More than AI Recalibration Behind Layoffs at Google and Microsoft

I promised yesterday that I would deliver this post today. That's one promise kept. If somebody let you down today, it wasn't me.

Back in the sepia-toned days of my early career – when I wore a zoot suit, a skinny tie, and a gangsterish fedora on workdays – companies tended to resort to layoffs only in extreme circumstances, such as severe economic recessions. Their reticence to slash staff wasn’t motivated entirely by selflessness. On mahogany row, the thinking was that it would be too costly and counterproductive to lay off scores of workers only to have to rehire them, or others like them, when better days inevitably returned. 

Today, the corporate aristocracy of C-level executives is less squeamish about reaching for the hatchet. In information technology, especially at the larger companies, executives have no compunction about slashing employees at the slightest economic provocation. Now, though, economic distress is not required. Increasingly, employees are summarily dispatched on a regular basis, even when companies are growing like kudzu and making copious profits. 

In the news recently, we learned of layoffs at both Microsoft Azure and Google Cloud, commercial entities that are not in any way experiencing a business downturn or faltering customer demand. They’re both growing furiously and generating exemplary profits. Even so, employees are being shown the door. 

At Google, employees working across several cloud teams discovered that their jobs were being eliminated, according to an internal document reviewed by Business Insider. The report indicated that teams focused on consulting, partner engineering, and sustainability were among those in the line of not-so-friendly fire.

Officially, when asked about the layoffs to growth businesses, representatives of Google and Microsoft mouthed the usual anodyne platitudes, superficial and only superficially plausible. It’s as if they’re singing from the same corporate hymnbook.  What follows is an example of what they’ve said, or not said:

"As we've shared before, we continue to evolve our business to meet our customers' priorities and the significant opportunity ahead," a Google spokesperson told BI in a statement. "We maintain our commitment to investing in areas that are critical to our business and ensure our long-term success."

To which, one can only reply: Well, who doesn’t? 

Crusade for Candor and Decency

The official Google statement does not say anything meaningful about why the jobs were axed. Other reports, quoting officials at Google and Microsoft, cited a need to allocate focus and resources to AI initiatives, which seems reasonable until you ask yourself why jobs in heady growth businesses had to be sacrificed on the AI altar. It isn’t as if Google and Microsoft are practicing zero-sum budgeting across their business units, where overall resources are set at a fixed overall level and must be finitely apportioned, prioritized (or not) on an austere basis of comparative merit. 

Despite the need to invest proportionally in growth markets such as AI, there is no corollary logic that demands investments in other growth markets should suffer for the furtherance of AI. In fact, if you consider that most AI workloads are running in public clouds such as Azure and Google Cloud, does it make sense to slash personnel and reduce resource allocations destined for cloud infrastructure? The two are mutually reinforcing. 

So, listen, there’s some dissimulation happening here, and I don’t like it. I’ve embarked on a personal (likely doomed) crusade to inject more candor and honesty into the technology industry. I can’t tolerate any more cant, equivocation, prevarication, and euphemisms. I’m not a scold, and I’ve mellowed with age – just ask, well, a select few people I know – but there are limits to my forbearance. The industry has slipped into a formulaic vernacular that is not only opaque and otiose, but frankly disrespectful to anybody of modest intelligence. 

If there isn’t a zero-sum budgetary imperative that necessitates sacrifices from the cloud business in propitiation to AI, what’s driving the new tech austerity? 

AI is being used as a pretext for a change in corporate philosophy, particularly as it pertains to desirable or acceptable measures of profitability. It’s no longer good enough to run thriving businesses with impressive and rewarding profitability. Instead, major corporations, private and public, must run as leanly and ruthlessly as possible while powering inexorable growth and generating chart-topping profits. The scales have shifted. What was good enough before isn’t good enough now. 

Investors, including VCs and private equity as well as major institutional investors, have long pressured tech companies to run leaner, to shed staff at every opportunity, including during flush periods of hypergrowth. Mark Andreessen, from his well-feathered perch atop his eponymous venture-capital firm Andreessen Horowitz, has railed against the “laptop class” of white-collar professionals recumbent in technology companies.  He has said that “good big companies are overstaffed by 2x,” whereas “bad big companies are overstaffed by 4x or more.” Google and Microsoft would necessarily fall into one of those categories, according to Andreessen’s unforgiving parameters, as would every other large company in the technology firmament. 

Perhaps taking his cue from his boss, Andreessen Horowitz’ David Ulevitch, former founder and CEO of OpenDNS, more recently said that “half the white-collar staff at Google probably does no real work.” 

Adverbs of Probability 

Notice the use of probably,” a relatively weak adverb of probability. Ulevitch, like Andreessen, is expressing a subjective opinion. He has no corroborating data to support his assertion, which he subsequently claimed was the among “the least controversial things I've ever said."  Well, allow me to object to that statement on purely logical grounds, since the comment is prima facie contentious because it is predicated on nothing more than one man’s opinion. There might be a way for Ulevitch to defend his claims with corroborating data, but he has not done so. No evidence is offered.  Ulevitch is entitled to his opinion, of course, but that fails to move the logical needle, taking him only as far as the indulgence of a credulous audience will allow. 

I don’t scientifically know the extent to which large technology companies are overstaffed; nor do I know whether half the white-collar staff at Google probably doesn’t do any work.  The point is, Andreessen and Ulevitch don’t know, either. Their views are wholly tendentious.  You can choose to believe them, but your belief, on its own, takes you no closer to an accurate quantification of per-employee productivity at Google. 

In attempting to parse these high-profile opinions – which wouldn’t have made the business or trade press if they’d come from an anonymous howler on Twitter (sic) – we are dealing with dogma rather than scientific knowledge, somewhat ironic considering that both men would contend that fears regarding the malevolent or destructive application of technologies such as AI result, at least partly, from the ignorance or misunderstanding of those harboring such concerns. 

Other prominent VCs, including Khosla Ventures’ Keith Rabois, have expressed similar views to those of their counterparts at Andreessen Horowitz. Rabois asserted that the wave of layoffs at Meta was justified, the employees sent packing merely “extraneous”. 

The private-equity giants tend to be more discreet regarding their views on tech-company overstaffing, but their actions speak volumes. The largest private-equity companies focus heavily on “value creation,” a euphemism for continual cost-cutting and headcount reductions, the purpose of which is to drive profitability and increase returns for investors.

That was part of Ulevitch’s rationalization, too. In advising Google to dump the human deadwood, he argued that the money the cloud giant spent on irrelevant jobs robbed wealth from shareholders, including pensioners and retirees. I thank him for thinking of superannuated codgers like me, but my gratitude for his altruism does not change the fact he has presented no persuasive evidence to corroborate his allegation that half of Google’s white-collar employees are recipients of money for nothing.

Is there some waste, some measure of idleness, at Google? Probably so. Notice that I used the same adverb of probability that Ulevitch invoked, suggesting that while I believe, from my own personal and professional experience, there are inefficiencies nearly everywhere, including at Google, I am not able to provide an accurate assessment as to whether the incidence ratio at Google is low, average, or high; Ulevitch, based on the lack of evidentiary data he presented, is in no better position.  

Accounting for the AI Factor 

 At this point, let’s turn our attention to AI, which appears to be the elephant in the room. It’s still early days for AI, which should be understood as a technological means of making automation increasingly intelligent. Automated intelligence has yet to reach its apotheosis, but we can imagine that the results, in the fullness of time, will be profound. 

 AI is likely to have a considerable impact on the job market, and not only in the technology sector; but I’m of the view that it would be a mistake to demonize AI as a job destroyer. 

My reasoning involves the relationship between subject and object. AI will be wielded to degrade many jobs, making them less valuable and remunerative for those staff them and perform them. The AI itself, however, is merely an inert artefact. It does nothing without people behind the curtain determining how it will be used and programming its purpose. It’s a tool, albeit a powerful one. 

AI can be used in multifarious contexts and for countless purposes, but I have a distinct impression that it will be used extensively for what the private equity mavens call “value creation,” invariably involving the degradation or diminution of established jobs and professions. How should we – the royal “we,” as in you and I – respond?

I think we can already see how the VC, PE mavens, and institutional investors will respond. They’re already responding. They’ll invest in AI both as a business proposition and as a means of scaling business efficiency and productivity within the companies in which they’ve invested. Given that they are already counselling companies to make staffing cuts now, with presumably minimal AI in use, doesn’t logic suggest that AI will be applied to make further staff reductions in pursuit of unprecedented profitability? 

There’s a clear track record in the form book here, so that last rhetorical question is more than my subjective opinion.

Will other jobs emerge to take the place of those relegated to purgatory or oblivion in the wake of the widespread utilization of AI? Quite probably, yes, but the jury is out on how quickly that will happen and whether those jobs will be as rewarding (in all respects) as the jobs left behind.

But I am moving too fast and careening toward some particularly knotty mangrove thickets. There will be plenty of time ahead to discuss what the “fourth industrial revolution” will deliver and how we should prepare for it.