
   a facebook post smile emoticon == Also, consider the bias of the
   article before it becomes an article of concern: This is an
   OPINION piece he is writing. It is fundamentally no different
   from what I wrote above, except he was published in Forbes and I
   was published on Facebook.
   [1]http://www.forbes.com/sites/kalevleetaru/ If you look at the
   other articles of the author you will see a similar bias. This
   does not invalidate his or validate mine but I rather see them
   as on nearly level footing. == But beyond my solution to these
   issues on a more esoteric scale, I'll switch to something far
   more basic: A PRODUCT should not be released to MARKET without
   being properly checked. These are not algorithms that are taking
   over. These are PRODUCTS being sold to customers. ==

References

   Visible links
   1. https://l.facebook.com/l.php?u=http%3A%2F%2Fwww.forbes.com%2Fsites%2Fkalevleetaru%2F&h=LAQGysGr6