4 ms·
> I wholeheartedly disagree. It is trivial to generalize. Careful observation resulting in nuanced findings is difficult precisely, because we tend to crave sim
by zmxz 4y ago
> I wholeheartedly disagree. It is trivial to generalize. Careful observation resulting in nuanced findings is difficult precisely, because we tend to crave simple answers.
Generalizing isn't always a bad thing, but this situation really is trivial to observe. There aren't too many possible outcomes.
If (boss + worker) make money, (boss + worker) keep the job.
If (boss + worker) don't make money, someone (or both) lose(s) the job because there's no $$ to pay them.
It's extremely trivial because it all revolves around upkeep. Can you keep up paying thousands of people (pretty large) salaries over an undefined time period? If yes = they keep the job, if not = they don't.
Most things are indeed trivial, especially in IT sector and employee compensation. I will accept we agree to disagree and I'm perfectly fine with it, but the wave of layoffs in IT sector is because of lack of money to pay salaries. The whole cause why there's no money, why work isn't optimal or why people aren't managed correctly is a different matter all together and plays a role.
The bad part is that IT workers refuse to accept modicum of responsibility and consequences and are using social media to present the "higher ups" as some kind of scheming, evil people who gather together around Stonehenge once a month and discuss new ways how to spite and hurt people. It's just silly.
- A4ET8a8uTh0 4y ago<< I will accept we agree to disagree and I'm perfectly fine with it, but the wave of layoffs in IT sector is because of lack of money to pay salaries. I was with you. I was genuinely nodding along ( including the other comment ) until this portion. That is a hard no. Notice how there is seemingly simultaneously not enough money in the budget for tech workers ( I do not want to limit this conversation to IT support, which is what many equate IT with -- did the use of that acronym change ? ) while there is indeed money for MS executives to enjoy the show[1], money for bonuses[2] and.. I was going to end with a joke but decided against it. I am genuinely unsure how you can reasonably argue otherwise with a straight face. [1]https://www.businessinsider.com/microsoft-execs-private-sting-show-davos-before-mass-layoff-announcement-2023-1 https://www.businessinsider.com/microsoft-execs-private-stin... [2]https://thrivetrm.com/executive-compensation-trends-benchmarks-a-zoom-in-on-mid-year-2022/ https://thrivetrm.com/executive-compensation-trends-benchmar...
- zmxz 4y ago> until this portion. That is a hard no. Ok, I accept your argument as valid. There are sufficient funds that could be used for salaries. Let's imagine for a second you are hiring me. You have $40 in bank and I cost you $1 a year. In past 5 years, I never made you a single $. I just cost you $1 a year, with slight increases. You paid me $5 so far. Would you keep paying me $1 a year or would you rather invest (the $5 you'd spend on me in next 5 years) into this new thing called AI, hoping you might get $10 out of it? Purely looking at numbers, deciding objectively and without emotion - what seems the best for you?
- A4ET8a8uTh0 4y agoIn the current state of AI? I guess it would depend heavily on the job at hand. My concerns with current iterations are privacy ( how do I know sensitive data ingested to feed the model won't be re-used ala co-pilot or gpt for other entities and some information should stay where it is ) and errors which necessitate human babysitting anyway. There is a clear benefit to automation ( in fact, its usually my spiel ), but it is going to happen anyway so I am not sure what the argument is in this instance. Would you be willing to elaborate?