Corporate Ozempic
Audio Recording by George Hahn
If you want to understand how AI is reshaping business, picture it as the other massive innovation of our time: GLP-1 drugs. Both shed weight by suppressing cravings; both exacerbate existing inequities (aka the rich get richer) before generating wider prosperity; and both are having a greater impact than projected as early adopters are hesitant to admit they’re using.
Secret Sauce
Nobody I know is on Ozempic. Yet, nearly everyone I know is on Ozempic. Either that, or gluten-free diets are suddenly delivering exponential results previously unheard of. People are hesitant to acknowledge they need a drug to lose those last 15 pounds — which doesn’t fit with our narrative that success is correlated to self-control. One of the most interesting, and discouraging, features of GLP-1 use is that the region with the greatest per-capita prescriptions is also (wait for it) the thinnest. This makes no sense … but it does. The Upper East Side of Manhattan is replete with people who can spend $1,000 a month to go from slim to skinny. In sum, GLP-1 drugs are not (yet) getting to the communities who really need them. I believe this will eventually happen, however, because the public health and economic benefits are just that staggering.
Corporate Ozempic
Similarly, my thesis is that firms (notably tech companies) have also discovered a weight loss drug and are also being coy about it. Recent financial news features two stories: layoffs and record profits. These are related. There’s no mystery to the surface narrative. A company lays off 5%, 10%, or even 25% of its workforce and, 6 to 12 months later, after severance pay and expenses are flushed through the P/L, its operating margin hits new heights. The ultimate peanut-butter-and-chocolate shareholder confection is Meta, which produced a singular Hall of Fame quarter in Q4.
All told, tech companies fired 165,000 people in 2022 and 260,000 in 2023, and they’re on pace for 270,000 in 2024. The media narrative is that these firms are shedding the excess weight they put on during the pandemic. The collision of remote work, stimulus, and social isolation triggered a historic revenue acceleration and a reflexive hiring binge in tech. And we’re now seeing the course correction. This phenomena is undeniable: Between 2019 and 2022, Amazon doubled its headcount — it tripled its corporate headcount in five years. Amazon’s pandemic hiring may be the greatest non-war scaling up of a private employee base in history. But every major tech player bulked up. Even prudent Apple grew its ranks 20%.
Overdone pandemic hiring was a reasonable explanation for the tech layoffs in late 2022. But 18 months later, there’s been sufficient time to reduce headcount. There’s no incentive for a company to slow-roll this process — a drip feed of layoffs drains company morale. There is something else going on.
Business results also don’t explain the deeper cuts. These companies are killing it. Meta’s revenue was up 16% in 2023; Alphabet’s, 9%. Microsoft’s most recent quarter registered 72% greater revenue than the like quarter in 2020, which (until then) was its highest-grossing quarter to date. Accordingly, Big Tech’s stock prices are at all-time highs, and the magnificent seven were responsible for 70% of the S&P’s Macho Libre year (up 24%). Based on the numbers, the pandemic hiring binge wasn’t a mistake, but the correct response to a step-change in business. They needed these people … until they didn’t. The layoffs are no longer a signal of economic conditions, but innovation.
It’s not just tech. While the broader economy is enjoying steady job growth and low unemployment (the longest period of sub-4% unemployment in 50 years), many firms have let large numbers of employees go — UPS, CVS, and Hasbro are among the companies that have announced layoffs of 1,000+ people in the past six months.
What’s really going on? I believe AI is playing a larger role in layoffs than CEOs are willing to admit. There have been hints: IBM’s chief said the company plans to pause hiring for positions that could be replaced by AI, and UPS acknowledged that AI factored into its recent layoffs. But as a general rule, expect a CEO to be reluctant to state on an earnings call that the fastest-growing technology in history is already giving her “the ability to lay off people without any impact on the top-line.”
That creates a messaging maze no Investor Relations or Corporate Comms group has faced before. Shortly after UPS’s CEO mentioned AI in the context of layoffs, a spokesperson clarified that “AI is not replacing workers.” A similar post-earnings call two-step occurred at IBM. It’s rare that growth and the size of your employee base are inversely correlated. And it doesn’t make for a good all-hands narrative: “We’ve had a record quarter, and we’re going to need fewer of you.”
It’s the denials that first raised my antennae: “We’re not restructuring because AI is taking away roles,” Alphabet’s Chief Business Officer Philipp Schindler told analysts on the company’s most recent earnings call. That’s the “I gave up gluten” of tech.
Craving Suppression
Ozempic and other GLP-1 drugs work by suppressing the brain chemistry that rewards us for eating. It doesn’t make us thinner directly, but reduces our cravings for food. Because the same neurochemicals drive other cravings, GLP-1 drugs hold tremendous promise for addressing many of the societal ills that flow from our superabundance. AI can have a similar effect on corporate cravings. And if consumers are willing to pay $1,000 a month to lose weight without cravings, what would a corporation pay to achieve the previously unthinkable: reducing costs while growing revenue?
As I write this, Nvidia is valued at nearly $2 trillion, up 16% after reporting its earnings on February 21. Put another way, in the 60 minutes following its earnings release, Nvidia added the value of Ford, Ferrari, and GM to its market cap. Maybe even more impressive, the company has added the value of Tesla in the past six weeks. Nvidia GPUs are legal performance enhancing drugs for corporations. Add “1980s East German Olympian” to one of the monikers assigned to Meta, as it’s tied with Microsoft for largest corporate buyer of these PEDs (Nvidia GPUs).
We Are Everything
There is a piece of every star in all of us — we are the universe, and the fundamental imperative of the universe is to expand. The rookie move I keep making as an operator is overhiring. And most CEOs I’ve known make the same mistake. It’s the siren call of the cosmos … just grow. And new employees provide the illusion that you are expanding. You can see them in the halls of HQ. They build out individual fiefdoms of senior managers and produce more products, features, sales calls, and revenue. More of everything feels like … growth.
The hard part is to build a business whose output is greater than the inputs. But as long as the output is visible and increasing, you feel you are playing your role making the universe a better place. Now, as a board director, I’m usually the person pushing back on ambitious (i.e., bat-shit crazy) hiring plans, and I’m the first to suggest cutting costs. That may be why many/most of my startups survived, but I was also never able to build a $1 billion company. To be clear, there’s definitely opportunities for outsized value creation (sometimes) in the pursuit of crazy growth. It produced a new generation of businesses (AMZN, NFLX, etc.). But, as I tell CEOs: Assume you are not Amazon.
AI by Another Name
The media portrays the impact of AI as a one-for-one proposition — Mary the copywriter losing her job to ChatGPT. But that’s not how AI is trimming corporate America. Instead, it’s picking off individual tasks and augmenting teams with more capabilities. Goldman Sachs estimates that AI could perform about one-fourth of the work done by humans today, but that two-thirds of jobs are exposed to some degree of AI automation. That automation is taking many different forms. UPS is using AI to determine pricing for contract proposals. Allstate’s AI is developing internal training programs, doing in a day what once took three weeks.
Jobs are being lost, but augmentation will be the broader story. Rather than copywriter Mary losing her job, Mary’s firm will train her on an AI tool that generates first drafts, takes approved product copy and converts it for catalog, web, and social use, and streamlines other tasks. Accordingly, Mary’s manager will expect her to generate three times the copy in the same time.
Managers can take on new initiatives and domains without the headache of hiring more humans. It’s growth without calories. I’ve founded two strategy firms (Prophet and L2) and loved everything about them, except for the clients and the employees — every additional hire creates complexity and thus increases risk. The AI revolution will inspire a golden age of startups with lower infant mortality, as there will be fewer people (i.e., less risk) required to get to sustainability.
Coy
CEOs are being coy about this, at least in public, because there’s a sense of fear surrounding the brave new world of AI. The illusionist’s trick in the Valley right now is getting the media to look over there (trimming fat) while they’re stuffing the rabbit into the hat here (replacing it with AI). In the next several quarters, however, I believe CEOs will come out in earnings calls and put it bluntly: “We’re going to be a smaller company that does more business thanks to AI.” Pundits will clutch their pearls for a hot minute until the stock explodes, and the secret hiding in plain sight will be visible to everyone. It’s corporate Ozempic. It’s not about less bread, but less craving for bread. Read: hiring people.
This will, correctly, raise concerns about a dystopia where nobody can find work. But AI will ultimately create jobs, as there will be new windows of attack against corporate titans. AI is currently a hormone therapy for mature companies looking to feel young again. Soon enough, though, a new generation of firms raised on the mother’s milk of AI will create an army of supersoldiers ready to attack bigger armies who are still fighting on horseback. I’m here for it.
Life is so rich,
P.S. Did you know you can listen to No Mercy / No Malice read by George Hahn? Subscribe here.
P.P.S. I’m teaching an Algebra of Wealth course for Section on May 13. Sign up now to get a simple formula for building long-term wealth, directly from me.
Yes, I think you are right. AI is indeed changing the whole landscape.
Hi Scott – thanks for this – it did make me wonder whether AI shaped futures have to be hysterical, high pressure and militarized. If so then AI=big yawn.
Change is the only thing that is constant.
Adapt or be left behind
employees must open their eyes
Time to launch that passion project using AN while drawing unemployment
Bezos, Leon Black, Jamie Dimon, and the Walton family have now sold a combined $11 billion in company stock this month—some for the first time ever
Technology has been replacing/enhancing/making industry more productive since the industrial revolution. This is another step along the way, rather than some kind of AI revolution.
Things will change, but based on what I see today, AI still needs a lot of supervision from people that know what’s going on. Whether that’s from doctors, programmers, writers, etc …
What I’m concerned about is that this has a lot of potential to make things worse in some areas where senior management blindly follows through some means of cutting costs.
I’m seeing this now in various customer service groups, where they are cutting costs by replacing people with machines. Most of these chatbots that companies are deploying now are mostly useless. I recently replaced one of my ISP/telecom companies, because their service was abysmal – talk to the chatbot, wait on hold for indefinite amount of time. In contrast, the competitor would have a person assist me with minimal/no wait times. If the service person couldn’t help me, he would know to pass me on to someone who could.
The false promise/wrong use of AI has the potential to make this all worse. I feel bad for the customer service people left behind dealing with frustrated customers.
The evidence regarding Ozempic is that, like the majority of other drugs, it stops working after some time and with Ozempic, people regain the weight they lost. Ozempic is heavily promoting its product now because they know, that very soon, they are going to have to put some extra warnings on their packaging.
I doubt corporations will re fill the jobs they cut thanks to their effective use of AI.
Lots of mistakes are being made with ChatGPT and other AIs. Staff are putting in personal information belonging to clients and sensitive company information – both are of course big no nos. Companies have not taken the time to installed guard rails, no AI policy means staff are just doing what they think is a good idea.
Google loses $70 BILLION in market value after Gemini AI exposed as woke
Google parent company Alphabet lost over $70 billion in market value, with stock prices falling as much as 4.4 percent on Monday following the pause on Google Gemini’s AI image creator and posts surfaced of Gemini’s responses in the chat feature.
According to Bloomberg, Melius Research analyst Ben Reitzes warned in a research note that problems that have arisen with Gemini may fuel the perception that Google is “an unreliable source for AI,” as many companies try to launch their own AI platforms.
What do you think was the rational behind overhiring in a pandemic?
While some folks already tell “AI is replacing almost any job”, I don’t read that exactly from your article.
Here is what I hear:
– Hiring like crazy was the default way “new managers” in hyper-growth companies. Often 50% of time doing just hiring.
– These managers over the last decade often grew into Directors, VPs etc. keeping their ego and narrative around “more people => more success”.
Thanks for high interest rates, LLMs and soon “AI Agents” (Meta is doing interesting Swarm Intelligence ideas into AI) a lot of repetitive work, things that can take days are now done within minutes / hours.
So now managers (who are often also let go) realize that number of hired people != success, it’s profitability again.
I love your use of Corporate Ozempic, as “Getting fat” is a normal corporate wording, so AI & LLMs are really the new corporate Ozempic.
Here comes the thing (which I think I’m reading between the lines):
A) Just laying off people doesn’t make you long-term success. Even big tech companies haven’t proven any real jump in their revenue (except Nvidia). They are innovative, yet getting 20-30% slimmer being still that “old corporate” doesn’t make you faster.
B) The age of “hypergrowth = hire like crazy” is gone. With LLMs and AI Tooling, you can do 80% of the past work with likely 20% of the headcount of the past.
Definitely crazy decade ahead. 2035 will look way different (or AI winter kicks in :D)
This is a very refreshing perspective on AI. You always have a refreshing perspective on whatever topic you tackle. Deep insight as one who is in architecture and construction and the possible impact AI use will have on my industry. Lots to think about. Many thanks
220 / 5,000
Translation results
Translation result
Perhaps, and this is my opinion, someone took advantage of the pandemic to balance working from home a little too much with doing their things from home and the companies noticed this and now, they also have AI at their disposal
Cool
In late 2022, Musk paved the way gutting labor force by 80%. While Twitter/X has certainly turned into a cesspool due to governance issues, technically it still stands. Having seen this Meta, Google, Microsoft and others confidently started cutting back on headcount. IMO, adapting AI into this mix, high-paying employment opportunities in tech can further diminish.
“AI will ultimately create jobs, as there will be new windows of attack against corporate titans”
To me this is the most interesting part of the article. Would love to hear you expand more as to what those “windows of attack” could look like in a future essay.
Astute observation, Scott. Are you suggesting that all young people ahould expect difficult times until 2030?
I have worked in this industry for several years. The concept of a “10x developer” is nothing new and the deep cuts at X have demonstrated that the core business can limp along even with a fraction of the former staff. My wife is an agency copywriter and hasn’t found much use for ChatGPT aside from generating a basic list of ideas.
There are 2 main issues with AI tools in programming: 1) they only work for the most common problems & tools and 2) they tend to degrade quality and lead to more long-term issues. Realistically, the current spectrum of skills among programmers is already much larger than any impact AI could have on the profession.
The reason we haven’t heard many stories about technical employees becoming dramatically more productive is because, for the most part, they don’t exist. It’s just a convenient buzzword for earnings calls while discussing layoffs designed to boost the stock price. AI does not deserve credit here.
Weight loss diets are only effective when combined WITH regular muscle gain (hint: lift weights!). If corporations see automation via AI as the new pill, they need to better figure how to utilize human skills and brain power to run the machines instead of casting talent aside. — Corp Ozempic over time will lead to muscle atrophy…..and sagging butts. Just sayin’
Great article. Scott, you have a brilliant ability to declutter topics and offer truly useful insights. I love the parallel analysis of AI and ‘secret’ weight-loss drugs. Much food for thought. 😊
Nice article Scott. I am surprised at the fear, uncertainty and doubt (FUD) that pervades throughout the comments. I think there is an inverse relationship between the FUD and the likely outcome. The massive productivity to come will likely raise almost everyone’s standards of living. The future is bright and everyone thinks it’s an oncoming train. Embrace change- it’s uplifting and enabling.
Perhaps not in your stratosphere, Scott, but for us on the ground we’ve been offered this bait before. Computer use revolutionized how work was organized, sold, and delivered. All were assured that a place could be found at that table for everyone–that is, if you could get yourself retrained yourself. From that point on to now, the seducing power of quick, easy, cheap has been sold to us with all kinds of promises. AI is just the next hollow promise we’re expected to swallow with gratitude. Even your article here pitches the bait: Mary will learn something new. Actually, no she won’t. Takes money and time to retrain your in-house people. Will Meta or Alphabet or MSFT spend that nickel? Think again.
You attempt to parallel Ozempic and Artificial Intel which, have to say, is a catchy approach. But then assure us that this AI stuff will launch a new era of jobs. It’ll raise all our boats and fix what ails us, so that a beautiful land of augmentation will take shape for all. Except it won’t. (Nor will its more interesting cousin, quantum computing.)
Outside of its circus tent, we view AI with a necessary pragmatism. It’s yet one more iteration of a business model that, to us in the trenches, means less of everything for everyone except the few who can afford it. In other words, sure, it does neat tricks, but can we afford what AI will cost us laborers socially, environmentally, and existentially?
Everyone is not better off. If they were it wouldn’t be progress, it would be a miracle.
As a tech worker, I can confirm the Ozempic theory, and as with the weight loss drug, it doesn’t create a healthy body, it just creates a non-sustainable effect. As these companies are basically shitting out jobs, like a body, they will also eventually endure “Ozempic ass”. In a corporate form, this results in not just “Mary producing 3 times as much”, but “Mary actually being overworked three times as much” because AI doesn’t actually do as much as CEOs think. This will ultimately be bad, because AFAIK, businesses still involve people.
Like weight loss in the slim, it isn’t because being slim is bad, it’s that the Ozempic user is increasingly vain, and is enduring the awful side effects of the body because of self loathing. In the corporate world this is “CEOs loathing their staff” – anger at the lack of desire of a full return to office, where they can stoke their egos at sycophantic employee simpering, greedily lapping at the adoration of their workers ( which is the sugary pies and fatty foods they crave ), C-level works out their resentment of not being fed what they want, and instead turns it on workers, ostensibly “shitting out jobs”, in revenge, instead of adopting exercise and good diet.
Do you know Debbie Downer? Really, there are other more optimistic scenarios. This is not the first technological upheaval.
I’m old enough to remember that “knowledge worker” was the way forward and that the “knowledge economy” was the promise of new jobs to replace manufacturing . Now in less than half my working life that promise is dead – all that knowledge is being sucked up by AI and regurgitated as something “new” but without the jobs for older workers who created the damn knowledge in the first place.
It’s not dead. Reducing lower level jobs has always produced more high paying jobs. Be flexible and learn new tricks and you will be fine- and probably better off. Society has benefited from every economic revolution. This is likely to be the same.
These are not just going to be lower paying jobs. Programmers, lawyers, accountants are going to be replaced by AI. The jobs that will not be replaced are those that require a body – trades, retail, nurses.
Is it really the AI? Honestly, I think this is corporate executive “virtue signaling” where CEOs believe these headcount moves will improve the company “brand” with institutional investors. I see precious little evidence that AI is a major contributor to productivity, and I’m pretty skeptical that it will be, other than in small ways without some major breakthroughs. Major breakthroughs are notoriously resistant to money being thrown at them. Real discoveries are the result of curiosity and strokes of inspiration that can’t be bought.
Besides the signaling the other thing I think is happening with these layoffs is vast corporate power born of decades of lax antitrust enforcement turning towards workers. Because large corporations know that don’t really have to compete they know they can tighten their belts and it won’t really affect their bottom line.
Speaking as someone in the controls industry, I have noticed that end user companies have in the last 1.5 years become more indecisive. It also looks like the roles in an organization that champion projects are no longer being filled so I think it’s that when everyone is doing less and making less and championing less projects, the status quo, which is to say, the big publicly traded incumbents win. Hence the increase in profits. That does not mean they are implementing AI in any meaningful way because they are just doing less. So corporate ozempic is a good term for it, but don’t be surprised if general malaise and apathy is the main driver, not AI taking jobs. But when you think someone else is making something that can replace a need to hire, it’s easier to just sit and wait. But when everyone sits and waits, no meaningful progress is made on any front.
Did you use chat gpt to edit? The word “phenomena” is plural form of “phenomenon”, which is the word I think you meant. I sure hope AI is learning from proper grammar! I also wish you’d addressed the side effects of Ozempic/AI more fully. What good is weight loss if you ruin your health in other ways, or increased profits resulting from firing humans (that cause increased unemployment) if you have no one left who can afford your product or service?
In the initial phase, it will be relatively more straightforward to replace employees with AI technology. However, in the subsequent phase, it will become easier to replace entire companies due to the increase in innovation. As you have mentioned, new initiatives will be less risky, and the exciting part is that there is now a wealth of intellectual capital available to accomplish that.
“But AI will ultimately create jobs, as there will be new windows of attack against corporate titans.”
That’s great in the long-run, but people have needs and responsibilities now. If your laid off in the AI-froth will your lenders be willing to give you a break on your mortgage payments? If we want to have a society of builders we need social support (universal healthcare, UBI, etc.) that give everyone a foundation of base security.
Scott, I first discovered as a superstar teacher video on YouTube. Following you offers great insight, but often there appears to be two unrelated things going on. One, you appear saddened by the very information you are delivering. Regarding ozempic, you do discuss the loss of jobs. Other times it can be similarly dispiriting. Second, you seem both embarrassed by the amount of success you’ve had, and at the same time embarrassed that you never reached billions in success. This makes you heroic in that, unlike most publicly successful people who claim even the credit they didn’t earn, you are refreshingly humble and self deprecating. You are also visibly uncomfortable relaying uncomfortable things. But here’s my real point. All if this seems to make you actually sad. You visibly struggle to smile, even when saying funny things. If any of this is accurate perhaps you can orient your delivery as an early warning system to your students and followers. You should capture a bigger audience than only the people who feel you will make them CEOs. I think you are pained by outcomes that affect workers, but you don’t lean into it. In the Izempic article you are talking to only future CEOs. Mary works for them. Why aren’t you talking to the Marys of the world as well. Yes, Young-CEO may chaff a bit, but maybe you would be doing something that aligns as much with your heart as it does with your head.
**discivered you**
Correction: “discovered YOU”
Nice and accurate. Great insight. Post more!
For an old guy, retired and 91, your weekly commentary is very to the point and interesting to me. I have not thought that much about the effects of AI, but you piece got me thinking. Thanks, Fred
Reducing social security eligibility age by 2 years produced weeks of rioting in France, what is going to happen when millions globally cannot find work… Rage against the machine
Yet, in America the unemployment rates is low and has sustained this low for over 50 years. Progress creates more jobs than it destroys. Always has and likely always will. Yes, tragically there are those that fall behind. If everyone was better off it would be a miracle- not progress.
“Accordingly, Mary’s manager will expect her to generate three times the copy in the same time. ”
To me, this is typical corporate bean counting mentality – “how can I squeeze more juice out of the human orange” Literally everything that I’ve read learned about AI is focused on cost-cutting and efficiency. Where’s the focus on improving our daily non-work lives? I’d gladly pay for an AI assistant that could handle a task like this:
Alexa (whatever), buy me two tickets to the upcoming Taylor Swift concert at the Hollywood Bowl. I want to be in the upper seating area and am willing to pay $1,000 per ticket. Use my credit card ending in XXXX to make the purchase”
Consumers aren’t buying Ozempic. The 10% are.
AI isn’t going to benefit most workers. It will benefit shareholders.
AI and robotics are rapidly invading a host of jobs that few would have believed endangered just a few years ago. Weeding fields, picking fruit and berries (only the ripe ones), radiology. It’s about to completely upend the massive gaming industry. Just a few examples. I saw my first computer when I was almost thirty. My AI course in graduate school was laughable. The field was a joke to everyone not part of it. It was kind of a joke for another generation, but today, I cannot even keep up with the fields it is revolutionizing. My overview understanding is obsolete every year now, and the technology hasn’t yet taken the big step of linking LLM’s and a host of specialized AI capabilities into the kind of “society of mind” that Minsky foretold. My point is, not just many fields, but almost every kind of work is on the block, not 100 years from now, but by the time babies born today start working. It seems to me that people who fear socialism are in the backwash of history—some kind of socialism is inevitable because very soon there will be almost no jobs that machines cannot do better. Even “creative” work is falling fast. We’re a decade away from being able to fully and convincingly generated virtual versions of any actor, at any age. We’re even closer to being able to generate the writing. It’s time to start planning for this in a way that isn’t piecemeal, but anticipates the social consequences a revolution larger than the discovery of agriculture.