As someone who works in data science, I can't even begin to describe the number of red flags swirling in my head. Even under the best of circumstances, I simply cannot trust a model trained on...
As someone who works in data science, I can't even begin to describe the number of red flags swirling in my head. Even under the best of circumstances, I simply cannot trust a model trained on "activities in Jira, Confluence, Gmail, chats, documents, dashboards..." would do a very good job at determining how engaged an employee was, let alone how vital they were to a company operating. I'm sure there were conversations with managers above these employees and some flagged by the model were spared because they were needed, but everything about this feels upsetting and hypercapitalistic knowing what I do about how poor model fit often is and the kind of person that would allow this kind of analysis to determine the future of their company.
If I'm being frank, this sounds like ruthless capitalism operating under any guise they can find that will allow them to operate without being persecuted too hard. Employers used to justify laying off minorities because they were 'lazy' or 'poor workers' because enough people believed that. In reality they wanted to avoid paying for more workers because they made the decision they could operate without them and pursue greater profit margin, or because the wages for that individual had increased too much over time and they could 'reset' it by hiring someone new at the bottom of the ladder.
I think the explanation is simpler than that to be honest. The CEO had a meltdown and desperately tried finding something/someone else to blame. The layoffs are a convenient way to save money on...
If I'm being frank, this sounds like ruthless capitalism operating under any guise they can find that will allow them to operate without being persecuted too hard
I think the explanation is simpler than that to be honest. The CEO had a meltdown and desperately tried finding something/someone else to blame. The layoffs are a convenient way to save money on the extreme short term and blame poor performance on "employees we've gotten rid of".
It's idiotic to the core, of course. But there isn't a world, no matter how ruthlessly capitalistic, where this makes good business sense.
As someone who works in data science, I can't even begin to describe the number of red flags swirling in my head. Even under the best of circumstances, I simply cannot trust a model trained on "activities in Jira, Confluence, Gmail, chats, documents, dashboards..." would do a very good job at determining how engaged an employee was, let alone how vital they were to a company operating. I'm sure there were conversations with managers above these employees and some flagged by the model were spared because they were needed, but everything about this feels upsetting and hypercapitalistic knowing what I do about how poor model fit often is and the kind of person that would allow this kind of analysis to determine the future of their company.
If I'm being frank, this sounds like ruthless capitalism operating under any guise they can find that will allow them to operate without being persecuted too hard. Employers used to justify laying off minorities because they were 'lazy' or 'poor workers' because enough people believed that. In reality they wanted to avoid paying for more workers because they made the decision they could operate without them and pursue greater profit margin, or because the wages for that individual had increased too much over time and they could 'reset' it by hiring someone new at the bottom of the ladder.
I think the explanation is simpler than that to be honest. The CEO had a meltdown and desperately tried finding something/someone else to blame. The layoffs are a convenient way to save money on the extreme short term and blame poor performance on "employees we've gotten rid of".
It's idiotic to the core, of course. But there isn't a world, no matter how ruthlessly capitalistic, where this makes good business sense.
Good discussion on HN: