> while quoting an HR executive at a Fortune 100 company griping: "All of these copilots are supposed to make work more efficient with fewer people, but my business leaders are also saying they can't reduce head count yet."
I'm surprised McKinsey convinced someone to say the quiet part out loud
> Software vendors keen to monetize AI should tread cautiously, since they risk inflating costs for their customers without delivering any promised benefits such as reducing employee head count.
That's easy. Reduce the headcount first, and then let the remaining team of poor and desperate, I mean, elite engineers and support teams <buzzword for use> AI for <more buzzwords for make dollars go up> /s.
When will boards replace executive leadership with AI? If Return to Office taught us anything, it was that we already need a couple, and the rest of them copy and paste. Well, AI can do that! Also /s, but maybe just 50%.
What most companies and CEOs fail to grasp is that with all the talk of headcount cuts from AI customers are expecting that AI will LOWER pricing and costs, not raise it. Challenge is that the cost cutting story is mostly vaporware (as many other studies have shown) so CEOs are in a tough spot. They can’t both boast to shareholders at how much cost savings they got from rolling out AI and then charge customers more.
All this is pretty textbook setup for how this bubble finally implodes as companies fail to deliver on their AI investments and come under fire from shareholders for spending a ton with little return to show for it.
AI in its present form is probably the strangest and the most paradoxical tech ever invented.
These things are clearly useful once you know where they excel and where they will likely complicate things for you. And even then, there's a lot of trial and error involved and that's due to the non-deterministic nature of these systems.
On the one hand it's impressive that I can spawn a task in Claude's app "what are my options for a flight from X to Y [+ a bunch of additional requirements]" while doing groceries, then receive a pretty good answer.
Isn't it magic? (if you forget about the necessity of adding "keep it short" all the time). Pretty much a personal assistant without the ability of performing actions on my behalf, like booking tickets - a bit too early for that.
Then there's coding. My Copilot has helped me dive into a gigantic pre-existing project in an unfamiliar programming language pretty fast and yet I have to correct and babysit it all the time by intuition. Did it save me time? Probably, but I'm not 100% sure!
The paradoxicality is in that there's probably no going back from AI where it already kind of works for us individually or at org levels, but most of us don't seem to be fully satisfied with it.
The article here pretty much confirms the paradox of AI: yes, orgs implement it, can't go back from it and yet can't reduce the headcount either.
My prediction at the moment is that AI is indeed a bubble but we will probably go through a series of micro-bursts instead of one gigantic burst. AI is here to stay almost like a drug that we will be willing to pay for without seeing clear quantifiable benefits.
Developers are hit too, although I don't expect that anyone will be replaced. I think AI is a productivity boost, it just takes less times to solve the small problems and get reasonable advice for aynthing beyond. Perhaps it reduced required headcount to implement some features. But companies that expel their knowledge workers for some AI solution probably won't survive long. Those that understand the tooling advantage, will get ahead though.
I love AI image generation, but many certainly do not enjoy the results. I can see some people skimping on paying artists.
First I thought translators would be hit hard by AI, but you probably still need them as well to be decently sure about correctness.
And it remains true that any creativity produced by AI is basically still just a function of the creativity of other people.
I hadn't ever tried Notion before but I sort of vaguely understood it was a nice way to make some documentation and wiki type content. I had a need for something like a table that I could filter that I would normally just do in Google Sheets. So I go check out Notion and their entire site is focused on AI. Look at what this agent can do, or that. I signed up and the entire signup flow is also focused on AI. Finally I was able to locate what I thought was their core offering - the wikis etc. And ended up pretty impressed with the features they have for all of that.
Now maybe Notion customers love all these AI features but it was super weird to see that stuff so prominently given my understanding of what the company was all about.
>Consultant says software vendors risk hiking prices without cutting costs or boosting productivity
From what I know of the firm, it looks like clients have come to the right place if they want a consultant with great experience at hiking prices without cutting costs or boosting productivity.
> Many software firms trumpet potential use cases for AI, but only 30 percent have published quantifiable return on investment from real customer deployments.
McKinsey has pitched my company on projects where their compensation is entirely outcome-based — for example, if a project generates $20 million in incremental revenue, they would earn 10% of that amount.
I have to admit, the results they demonstrated — which we validated using our own data — were impressive.
The challenge, however, is that outcome-based contracts are hard for companies to manage, since they still need to plan and budget for potential costs upfront.
So even when you have measurable benefits - it's still not so easy either.
EDIT:
To clarify the issue — companies are used to budgeting for initiatives with fixed costs. But in an outcome-based contract, the cost is variable.
As a result, finance teams struggle to plan or allocate budgets because the final amount could range widely — for example, $200K, $2M, or even $20M — depending on the results achieved.
Additionally, you almost then need a partial FTE just to manage these contracts to ensure you don't overpay because the results are wrongly measured, etc.
None of these challenges are insurmountable, but it's also not easy for companies either.
- The moment AI is actually good enough to replace us, it will also be incredibly easy to create new software/apps/whatever. There could/would be a billion solo dev SAAS companies eating the lunch of every traditional tech org.
- People (Executives) seem to underestimate just how much of the work is iterating and refining a product over a long time. Getting an LLM good enough to complete a Jira task is missing the point.
- IMO LLM's are also completely draining the motivation of workers. A lot of software devs are intrinsically motivated by solving the problem. If your role is being watered down to "prompt the chat bot and baby sit what comes out", the motivation disappears. This also absolutely destroys any of the creativity/discovery that comes out of solving the task hands-on.
We have to accept that sometimes technology that was envisioned to change the future one way, may be beneficial in other ways instead - and that's okay. We are very clearly still in the phase of "throw AI at everything and see where it is useful." For example, just yesterday I was sent a contract to sign via DigiSign. There was a "Summarize contract with AI" button. Having read the contract in full, I was curious how good the summary would be. The summary was very low fidelity and did not go into the weeds of the contract and I would be essentially signing the contract blind. Although AI is pretty good at summarizing key point of things like articles and conversations, this was a very poor use case imho. But hey, they tried it and hopefully see it is a waste. Nothing wrong with iterating we just have to converge on acceptable use cases.
I am curious if the timing have impacted the inability to measure a benefit. AI is rolling out at the same time as widespread return to office campaigns. Remote work was widely studied and touted as improving efficiency, but no one is showing the drop for RTO. Is AI in part just balancing it out? There's also an ongoing massive brain drain. Many companies are either laying off their most tenured and competent employees, or they are making life miserable for them in the hopes that they quit.
All of this said, using AI in your back end takes a huge amount of time from your users and employees. You have to vary multiple prompts, you have to make the output sane, touch it up, etc. The most useful part of AI for me has been using it to learn something new, or push through a task that I otherwise couldn't do. I was able to partially rewrite a logging window to reduce CPU use significantly. It took me over two weeks of back and forth with AI to figure out a workable solution and implement it into the software. I competent programmer probably could have done it better than I did in less than an hour. There's no business benefit to a help desk person being able to spend 2 weeks writing code that an engineer would be much better suited to handling. But maybe that engineer could write it in 10 minutes instead of an hour if they used AI to understand the software first.
This is because they're trying to reduce the wrong headcount. The largest inefficiencies in corpo orgs lie in the ways they organize their knowledge and information stores, and in how they manage decision making.
The rank and file generally have a really good grasp on their subset of the domain -- they have expertise and experience, as well as local context. Small teams, their managers -- those are the ones who actually perform, and deliver value.
As you move up the hierarchy, access to information does not scale. People in the middle are generally mediocre performers, buried in process, ritual and politic. In addition to these burdens, the information systems do their best to obscure knowledge, with the usual excuses of Safe and Secure (tm) -- things are siloed, search does not work, archives are sunsetted, etc.
In some orgs tribalism also plays an outsized role, with teams acting competitive, which largely results in wasted resources and seven versions of the same failed attempt at New Shiny Thing.
Then as we look higher yet in the hierarchy, the so-called decision makers don't really do anything that cannot be described as "maximize profit" or "cut costs", all while fighting not to get pulled down by the Lord of the Flies shenanigans of their underlings. They are the most replaceable.
A successful "AI Transformation" would come in top-down, going after the most expensive headcount first. Only truly valuable contributors would remain at that level. Organizational knowledge bases would allow to search, analyze and reason about the institutional knowledge accrued in corporate archives over the years, enabling much more effective decision making. Meanwhile, the ICs would benefit from the AI boost, outsourcing some menial tasks to the machine, with the dual benefit of levelling up their roles, and feeding the machine more context about the lower-level work done across the org.
I shared this recently with regard to teams cross training.
I worked in tech support at a big company long ago. Tech support, sales, and engineering used to have a week (for each employee) where we would leave our team and follow the other team around.
It provided incredible efficiency. I now knew what sales was talking about when they called me, they understood how I worked, engineering and I got along so well they used to invite me to their team when they had lunch catered.
Who didn't we need anymore? The middle managers between the groups who brokered what info each group could see, and how we communicated among groups. We solved problems before they started all on our own.
The middle managers won in the end, ending the cross training, too costly they said, but I think they realized that we just didn't need them / weren't engaging them anymore...
> For every $1 spent on model development, firms should expect to have to spend $3 on change management, which means user training and performance monitoring
I think the general point here is true, but it's also brilliant framing from a company selling consulting services.
38 comments
[ 3.2 ms ] story [ 42.4 ms ] threadI'm surprised McKinsey convinced someone to say the quiet part out loud
That's easy. Reduce the headcount first, and then let the remaining team of poor and desperate, I mean, elite engineers and support teams <buzzword for use> AI for <more buzzwords for make dollars go up> /s.
When will boards replace executive leadership with AI? If Return to Office taught us anything, it was that we already need a couple, and the rest of them copy and paste. Well, AI can do that! Also /s, but maybe just 50%.
All this is pretty textbook setup for how this bubble finally implodes as companies fail to deliver on their AI investments and come under fire from shareholders for spending a ton with little return to show for it.
These things are clearly useful once you know where they excel and where they will likely complicate things for you. And even then, there's a lot of trial and error involved and that's due to the non-deterministic nature of these systems.
On the one hand it's impressive that I can spawn a task in Claude's app "what are my options for a flight from X to Y [+ a bunch of additional requirements]" while doing groceries, then receive a pretty good answer.
Isn't it magic? (if you forget about the necessity of adding "keep it short" all the time). Pretty much a personal assistant without the ability of performing actions on my behalf, like booking tickets - a bit too early for that.
Then there's coding. My Copilot has helped me dive into a gigantic pre-existing project in an unfamiliar programming language pretty fast and yet I have to correct and babysit it all the time by intuition. Did it save me time? Probably, but I'm not 100% sure!
The paradoxicality is in that there's probably no going back from AI where it already kind of works for us individually or at org levels, but most of us don't seem to be fully satisfied with it.
The article here pretty much confirms the paradox of AI: yes, orgs implement it, can't go back from it and yet can't reduce the headcount either.
My prediction at the moment is that AI is indeed a bubble but we will probably go through a series of micro-bursts instead of one gigantic burst. AI is here to stay almost like a drug that we will be willing to pay for without seeing clear quantifiable benefits.
And I did not speak out
Because I was not an artist
I love AI image generation, but many certainly do not enjoy the results. I can see some people skimping on paying artists.
First I thought translators would be hit hard by AI, but you probably still need them as well to be decently sure about correctness.
And it remains true that any creativity produced by AI is basically still just a function of the creativity of other people.
Now maybe Notion customers love all these AI features but it was super weird to see that stuff so prominently given my understanding of what the company was all about.
https://www.mckinsey.com/industries/technology-media-and-tel...
From what I know of the firm, it looks like clients have come to the right place if they want a consultant with great experience at hiking prices without cutting costs or boosting productivity.
"Only" 30%. Interesting framing.
I have to admit, the results they demonstrated — which we validated using our own data — were impressive.
The challenge, however, is that outcome-based contracts are hard for companies to manage, since they still need to plan and budget for potential costs upfront.
So even when you have measurable benefits - it's still not so easy either.
EDIT:
To clarify the issue — companies are used to budgeting for initiatives with fixed costs. But in an outcome-based contract, the cost is variable.
As a result, finance teams struggle to plan or allocate budgets because the final amount could range widely — for example, $200K, $2M, or even $20M — depending on the results achieved.
Additionally, you almost then need a partial FTE just to manage these contracts to ensure you don't overpay because the results are wrongly measured, etc.
None of these challenges are insurmountable, but it's also not easy for companies either.
- The moment AI is actually good enough to replace us, it will also be incredibly easy to create new software/apps/whatever. There could/would be a billion solo dev SAAS companies eating the lunch of every traditional tech org.
- People (Executives) seem to underestimate just how much of the work is iterating and refining a product over a long time. Getting an LLM good enough to complete a Jira task is missing the point.
- IMO LLM's are also completely draining the motivation of workers. A lot of software devs are intrinsically motivated by solving the problem. If your role is being watered down to "prompt the chat bot and baby sit what comes out", the motivation disappears. This also absolutely destroys any of the creativity/discovery that comes out of solving the task hands-on.
All of this said, using AI in your back end takes a huge amount of time from your users and employees. You have to vary multiple prompts, you have to make the output sane, touch it up, etc. The most useful part of AI for me has been using it to learn something new, or push through a task that I otherwise couldn't do. I was able to partially rewrite a logging window to reduce CPU use significantly. It took me over two weeks of back and forth with AI to figure out a workable solution and implement it into the software. I competent programmer probably could have done it better than I did in less than an hour. There's no business benefit to a help desk person being able to spend 2 weeks writing code that an engineer would be much better suited to handling. But maybe that engineer could write it in 10 minutes instead of an hour if they used AI to understand the software first.
The rank and file generally have a really good grasp on their subset of the domain -- they have expertise and experience, as well as local context. Small teams, their managers -- those are the ones who actually perform, and deliver value.
As you move up the hierarchy, access to information does not scale. People in the middle are generally mediocre performers, buried in process, ritual and politic. In addition to these burdens, the information systems do their best to obscure knowledge, with the usual excuses of Safe and Secure (tm) -- things are siloed, search does not work, archives are sunsetted, etc.
In some orgs tribalism also plays an outsized role, with teams acting competitive, which largely results in wasted resources and seven versions of the same failed attempt at New Shiny Thing.
Then as we look higher yet in the hierarchy, the so-called decision makers don't really do anything that cannot be described as "maximize profit" or "cut costs", all while fighting not to get pulled down by the Lord of the Flies shenanigans of their underlings. They are the most replaceable.
A successful "AI Transformation" would come in top-down, going after the most expensive headcount first. Only truly valuable contributors would remain at that level. Organizational knowledge bases would allow to search, analyze and reason about the institutional knowledge accrued in corporate archives over the years, enabling much more effective decision making. Meanwhile, the ICs would benefit from the AI boost, outsourcing some menial tasks to the machine, with the dual benefit of levelling up their roles, and feeding the machine more context about the lower-level work done across the org.
I worked in tech support at a big company long ago. Tech support, sales, and engineering used to have a week (for each employee) where we would leave our team and follow the other team around.
It provided incredible efficiency. I now knew what sales was talking about when they called me, they understood how I worked, engineering and I got along so well they used to invite me to their team when they had lunch catered.
Who didn't we need anymore? The middle managers between the groups who brokered what info each group could see, and how we communicated among groups. We solved problems before they started all on our own.
The middle managers won in the end, ending the cross training, too costly they said, but I think they realized that we just didn't need them / weren't engaging them anymore...
I think the general point here is true, but it's also brilliant framing from a company selling consulting services.