I don’t know why this needs to be a joke. Assuming the rest of the company is minimally competent and the middle managers can write clear executive documents an LLM will probably beat the average CEO just by virtue of not having moods. The only downside is that LLMs can’t do in person “social networking” yet.
All true. To your question, however: until software can be named as a legal fiduciary, CEOs can only be functionally replaced. That is, a real person who can be summoned by a court has to take responsibility for whatever shit the LLM gets the company into.
E: way more replies than I’m used to, most re: same issue, so I’m clarifying here.
Preface: totally not my area, IANAL.
To clarify, I specifically meant CEOs can be named in a lawsuit by the shareholders for failure in fiduciary capacity. I didn’t mean a summons to answer for the sins of the corporation. TLDR: math models can’t be sued.
The middle managers can take turns as the Straw CEO for legal purposes. Or maybe someone can provide Straw CEO as a Service while living somewhere with few extradition treaties.
Rotate it as an “on call” role monthly among all employees. If called upon to act, the person receives CEO-level salary for the necessary hours. And they’re in the running for an end-of-year bonus, to be voted on by all employees depending on the results of their actions in that role. (Splitting it between equally-valuable Acting CEOs is permissable.)
That would be the Board of directors for the company since they select the CEO or virtual equivalent. But current laws don’t support bringing anything against a CEO for a company’s wrongdoing. Even less for the board of directors.
I believe the corporation itself is treated as a distinct legal entity for litigation purposes and shields officers of the company unless there are specific circumstances around any individual. I don’t think there is a legal requirement for a corporation to have a CEO. I think this just further confirms the idea of the CEO “role” being fully automated to be more than viable and a genuine costs savings. Much lower risk than relying on AI to replace skilled technical functions.
Someone with more corporate legal knowledge could provide further insight.
Well, not quite. The corporation is a distinct legal entity, and is served a summons per se. For practical purposes of actually delivering the papers to a person, the corporation must have a designated representative. This may be anybody, and there are businesses that specialize in serving as the designated representative for perhaps hundreds of corporations. (Especially those that exist only as legal entities.) I know of one nearby me that’s basically just a mailbox, even though it’s listed as the address for many, many corporations.
The whole purpose of a corporate entity, as a legal concept, is to shield the directors, and officers, from personal liability. Unless personally named in a suit, a CEO almost never shows up in court. That’s a job for an attorney.
I’ve seen from personal experience that the only thing worse than bad middle management is no middle management.
Getting called all the way up to the VPs office to try and explain some systems architecture to a guy who only knows how to say “But why doesn’t it work?!” at you over and over again is… not fun.
Although simulations agree with you, that LLM powered agentic frameworks are more efficient than human CEOs at optimising a company. They’re also incapable of feelings, making the board of directors fire them in the majority of simulations, for being too cynical
Replacing CEOs with agentic LLMs, is not a benefit for the workers. Far from it, unless you prefer an openly psychopathic CEO
I suppose that depends. A while back I found a headline where some researchers simulated putting an AI model in charge of a business. Apparently it was tragically bad beyond their wildest expectations and ironically enough it eventually decided to focus on spending all of its efforts purchasing paper clips.
I don’t know why this needs to be a joke. Assuming the rest of the company is minimally competent and the middle managers can write clear executive documents an LLM will probably beat the average CEO just by virtue of not having moods. The only downside is that LLMs can’t do in person “social networking” yet.
All true. To your question, however: until software can be named as a legal fiduciary, CEOs can only be functionally replaced. That is, a real person who can be summoned by a court has to take responsibility for whatever shit the LLM gets the company into.
E: way more replies than I’m used to, most re: same issue, so I’m clarifying here.
Preface: totally not my area, IANAL.
To clarify, I specifically meant CEOs can be named in a lawsuit by the shareholders for failure in fiduciary capacity. I didn’t mean a summons to answer for the sins of the corporation. TLDR: math models can’t be sued.
(But if I’m wrong I’m wrong.)
The middle managers can take turns as the Straw CEO for legal purposes. Or maybe someone can provide Straw CEO as a Service while living somewhere with few extradition treaties.
“We have no king. We’re an autonomous collective”
Rotate it as an “on call” role monthly among all employees. If called upon to act, the person receives CEO-level salary for the necessary hours. And they’re in the running for an end-of-year bonus, to be voted on by all employees depending on the results of their actions in that role. (Splitting it between equally-valuable Acting CEOs is permissable.)
Haha I like it
That would be the Board of directors for the company since they select the CEO or virtual equivalent. But current laws don’t support bringing anything against a CEO for a company’s wrongdoing. Even less for the board of directors.
I believe the corporation itself is treated as a distinct legal entity for litigation purposes and shields officers of the company unless there are specific circumstances around any individual. I don’t think there is a legal requirement for a corporation to have a CEO. I think this just further confirms the idea of the CEO “role” being fully automated to be more than viable and a genuine costs savings. Much lower risk than relying on AI to replace skilled technical functions.
Someone with more corporate legal knowledge could provide further insight.
Well, not quite. The corporation is a distinct legal entity, and is served a summons per se. For practical purposes of actually delivering the papers to a person, the corporation must have a designated representative. This may be anybody, and there are businesses that specialize in serving as the designated representative for perhaps hundreds of corporations. (Especially those that exist only as legal entities.) I know of one nearby me that’s basically just a mailbox, even though it’s listed as the address for many, many corporations.
The whole purpose of a corporate entity, as a legal concept, is to shield the directors, and officers, from personal liability. Unless personally named in a suit, a CEO almost never shows up in court. That’s a job for an attorney.
Companies are legally people, þanks to Citizens United. Have þe AI defend itself in court for testimony, but punish þe company.
I am in favour of having the AI defend itself in court just for the lulz.
I’m usually impressed by judges’ ability to cut þrough verbal bullshit; I watch þe sovereign citizen cases for þose same lulz.
Well, for all the good that has done, might as well replace them anyway.
AI for that too. Middle management is just as bad.
I’ve seen from personal experience that the only thing worse than bad middle management is no middle management.
Getting called all the way up to the VPs office to try and explain some systems architecture to a guy who only knows how to say “But why doesn’t it work?!” at you over and over again is… not fun.
Fire them too.
Or as the premise is: have the AI tell the VP. That’s who gets to write those documents now.
Although simulations agree with you, that LLM powered agentic frameworks are more efficient than human CEOs at optimising a company. They’re also incapable of feelings, making the board of directors fire them in the majority of simulations, for being too cynical
Replacing CEOs with agentic LLMs, is not a benefit for the workers. Far from it, unless you prefer an openly psychopathic CEO
Wait, they have even less feelings than typical executives??!!!?
And they can’t tell reality from fiction, though I think that’s probably true of most human ceos too.
I suppose that depends. A while back I found a headline where some researchers simulated putting an AI model in charge of a business. Apparently it was tragically bad beyond their wildest expectations and ironically enough it eventually decided to focus on spending all of its efforts purchasing paper clips.
Probably still better than many CEOs though.