30
votes
The latest AI use cases appear to be built specifically for managers and executives, and literally nobody else
Link information
This data is scraped automatically and may be incorrect.
- Title
- Is AI Being Built For The Manager Class Alone?
- Published
- Jun 1 2024
- Word count
- 1730 words
I find this whole idea absolutely hilarious:
The person that has this end goal thinks this will make them 5x as productive when in reality it would just mean that their job is no longer required (for whatever definition of "required" exists for someone that is spread so thin that they don't have any real context for their current decisions).
As dense as managment can seem at times, they are aware of the threat.
This is somthing that really worries me about contemporary corporate culture. I've probably said this dozens of times over several sites already. I just looks like all managerial schools teach in a way that treats every company as the same company. And to keep the MBA industrial complex going, it has been loudly communicated to shareholders and students that MBAs are a golden ticket to success.
Now it's easy to replace executives/managers with data models because these companies have been run by the same algorithm for almost 2 decades. There's value in leadership having extensive vertical knowledge of the company. People with real product/process experience, customer insight and empathy for workers.
Some of my favorite stories are people improvising or innovating thier way out of seemingly impossible problems, because they understand thier job.
But in leiu of that, leadership and decision making is done by quarterly KPI, tend chasing and spreadsheet. And the second managers demonstrate any success, they can take a step up at the next place. I think that's why every company is falling into the same strategies of shrink(+in)flation, bloating costs, ehshitification, whaling and layoffs. It still blows my mind how so many major game studios managed to make the exact same mistakes and it's no coincidence.
It was really apparent in creative spaces but now we're at the point where parts are popping off planes, McDs is a luxury brand and being asked to tip a billion dollar company.
The typical online discourse around management is that they're totally useless, they're generally stupid, and the roles only exist to keep incompetent people rich.
I'm not of that opinion whatsoever. I think that management is a real skill, and that skill is just as difficult to do well as software programming, mechanical engineering, or any other hard technical skill that is generally respected online. However, that's with the caveat that I don't think you can effectively manage a technical company without deep experience in that field.
I know a lot of great non technical managers. They're able to delegate, trust their experts, and lead very effective teams. They don't, however, have a feel for strategy and the possibilities of what their teams are capable of.
I've never met a non technical manager that has a realistic grasp for what is, and isn't possible, and the most common issue they have is that they'll have a pie in the sky idea that may or may not be possible, their engineers tell them it isn't, because that's generally what engineers do, and they effectively have two choices here. They can believe their engineers, which is the safe bet, life carries on as usual, and everything is fine in the short term. Long term, you're not improving, and you're being left in the dust by the rest of the industry. Or B, you can call bullshit every time you don't get your way. This will piss off the people that work for you, you'll lose a lot of good talent, but you'll be right by dumb luck a lot of the time too. Steve Jobs made apple the most profitable company in the world this way. The issue is that most people who work for you will hate you, you'll eventually build up a company culture built on a cult of personality, and your company will become stuck in a incestuous pattern where the only good way to do things is the way you've decided to do things.
An effective technical leader has worked in the trenches and has a good idea of what is and isn't possible. When his engineers say it can't be done, they can actually use some finesse to call BS to that and back it up with a theory that can be tested. They see the possibility of the technology they're working with and it's potential applications. The issue with just promoting engineers to managers though, is an opposite problem. Even though there are plenty of engineers with great interpersonal skills, contrary to the stereotype, engineers chronically suffer from a form of organizational myopia. They tend to be more concerned with the elegance of the product, the rigor of the process, and the technical achievements of the team rather than the perception the product actually has in the marketplace. Being a manager that started as an engineer, it's something I suffer from and constantly have to consciously battle. Managers from a business background typically don't have that problem. Their interest in a piece of technology is as a product, first and foremost. They don't really care about the technical aspects except as a means to attract and retain customers. That gives them an edge because ultimately, their view is the more correct one if considered through the lens of the company's priorities.
If you can get a manager with an engineering background that is effectively able to resist that urge though, you have someone who is almost guaranteed to be successful.
People with that kind of vision plus the soft skills needed to actually manage people effectively are some of the rarest and most valuable people in the world.
The typical strategy of "throw a high performing MBA at it" probably isn't as effective as business schools would have you believe though.
It also kind of plays into an observation I made in a brush with management earlier in my career: being a mediocre manager is easy. Attend a bunch of meetings in which it's not required that you know anything or demonstrate any particular insight, do a bunch of paperwork and rote administrative tasks you're never meaningfully evaluated on, and… don't get caught doing lines in the bathroom, I guess? In a lot of cases, the org doesn't seem to have any way to evaluate managers outside the performance of the group they're managing, which for mediocre managers is almost completely divorced from their actual management.
As far as I can tell, this tendency extends all the way up to the executive, though I'll freely admit I've had relatively little personal contact with the C-suite of larger companies.
If there's one class of jobs LLMs could actually plausibly replace, this is it, though I'm just guessing it won't actually play out that way.
(Being a good manager is extremely challenging, to be clear. Actually carrying out the ostensible responsibilities of the position is a ton of work—including a lot of emotional labor and some extremely unpleasant responsibilities—even if your employer never evaluates or rewards you for that work.)
It's funny from the perspective that a person who would want to overbook meetings in this way doesn't have much imagination. It's like telling someone from 200 years ago about cars, and they would want to know if you could fit a horse in the car.
Meetings are a required thing because there is friction in communication between people in a group. If AI was useful, it would remove the need for meetings by helping give the information to everyone as needed at their own pace. It wouldn't be some bot sitting in a zoom chat.
I suspect what it really means is that everyone can pretend to be in all the meetings all the time, and get on with some actual work. At least until being in pointless meetings ceases to be a form of presenteeism and we find something new to crush people's souls with.
After reading Manna yesterday, this sounds suspiciously familiar...
Aaah, thank you. I've been looking for this story for ages. I remember it making quite the impact when I originally read it because it makes a great point. For some jobs, the 'low skill' menial tasks are HARD to automate. You need a machine with eyesight, perception, balance, etc. On the other hand, the 'high skill' tasks can be automated with a cell phone if the AI is good enough.
Seems like this future is becoming a reality. People will walk and climb and crawl and flip things while receiving orders from machines. Seems like our last competitive advantages to fall to automation will be the lower level processes our brain has 'automated'.
To do what he wants it'd need critical thinking skills, it'd have to be AGI. If we had AGI, what do we care of meetings? We're at the point of choose a future where we either don't have to work, are forced to work because some ass convinced people humans require work to live, or we can't work and are starving because capitalists won and AI's have all the jobs with the rich owners at the top.
In my industry, community financial institutions, the focus has been on improving processes. With limited resources in small companies, it's a perfect place to implement AI to give us more time back in our day to focus on the important things. The other area is customer interactions, everyone is all over the AI customer service wagon. I have seen some really cool stuff to change my mind about talking to an actual human.
We have been using MS Copilot for meeting notes and next steps, it has been a life saver. We can focus on the meeting and It helps us keep track of all items discussed. The other uses in the other MS products are great as well. It can really build a nice PowerPoint from an outline.
My first thought: "You've reached elight's life model decoy..."