Hear me out.. Data Scientists and Python Module Developers.
Companies have over the past 10 years been fed the Big Data bullshit pitch, and now have a cookie cutter "Machine learning" setup. E.g. 1. Firehose(Twitter data feed) or via GNIP or some similar Data Vendor, 2. Some SQL/nosql way to query the data 3. Some notebook environment for the data scientists to try out medium blogs/stack overflow, and 4. Some ML pipeline batch scoring historical data, real-time scoring single events.
Every data scientist has been following the same Tensorflow/keras/scikit medium tutorials to "Get Value" out of the data they pay for.
The result? The loudest "crazies" of the internet have shaped many of the batch training jobs, and now have a HUGE weight associated to each message. From this side, it feels like it's a self fulfilling cycle of Cancel Culture, with the data scientist's dashboard ending up on some PowerPoint to the sales kids to backup the "right choice" and no one asking why someone's tweet has such a weight
Hear me out.. Data Scientists and Python Module Developers.
Companies have over the past 10 years been fed the Big Data bullshit pitch, and now have a cookie cutter "Machine learning" setup. E.g. 1. Firehose(Twitter data feed) or via GNIP or some similar Data Vendor, 2. Some SQL/nosql way to query the data 3. Some notebook environment for the data scientists to try out medium blogs/stack overflow, and 4. Some ML pipeline batch scoring historical data, real-time scoring single events.
Every data scientist has been following the same Tensorflow/keras/scikit medium tutorials to "Get Value" out of the data they pay for.
The result? The loudest "crazies" of the internet have shaped many of the batch training jobs, and now have a HUGE weight associated to each message. From this side, it feels like it's a self fulfilling cycle of Cancel Culture, with the data scientist's dashboard ending up on some PowerPoint to the sales kids to backup the "right choice" and no one asking why someone's tweet has such a weight
Well that doesn't sound very good. Is there any hope of changing it or should it just get burned to the ground and start over?
Learn to code!
You're way in the deep water. I'm still trying to program simple things on my Raspberry Pi