
Well, a PHB *is* controlled by a trained neural net.... Trouble is, so am I....
Nearly half of US office workers expressed concern that AI might take their jobs in a February survey by investment banking biz Jefferies. Bankers aside, managers may want to worry, too. Researchers at ESMT Berlin argue that AI can help manage research projects, allowing them to operate at greater scale and efficiency than …
"what if organizations shared worker metrics data from AI management systems and that influenced future hiring decisions involving said workers"
In a world where installing a tattletale in your car to radio your insurance about how you drive, supposedly to lower your insurance bill, is becoming common practice, that is not a what-if scenario.
It is blindingly obvious that any AI management system will automatically and almost immediately be taken by upper management as criteria for advancement, not to mention continued employment.
Whether that will be a good thing or not, well, I guess that depends on what your current management feels like to you.
[quote]Whether that will be a good thing or not, well, I guess that depends on what your current management feels like to you.[/quote]
where I am presently, it would be a rapid race to see which side can stab the other in the back first :o)
and, as we DO outnumber them, maybe our results would hold greater weight, in which case, so long middle management :o)
but, as in all cases, they will probably have some form of control over it all, making this but a wish
So they've looked at some fairly basic efforts and drawn a conclusion.
Not quite so sure I follow their conclusion that
a) any of this stuff actually works; the human efforts often don't, so a crude AI approximation seems unlikely to
b) that turkeys will vote for Christmas & let these tools get a foothold instead of making sure they go nowhere. You don't actually expect managers or HR to eliminate their own jobs do you?
"You don't actually expect managers or HR to eliminate their own jobs do you?"
Middle managers and HR drones also have bosses, all the way up to the top. It's only the C-level and board who don't really answer to anyone else (bar their own infighting and backstabbing) who would never replace their own jobs, as far as they are concerned, anyone below them (ie everyone) is fair game.
On a more pragmatic note, I find that vast layers of middle management are a prime indicator of a poor company. Flatter company structure allows better communication and more flexibility. What typically limits a flatter structure is how many direct reports a manager can handle. I would think that having a good manager (all the manager jokes notwithstanding, every company has a few of these) with access to a precisely-scoped AI assistant / AI tools could effectively manage larger teams, allowing larger companies to have the benefits of a flatter structure
The "researchers" have concluded that the middle management can now be done by AI. Unfortunately, the lowest ranks of workers have already been written of as replaceable by AI, which means that the middle management would have been left with nothing to do anyway but try to manage the AI, so it's only logical to let the AI manage itself and get rid of the managers too. That leaves nobody but the executives - who will be found to be less able to look at the big picture and make the best strategic decisions than the AI. So they'll have to be replaced too. The end result is that the company now consists of nothing more than a bunch of virtual machines in a cloud somewhere telling each other what to do, generating lots of activity and producing nothing.
Pfffrrrrrt! Their paper is deliberatly miss-titled to attract press attention IMHO. Instead of:
"Algorithmic Management in Scientific Research"
a better title would have been (based on the 16 projects listed in "Apendix 2" of their paper, through the SSRN link):
"The Potential for Algorithmic Management of Crowdsourced Projects".
There's no "Albert Einstein"-style scientific research in their analysis -- it's more stuff one gets high-schoolers (and their parents) to help out with on their ample free time, for free (to keep them away from drugs, among others). Valuable, but not quite the unification of the laws of physics.
Still, the prospect of replacing bosses with more broadly sympathetic AI, with its inherent potential for prompt-engineered hackorama, is just delicious (polish-up on those black hat skills!)!
we could neutralise the AI powered manglement by introducing the AI powered BOFH I mentioned yesterday.
Although the thing is.... is that they begin to suck in more and more computing power as they try to outwit the other, until they've spread to every computer connected to the internet and brought down the world we've built for ourselves......
MBAs are predicated on the theory that any manager can manage anything, they don't need to know any details of the work being managed and only need to pay attention to the "big picture".
Current generative LLM AI is awful at detail, but good to very good at the vague handwavey "big picture" stuff.
Thus, if MBAs are valid, anyone with an MBA can be trivially replaced by a few GPU cores.
It's mostly puff piece, scaremongering but there's always a ring of truth to these things.
Not only have they f'ked with housing, f'ked with cost of living, educaiton is f'ked and now there's no hope of a decent, well paid job anymore as a manager and there are good ones. It's no wonder kids aged 18-25 are now the biggest anti-depressant consumers on the planet! They're basically f'ked before they're even born.
I count myself lucky that fingers crossed, I'm now only 8-10 years max before I'm done and heading to retirement, get my parole from the looney bin that is modern life.
"AI predicts the probability of participant disengagement and provides interventions (i.e., messages) to increase user motivation," the paper explains. "The AI balances the trade-off between messages sent too early (disrupting the workflow and addressing a problem that was not acute yet) versus messages sent too late."
This portrays good management as being the optimized timing of messages of encouragement. That is a ridiculous. And really, that task could be done with an Eliza level of technology.
At least the following is one small part of managing of a creative team: Getting a clear picture of what employees are doing by asking them to explain it and the problems they face - and in the process the employees get clearer picture of their own roadmap, and often end up answering their their own questions to their own stumbling blocks.
I could see how an interactive chatbot could maybe help with that small part when used to by an employee to assemble reports - by asking questions to clarify what's being said. Notice this is not generative but rather introspection assistance. Oddly it is not a feature I have seen promoted anywhere in the AI hype syllabus. Probably because it implies that the real source of creativity is the employees themselves rather than the AI, and that's not something that appeals to AI pushers.
A modern day chatbot does not offer the same quality of knowledge based suggestions and insight that an experienced manager should be able to offer. It does not capture the overall team strategy and direction. It does not evaluate except using the most superficial criteria.
"what if organizations shared worker metrics data from AI management systems and that influenced future hiring decisions involving said workers"
Well, I imagine these organisations better have a really large, really good, and very non-AI powered legal department, and enough money to cover getting slapped with GDPR based lawsuits so hard, they won't know their arses from their elbows any more.
Many, many years ago when the "threat" was offshored work, I was hired software development help for a large multi-national company. A colleague who was more senior in the process and more focused on requirements gathering and user liaison mocked me a little and suggested that my job would be disappearing to a different continent. I pointed out that spreading development across the planet was hardly sensible and asked him to explain why such a large organization with offices on virtually every continent, should want to see development in Asia and requirements gathering in Europe. His face fell as it dawned on him that I could be right.
Society needs to address this outsourcing of jobs to AI because unless the employers are forced to compensate society, we will start to see a lot of people competing even for the menial low paid work and their bosses.
I mean we didn’t give a shit when entire towns and cities were left destitute by offshoring industrial labour, leading to huge spikes in suicides, violent crime, reliance on benefits, and generational poverty that still haven’t gone away 40 years later in some parts; why should things be different this time?
And why *should* things be different this time?
int main(enter the void)
...