If you're a CEO running a company, you can go out and hire McKinsey and tell the board, "I'm a great CEO, that's why I brought in McKinsey. They're the BEST". McKinsey then proceeds to charge you a bunch of money and leaves you with some very slick powerpoints containing a bunch of advice, roadmaps, org charts, and action items.
If your company has a few good quarters after that you can say, "Look at what a great CEO I am, I brought in some best in class consultants and now our stock price is through the roof!". If on the other hand your company struggles, misses earnings targets, etc., you can say, "It's not my fault we're doing badly, I brought in a team of best in class consultants and even THEY couldn't help us turn things around yet - it must be the macro economy, industry headwinds, etc. Imagine how much worse we'd be doing if I didn't bring McKinsey in!"
They're also a normalized industrial espionage and wink-and-nudge collusion service. They call all that stuff "best practices".
In a way that cover's everyone's ass legally* of course. But that's what they're doing.
EDIT: * - by "legally" I mean allow information to be shared in a way that precludes legal repercussions but I do not necessarily mean adherence to the law.
Companies benchmark and consider industry best practices all the time. If they don’t they’re idiots.
Im not a huge fan of management consultants generally, but consultants and analysts can provide useful third party perspectives, in part because they talk to a lot of companies that you may have mess directly with access to.
Take McDonalds for an example, they are running around 30% margins and you see on social media customers complaining about prices.
Why don't those customers seek an alternative business?
The rest of them also follow pricing "best practices" so you have no where to run to.
There must be a reason McDonalds earns so much more. As far as I understand, McDonald’s has a significant real estate component to its business (which is technically selling its brand/real estate to franchisors, not selling food), bolstered by its very strong brand with loyal customers (that the others don’t).
https://www.mashed.com/178309/how-much-mcdonalds-franchise-o...
> The cost of running a business, especially a restaurant, can really eat into its profits. At the end of the day, McDonald's only keeps around 16 percent of the revenue its company-owned stores make, but it keeps 82 percent of the revenue franchisees pay out to it.
For instance, a company has to figure out how much to pay people. If McKinsey comes in and says "pay them less and we'll tell everyone else to pay them less" that just doesn't work. Why would any company that is 1 of N decide to pay them less when they can just pay them slightly more and attract the talent from all these other places. There is an incentive to cheat and without a legally binding agreement all collusion services will collapse. And this even ignores new entrants. You can have individual (illegal) agreements between companies like Google and Apple where they agree to not hire their employees, but those break down and have nothing to do with consultants (why would you want a middle man in an illegal activity?)
The reality is there is a market for most things. To hire an analyst it'll cost you X. McKinsey and other try to discover that and tell their clients. But they can't change the nature of the market as much as someone installing an HVAC unit can change your temperature but giving you a rigged thermostat.
Big players can collude. The CEOs can know each other by name and call around. Steve Jobs etc. did that.
Business consultants exist for a purpose that can be described as "how can we squeeze this enterprise or deal for as much as we can?" Sometimes that's squeezing your customer base through anti competitive practices. Sometimes thats offshoring (labor arbitrage). Sometimes it is M&A (realizing "efficiencies"). Sometimes it is squeezing your existing labor pool as hard as you can. Or a combination of the above.
https://www.simonandschuster.com/books/Goliath/Matt-Stoller/...
What McKinsey and the rest of the scumbags have taught CEOs is to be aggressive, cheating and ruthless at cutting costs no matter what. That's how we ended with almost all manufacturing and associated know-how getting shipped over to China (sometimes literally - they bought an entire steel manufacturing plant in Germany's Ruhrpott, tore it down and reassembled it in China [1]!), that's how we ended up with the clown show that is Boeing, that's how we ended up with "employee loyalty" not being a thing any more in either direction.
[1] https://www.wiwo.de/unternehmen/wiwo-history-tatort-dortmund...
Not if none of the competitors do, that was the point of the comment if they all think the competitors aren't competing on price then it becomes an informal price fixing scheme.
I think your second point is actually the reason why this will have a big impact on the industry, since McKinsey et al generally operate on a billable hours basis, and automation / AI will exert significant downward pressure.
Initially, I imagine they will staff smaller teams at a higher hourly rate (mostly by replacing some number of analysts and associates with automation/AI). Long term, I can imagine some firms will attempt to make up for the loss of revenues by expanding into SMBs ("Own a struggling Plumbing business? Get help from McKinseyAI!") or by dramatically reducing the size of their staff (as you noted, their product is mostly executive insurance, and for that you really only need the Partners).
The retention of accountability and access to otherwise-inaccessible information will continue to be among the hottest commodities.
If you’ve ever wondered what makes “shady data brokers” so powerful in cyberpunk stories, it’s not their collections of stolen credit card numbers - it’s insider knowledge of what your trillion-dollar competitor is doing.
https://www.goodreads.com/book/show/566213.The_Secrets_of_Co...
You can see how AI is already transforming consulting at the Big 3.
its unreliable insurance, since board may not buy this excuse.
But now you just need 2 associates instead of a team having you as part of their portfolio of clients. The fact that many the consultants are going "AI won't come for us" has me betting against the consultants lol.
Particularly for junior professionals, the expectations placed on them have suddenly increased tenfold, while the expectation to utilize generative AI for the most intellectually and creatively demanding tasks has diminished the mental reward. As a result, the most unsatisfying manual tasks are only ones left to be done - needing most of their attention.
I’m sure they are talking about the meaning of the deck, but they’ve already long offshored the creation of the decks to India. They make notes during the day, send the notes to India, receive decks the next morning.
Dunno about the rest, can’t see it.
AI agents can't do sql queries a lot of the time, but they sure can autofill the remainder of most excel sheets (and AI agents can't really autocomplete sql queries yet for some reason lol).
Yes, I use claude to code sometimes, but I use claude much more as an assistant to brainstorm tasks and mindmap than actually doing software engineering because vibe coding builds brittle projects. Whereas one-off decks are a perfect use case
Now replacing people that require a license or some other sort of credentials like lawyers, doctors or accountants... That is much harder.
Even sales - I hear of companies building bots that do outreach with human voices.
But the professional careers that were once considered "prestigious" (strategy consulting, investment banking, law, medicine) will be the most disrupted bc labor costs are the highest in those.
Imagine a lawyer doing discovery in minutes, rather than days or weeks. Imagine a doctor that can diagnose you from your smart phone.
The world is not ready for those changes.
For instance, in paediatrics one of the trickiest types of case to deal with are those relating to safeguarding. The tact required to navigate those situations, eg. correctly identify whether an unexplained bruise is a signal to flag a safeguarding issue or has a reasonable explanation, is not something that can be automated and would likely receive a lot of pushback.
It is worth noting that in 1930 Keynes predicted that the work week would shrink to 15 hours thanks to advances in automation.
I have great confidence in our abilities to create more work for each other.
That effectively happened. Then we invented staff meetings & quality circles.
These recent visits I saw junior doctors, one of them literally pulled Google to ask a question.
Any job that requires “memorising” things can be automated with generative AI.
The ability to iteratively observe and inquire to explore the state and the limits of the patient is NOT something automated docs will be able to do anytime soon. Diagnosis is a trivial task compared to learning who the patient is and how involved they actually will be in managing their own health.
This is because we have a lot of people who don't like the idea of having to go work while others sit around on their ass getting checks. I get it. It's the most natural thing in the world, but we've got to evolve past it and we've got to evolve past it right now. Not 50 years from now, not 10 years from now, right now.
A shitload of people are going to be out of jobs very soon. Sooner than anyone thinks, actually.
And there's really only two paths this goes. We institute Universal Basic Income and some kind of program or programs to help you learn a skill that doesn't lend itself to automation well... or... the wealthy continue to horde everything and half the population starves to death... which doesn't happen, because when people get hungry, they start to kill the wealthy. And frankly even with a swarm of murderdrones, good luck stopping 200,000,000 people from killing you.
Evolving towards Star Trek's Federation is the way to go, I just hope rich people aren't too damn stupid to do it.
$7 a day would be a step down for first-world (and second-world) economies. I think people would be very unhappy. Very very very unhappy.
I asked Gemini today how to use my (paid) credits to get more RAM. The conversation was roughly like this: "You have to click on this option". "That option doesn't exist". "You are right, that option doesn't exist. You should consider paying for an upgrade".
My theory about chatbots is that, in general, they are only replacing humans in tasks in which no one really cares whether the problem is solved at the end: if I'm unhappy about my internet connection and the company has a monopoly, then all my call does is ensuring that the company doesn't spend more than legally necessary on a problem they never intended to fix.
All AI is doing is making it easier for you and I to be ignored at scale. So while I don't doubt that some doctors will diagnose people from my phone, they will be shit solutions to the wrong problem.
I think this is exactly right, and possibly a good thing relative to the alternative. If AI were able to automate 20% of office-type jobs, you'd have a lot of people out of work, but likely not enough for anyone to do anything about it at a societal level. Those people would just be given some platitude in the same way out of work coal miners or assembly line workers were snidely told to "learn to code".
If every or nearly every office-type job gets automated, it would force us to address the issue in a meaningful way.
Think we’re still a while away on many of the other knowledge work ones. Often they just don’t have a good source of training data that has similar depth to GitHub/stack overflow etc.
eg law - yes you have law books. But that doesn’t have the same 1:1 relationship as code does. A lawyers output is not law books
I've felt this personally as a dev. We are getting squeezed to move and deliver faster because we have AI.
The reality is a lot simpler (and darker) when you realize who the buyers of this technology are, and what they want from it. And what they want is always the same...more, and faster.
That said, the article is on point about how junior and associate roles are now on the line.
You can be dismissive all you want on here, but at the end of the day, FAANG, MBB, YC, and BBs - not MBAs - are the LDPs of the 21st century, and if you don't like leadership at companies today, it's only going to get worse, because the people who survive this will be even more mercenary.
War is extremely lucrative because there will be "reconstruction" contracts once the dust settles. For McKinsey, these are equivalent to "restructuring" businesses after economic recessions.
Trillions were lost in the March 2000 dotcom collapse and the September 2008 crash. McKinsey, Bain etc emerged unscathed. There's plenty of business to go around even if the economy is in shambles.
They'll force it down people's throats and then claim that they have tons of users. They don't even both trying to sell the idea that their product is good for humanity. They're doing their level best to kill art, to kill journalism, to kill medicine, to end the job market as a source of opportunity...
Let's be clear: the AI playbook is the same playbook SV always has. Get a lot of people to come in and look around, then lock them inside and start to exploit them for your gain. It's just the scope of the control they seek now is orders of magnitude more than they ever have before. They are trying to convince people to stop thinking for themselves, to stop doing art, to stop valuing art, and in general, for the engineers and employees they've already locked into their power structure, never to question that this is the way of the future
These people should be a lot more worried lol.
Then, it'll be checkmate.
> 5,00,000+ subscribers [sic]
?
Instead of the ones(10^0)-thousands(10^3)-millions(10^6)-billions(10^9)-... system followed in most other parts of the world, the Indian numbering system uses ones(10^0)-thousands(10^3)-lakhs(10^5)-crores(10^7)-...
So, for example, half a million subscribers (500,000) would translate to 5 lakh subscribers (5,00,000).
1,23,45,67,890.
Splendid.