Despite its very obvious trust and safety issues, seemingly every major company is trying to cram generative AI into their roadmap. Snap, for whatever reason, has decided that the best allocation of their resources was to add a ChatGPT-powered AI bot to Snapchat to “recommend birthday gifts for your BFF” or “plan a hiking trip for a long weekend,” features that don’t really make sense in a chat app. Or, for that matter, at all.
Meta, a company that predominantly makes money through advertising, has created an entirely new product team under a former Apple VP built to “[quickly] integrate the latest [generative AI] technologies into its products,” as well as open-sourcing their own large language model “to help researchers advance their work in this subfield of AI,” and crucially appear to Wall Street as a company that was doing something with generative AI.
Salesforce has also announced their own generative AI product, called “Einstein,” that can allegedly generate leads and close deals according to CEO Marc Benioff, though I can find absolutely no information about when or how it will do so. What I can find is Marc Benioff discussing or launching a product called Einstein AI in 2021, 2019 (“you speak to Einstein, Einstein speaks to you”), 2018, and 2017 (when Einstein had some sort of partnership with IBM’s Watson to “boost the range of predictive analytics it can provide clients”).
Benioff, I note, is no stranger to jumping on a bandwagon. In 2016 Salesforce was the “first comprehensive AI platform with CRM” — a phrase that may seem superficially impressive, but actually doesn’t mean anything when you think about it. In 2021, blockchain briefly became “the future of apps” as part of the Salesforce Nocode Platform, and was as easy to install as Einstein AI, which is confusing, because it seems that in May 2019, Benioff promised that “every salesforce app [could] have full Blockchain…All in clicks not code.”
The reason that none of these products seem to ever have a firm connection to any real use case or product is that these announcements are not for customers. Benioff’s 2019 announcement of “full blockchain” coincided with a brief bull run in cryptocurrency, and his 2021 announcement coincided with Bitcoin’s brief flirtation with a $60,000 valuation. And Benioff’s new push into the nebulous world of adding AI to a CRM for the fourth or fifth time is likely to placate a push by activist investor Elliott Management and to try and stabilize a stock that has dropped from $210 to $167 in the last year.
That’s because Benioff, like many valley executives, is not concerned with the calamitously low morale at his company, or innovating in the sales space, but sending signals to Wall Street that Salesforce will grow eternally. Despite his “ohana” attitude and everybody generally liking him across the valley, Marc Benioff is running the same playbook as Elon Musk - making a vague, exciting statement that makes the company seem in control of every single nascent market opportunity, even if they never actually seem to meaningfully enter into it or do anything related to it.
Salesforce can lay off 8000 people while paying Matthew McConaughey $10 million a year, all while doing billions of dollars of stock buybacks, with their net profit margin dropping 60.7% year-over-year. And yet Benioff is applauded and Salesforce’s shares have jumped over 10% because revenue grew 14% year-over-year.
Salesforce, like many companies, plays interesting word games with their earnings, such as claiming they have a “GAAP operating margin of 4.3%.” That’s because Salesforce’s net income for the first quarter of 2023 was negative $98 million on $8.384 billion in revenue. They lost $28 million in Q4 2022, down from a profit of $210 million in Q3 2022, and, confusingly, $28 million of profit in Q2 2022.
What I’m saying is that Salesforce’s leaders (particularly Marc Benioff) don’t even need to run their company well - they just need to both keep revenue numbers on an upward trajectory and keep the stock market believing that they’re continually entering new markets, even if they never actually seem to do so. Salesforce doesn’t need to make a better CRM, or actually integrate AI in a useful way, because Marc Benioff is able to march onto CNBC and tell Jim Cramer that he wrote a book and businesses “must value stakeholders as much as shareholders.”
As I’ve hinted at, this is the Elon Musk playbook - an endless font of promises backed by convincing-enough verbiage and marketing collateral to make something look real. Musk has made so many promises that someone had to make a website to keep track of them, creating a constant air of “innovation” around Tesla as a company and a stock, all without ever having to actually do anything. Nobody in Wall Street seems to care that Musk’s promises - and vehicles - are regularly broken, because has realized that the markets don’t really care if his company is good, or whether it makes a lot of money, but whether it has the appearance of “growth” and “market share.” Despite being distinctly unlikeable, Musk is a savvy manipulator, and he’s well aware of what Tesla needs to resemble for it to continue to have its outsized valuation.
Where’s Your Ed At is a free newsletter, but if you like my work and want to kick me a few dollars, you can do so here. I really appreciate your support.
Except Musk took it a step too far at his first-ever Investor Day, where he rambled about “a clear path to a sustainable-energy Earth,” “re-powering the grid with renewable fuels,” and how Tesla would cut costs in the future. The metric-laden presentation was part of Musk’s "Master Plan 3,” a vague roadmap for Tesla that follows 2016’s “Master Plan, Part Deux,” a plan that has yet to be fully completed because, much like the Master Plan 3, it is incredibly vague and makes massive promises like “you will be able to summon your Tesla from pretty much anywhere” and have it “make money for you when you aren’t using it.”
Yet Musk’s presentation lacked any meaningful updates on anything that the company would actually be launching. No new cars, no new software, no cool new doohickeys. Despite being entirely about the future, Musk gave us very little insight into how we’d get there. After years of tolerating the fact that Teslas never quite seemed to live up to their company-driven hype, investors are finally asking a very simple question: what is it you intend to do to keep this company growing? Note that I am not saying “growing sustainably” or “making profit.” Tesla is profitable, though how profitable it will stay is somewhat dependent on the future of EV credits. Musk has kept the company going on delivering just enough to placate investors, but his disastrous acquisition of Twitter (and the preceding sale of billions of dollars of Tesla stock) has led some to worry that perhaps the man may not match up with the myth of Musk.
It’s ironic because Musk is arguably the biggest beneficiary of the market’s preference for new, shiny objects. His vast fortune has been grown through living on the bleeding edge of paper innovation by promising Tesla would produce fleets of robotaxis by either 2020 or 2024, or tunnels to reduce traffic never actually materialized, or the ability to summon your car across the country by the year 2018 (the year is 2023). These promises never really had to be fully realized, because Musk was (is?) the king of delivering exactly enough to appear innovative without ever having to actually innovate. When Musk hypes a plausible-enough sounding concept, he connects it to a current product that exists (such as Tesla’s current summon feature, which has its own sets of issues) and then connects it to a possible yet stupid concept like hitting a button and your car somehow driving 3000 miles.
I believe that Musk has realized that the markets don’t care about innovation - they care about disruption. The two terms may seem synonymous, but they’re fundamentally different. One refers to actual technological advancement, whereas the other refers to the appearance of innovation, whether it exists or doesn’t.
The market doesn’t want companies to change the world or take huge risks - they want to see you grow eternally rather than create a new way to do something. And that’s because you can’t really measure innovation as a neat figure as with revenue or customer acquisition. It’s not that innovation is conceptually ephemeral or vague, but rather that it’s borne from long-term effort and is inherently risky. When someone tries to build something new, there’s no guarantee it’ll actually work, or that it’ll find a market. Innovation requires experimentation and a willingness to fail, along with an understanding of the timelines something must be given to test its efficacy, which are usually longer than the three or four quarters public tech companies will give a concept.
Venture capital may take “risks,” but it takes it with proven founders (read: white guys they like such as Adam Neumann making a slightly different way of renting an apartment), or with things like Web3 where it can make a quick buck, or with whatever “big idea is getting all the money” like generative AI or the metaverse.
And public companies follow suit - once Meta announced their “metaverse,” Microsoft immediately rushed to discuss their metaverse strategy, with Satya Nadella saying that he “could not overstate how much of a breakthrough [the metaverse] is” despite there being no fundamental change other than the logo outside of 1 Hacker Way. Microsoft ostensibly spent nearly $70bn to acquire Activision Blizzard to develop its metaverse play (or, at the very least, tried to market the deal to acquire the company as a metaverse play), no matter how unproven the metaverse was as a marketable concept, or how likely regulators would step in and halt the deal over anti-competition concerns.
Ironically, Zuckerberg is sort of trying to innovate. Reality Labs lost $13.7 billion in 2022, and yet he continues undaunted. The problem is that he’s also miring his research and development in endless bureaucracy, leading to legendary designer and Oculus CTO John Carmack to leave the company. Zuckerberg may be interested in innovation, but he managed to ostracize someone that was formative to creating some of the most important virtual environments in history. One might be forgiven for thinking that Zuckerberg is really just burning cash as a means of appearing innovative without trying, especially considering how much of Meta’s “announcement” video was outright bullshit.
As Preston Gralla of Computerworld wrote, two years later Microsoft has almost entirely abandoned the concept of how “human presence is the ultimate connection” for AI-powered search, which Nadella says he has “not seen something like since 2007.”
Roblox - an online game for children launched in 2006 that went public in 2021 - rode the hype of the metaverse to an all-time high of over $130 a share based entirely on another company (Meta) saying that online worlds were important. Did it matter that Roblox has never lost less than $70 million a quarter, or that it’s a fundamentally unsafe and exploitative environment for kids? No. Only when the metaverse hype died down did things return to their pre-peak normality, with the stock dropping to $36 a share as of writing this sentence.
Finally, the Match Group declared in November 2021 they’d be "creating a dating metaverse,” only to kill it less than a year later after a poor showing in their Q2 2022 earnings. To bring this “dating metaverse” to life (or, at least, try to), it acquired South Korean app maker Hyperconnect for a cool $1.73bn. Money well spent, I guess.
The markets are only interested in symbolic innovation - the flashy term or idea that will suggest you will be a “growth company” (even if you don’t make any money) or part of the next big thing that consumers will spend money on. Venture capital may claim to love “innovators,” but what they really want is the quickest possible multiplier of their initial investment, which is why a lot of people are equating the hype around generative AI to that of the cryptocurrency boom of 2021 and early 2022. Venture capital may want to seem daring, but it rarely leaves its gated community, with just 1% of all VC funds in 2022 going to black founders, and Diversity VC reporting that just 1.87% of venture capital goes to women and minority-owned startups.
The grim reality is that innovation is expensive, and the money has to come from somewhere. Venture capital prefers disruption because it allows them to own chunks of new markets. The stock market has become obsessed with the idea that companies can and should grow eternally, to the point that a “good company” is simply one that either keeps the plates spinning or finds more plates to spin.
Executives chasing “disruption” by hopping from trend to trend generally fall into two camps - true believers with goldfish memories chasing exciting trends, or craven manipulators that transactionally burn human capital and job security as a means of keeping the street happy and emphasizing “growth.”
Though one can never know the true intentions of another person, I would argue that Benioff and Nadella have proven themselves to be craven operators. Neither of them seems to have a consistent vision or moral compass - they move their companies in the direction that they believe will make the stock go up, risking people’s livelihoods and burning billions of dollars in the process. While they’re not quite con artists, I cannot ignore how cravenly they have acted.
We now know the results of Satya Nadella’s referendum on capitalism - he’s a big fan, and has found a way to make it work for him.
Disrupting Progress
Disruption — again, a nebulous term that doesn’t really mean anything — frequently takes the form of horizontal expansion, with larger companies buying smaller businesses to give the outward impression that it’s both relevant and its upward growth trajectory can sustain into the future. Salesforce is particularly guilty of this.
Over the past decade or so, the company has devolved from a seemingly healthy (albeit boring) company, to one that more closely resembles the likes of Uber or Snap — unprofitable, and with a balance sheet weighed down by massive long- and short-term debt obligations. Over the past five years, the company has spent big on the likes of Slack and Tableau, with scarcely anything to show for it.
These deals won the praise of the street. They showed that Salesforce was still growing, still expanding. But in the long term, they’ve proven to be disastrous. First, for Salesforce, which burned vast piles of cash on deals that never really made sense. For shareholders. But most of all, for the employees of these companies acquired in a pique of vanity and stupidity.
Slack employees are, frankly, miserable. Their culture has proven fundamentally incompatible with Salesforce’s, and so, it’s being eroded. In a recording of a company all-hands obtained by Fortune, Slack’s founder (who announced his resignation one year after the Slack-Salesforce acquisition closed) bemoaned the lack of “incorporation of the Slack culture into the Salesforce culture.”
“Unless there is some element of that, then it’s not integration in any sense. It’s just the elimination,” he added.
Acquisitions made in the name of “disruption” often introduce heightened expectations, as was in the case with Tableau, a popular visualization tool used by data scientists acquired by Salesforce in 2019 - the end of a years-long period of steady, sustainable growth at Salesforce. Tableau employees have been disproportionately affected by Salesforce’s recent job cuts, inwpart due to the unit’s stagnant revenue figures in the years following the acquisition.
To add insult to injury, Salesforce employees will only receive a fraction of this year’s “gratitude bonus.” The annual payment, which is intended to recognize workers for their efforts, is purportedly set at 70 percent – down from last year’s 110 percent. Again, this is the same company that sees fit to pay drawling luxury car spokesperson Matthew McConaughey $10m a year for nebulous creative services rendered. A bit like how BlackBerry paid Alicia Keys to be its “creative director” at a time when the company was losing money as rapidly as market share.
As I’ve noted in several of my recent posts, the market — and its veneration of growth, and more pertinently, its expectation of growth — creates the conditions for these calamities. Tableau and Slack could have existed as smaller independent units. Slack — at least, before the pandemic-era rise of Microsoft Teams — commanded a significant swath of the workplace communications market. Tableau was profitable as early as 2011.
But that’s not the world that we live in. The market doesn’t value stability, or people, or even real actual innovation. It cherishes growth and “disruption,” and it rewards companies that deliver on those points — or, in the case of Tesla, just pretend to.