this is actually deeply irritating to me ive worked at lots of tech companies large and small in data/ml roles you have to understand every one of them is held together with the equivalent of duct tape and fortuitously growing barnacles, no one understands how the systems workhttps://twitter.com/eigenrobot/status/1106374706674847745 …
-
Show this thread
-
in 2015 or so I did some data pipelining at Amazon, the premier cloud data place most of my time was spent building and refining a script that let me work around problems inherent to their *internal* ETL platform go and ask an AMZN veteran about data warehouse they might cry
2 replies 1 retweet 30 likesShow this thread -
every product team has their own algorithms and often the use of these algorithms is to undermine algorithms operating on another part of the site over which a team has no control most of these algorithms were written by people who left the company and left no documentation
3 replies 2 retweets 35 likesShow this thread -
another team I worked on had a production pipeline that consisted of a bunch of R scripts with hard-coded dates that had to be manually updated every month and run in a particular order while manually moving files around this was at a company youve heard of
3 replies 0 retweets 32 likesShow this thread -
and this is all just data work I have no idea what the poor devs and ML engineers are working with but bless them
4 replies 0 retweets 19 likesShow this thread -
ANY TIME you hear about a tech company doing some dumb thing "malevolence from the top" should be your last explanation No One Knows Anything especially the execs! My god you think they have time for this shit "why did your company do this" honest answer: "no idea"
6 replies 4 retweets 56 likesShow this thread -
Replying to @eigenrobot
Might this also apply to governments and perhaps militaries - all large organisations? I sometimes get this impression from military history... Are humans assuming too much agency again?
2 replies 0 retweets 1 like
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.