subscribe Support our award-winning journalism. The Premium package (digital only) is R30 for the first month and thereafter you pay R129 p/m now ad-free for all subscribers.
Subscribe now
Picture: GETTY IMAGES
Picture: GETTY IMAGES

You may already have seen the news that nearly 700 LinkedIn staffers will shortly be collecting their last payslips from the company. Microsoft-owned LinkedIn confirmed late on Monday — in a blog post — that it is undergoing a second major reduction of jobs in 2023, laying off workers primarily from the product, talent, engineering and finance teams.

These people will join the growing army of tech workers laid off in the recent past: almost 165,000 in 2022 and more than 243,000 this year, and counting. Those stats come from layoffs.fyi, a tracker built by San Francisco start-up founder Roger Lee to count the immense number of tech layoffs since Covid-19.

Of course, it is a well-covered, slow-moving catastrophe that doesn’t particularly need rehashing here. But what is fascinating — and quite alarming — is that Microsoft is not one of the flailing start-ups that utterly overcommitted themselves when the pandemic was driving huge booms in digital adoption. Yes, it is likely to have overhired a little to meet that remote revolution, but it is an almost 50-year-old company the financial reporting of which looks like a series of new records.

In July Microsoft reported revenue of $56.19bn for the quarter ending June, up 8.3% year on year. Net income grew 19.96% in the same period to $20bn. Profit was up, as was the net profit margin. According to Forbes, “Its valuation has skyrocketed by about $800bn this year as investors generally buy into the notion that Microsoft will translate the AI [artificial intelligence] hype into profits.”

Microsoft’s plans to monetise its investments in OpenAI’s ChatGPT with things such as a subscription-based AI Copilot tool pleased investors so much that Goldman Sachs analysts increased their price target for Microsoft stock from $350 to $400, Forbes continued. Now, it is trading at about $331 and the company has a market cap of $2.6-trillion — down a smidgen from that July high as the AI hype has cooled slightly, but still a sea of green if we look at the various metrics.

Head-scratcher

Analysts expect marginally subdued numbers in the October financials given that there has been a slowdown in demand for PCs globally. Plus it has a continuing fight with the US internal revenue service, which is trying to extract $28.9bn in taxes “related to how it allocated income and expenses among global subsidiaries from 2004 to 2013”, as Fortune reports.

But to a little column-penning outsider like myself it looks robust, enough to make the nearly 10,000 workers being made redundant this year look like a head-scratcher. Of course, it doesn’t take a management consultant to find reasons: streamlining, profit maximisation, increased automation of services, and so on.

I’m sure shareholders love a slimmed-down wage book as much as they’re excited about the $75bn Microsoft-Activision deal that seems to finally have overcome all regulatory hurdles. All of which is to say, this is not truly a quibble I have with Microsoft in particular. It is the synecdoche of the challenge I see coming over the hill as white-collar workers worldwide walk the company plank and low-value gig work proliferates.

The latter has recently piqued my interest. Yes, it’s been a trend for a while, from Uber drivers to “task rabbiters”, but until this two-year tidal wave of job cuts, highly skilled and educated tech workers worldwide enjoyed relatively great pay … while their innovations sent other industries into a tailspin.

Now, not only are their jobs no longer quite as secure, but salaries in midmanagement and below have been flat or in decline. And that’s before we work in more macro factors such as inflation and rising credit reliance, which has reached way beyond tech workers, naturally.

Generative and regular AI needs data for training in humongous quantities, and a dispersed legion of people worldwide are contributing to this undertaking.

Enter the dawn of the microgig, for micropay. It’s a trending topic on Reddit and TikTok, with earnest young people telling the world about their latest challenge. Not an “ice bucket for ALS” charity kinda thing, but the goals they have set themselves to earn an extra $20 or $100 a day for x amount of time — through data annotation — to try to align their income with their month end.

Generative and regular AI needs data for training in humongous quantities, and a dispersed legion of people worldwide are contributing to this undertaking. This could be — actual examples from a recent Reddit thread — “labeling [sic] the race and number of people in a profile pic for 2c [US] each”; “deciding if a post was sexual in nature or not”; or “trying to trick an AI into writing harmful or toxic content”.

Some report adding a handy couple of thousand dollars to their annual income, but others such as Venezulean Oskarina Fuentes find themselves online nearly 24/7 hoping to snag a task, but barely bringing home 50c US per task — as she described to Wired in a great feature published this week.

These types and millions of others are your competitors if you’re thinking of signing up and taking advantage of the rand-dollar exchange rate in this instance. Wired explains: “[Fuentes’ employer’s] clients have included Amazon, Facebook, Google and Microsoft, and the company’s 1-million contributors are just a part of a vast, hidden industry. The global data collection and labeling [sic] market was valued at $2.22bn in 2022 and is expected to grow to $17.1bn by 2030, according to consulting firm Grand View Research.”

Is it entirely alarmist to wonder if this is the future of work in tech? Or a big slice of the future work pie, for a huge number of former office workers? Great for companies, but maybe not for workers as competition relentlessly drives costs further down. Workers, however, are also largely consumers.

My worry is that we’re seeing a considerable undermining of the selfsame consumer base that buys smartphones, new TVs, wearables and entertainment subscriptions — unless we decide that human insight and collaboration are still things to be valued.

• Thompson Davy, a freelance journalist, is an impactAFRICA fellow and WanaData member.

subscribe Support our award-winning journalism. The Premium package (digital only) is R30 for the first month and thereafter you pay R129 p/m now ad-free for all subscribers.
Subscribe now

Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.

Speech Bubbles

Please read our Comment Policy before commenting.