Germ-killing AI algorithms are being deployed against unions

Last week, we finally got some good news: researchers with the help of artificial intelligence have managed to create a special antibiotic capable of killing very rare superbugs that have hitherto resisted even the most powerful antibiotics. An astonishing algorithm, created by an artificial intelligence program, has mapped one hundred million chemical compounds to the key proteins of Cinetobacterbaumannii bacteria, which cause pneumonia and wound infections so severe that the World Health Organization has classified them as one of three “critical threats” to humanity. From mapping, AI set out to create an effective drug, using innovative mechanisms, comparable to existing antibiotics. Without the help of AI, it would be impossible to produce this life-saving antibiotic. This is clearly a shocking scientific victory.

Unfortunately, this issue has a dark side. Remember Chris Smalls, the Amazon warehouse worker who dared to stage a strike at the company’s Staten Island facility to protest working conditions during the pandemic? Smalls suddenly became famous when it was revealed that after his firing, rich and powerful Amazon executives spent hours on a conference call planning a smear campaign to undermine his grievances. However, a few years later, Smalls succeeded in creating the first, and still the only, officially recognized union of Amazon workers in the United States. Today, those successes are threatened by the same AI algorithms that produced the aforementioned germ-killing antibiotic.

Smalls’ union was a bitter defeat for Amazon management, which for years had perfected methods to defeat any union effort by any means, fair or foul. In a training video leaked in 2018, managers were trained to look for early warning signs of unionization. They were advised to use surveillance cameras outside Amazon warehouses to spot workers who linger after their shifts, possibly with the aim of persuading colleagues to unionize. or eavesdropping on their conversations to discover any references to phrases such as “a living wage” or “feeling overwhelmed,” which are seen as precursors to discussions that gradually boil down to the need for union.

Soon enough, the software replaced, or at least helped, the primitive methods of monitoring the bosses. In 2020, Recode reported that Amazon purchased GeoSPatialOperatingConsole (or SPOC) to monitor workers exposed to unionization. Later, Vice revealed how Amazon HR is cracking down on employee email lists and locking down Facebook groups to prevent late work, strikes, etc. This program categorized the characteristics and behaviors of workers into groups that either contained or did not contain those characteristics associated with union orientation. The program’s predictive power frustrated Amazon, so management continued to rely on field managers monitoring workers the old-fashioned way.

All of these things have now been overshadowed by artificial intelligence programmes. Why eavesdrop on employees or have software read their Facebook posts and pages when a central AI can simultaneously scan every Amazon warehouse for union-friendly phrases and behavior? automatically and in real time. And at zero cost! Unfortunately, new AI methods of union-busting are based on the same scientific discoveries that led to the creation of germicides.

Before AI, researchers would classify molecules based on whether or not they contained certain groups of chemicals. This was not quite as different and ineffective as Amazon’s SPOC program that ranks employees based on their tendency to unionize. By contrast, AI germ-killing programs rely on neural networks and machine learning models that are able to computationally explore chemical spaces (i.e., cyberspace) that would take humans decades to explore. Then they are trained to analyze the molecular structures of the microbe’s proteins and discover which compounds are most likely to kill it.

Exactly the same applies to AI programs aimed at defeating trade unions, with the only difference being that instead of chemical vacuums and molecules, they scour warehouses and focus on workers whose real-time data is constantly uploaded to the software, thanks to the electronic devices those must carry. Workers are everywhere in the warehouse or factory – even to the toilet. Also using neural networks, these AI-based systems learn how to automatically strategize to neutralize a specific target without caring whether the target is a bundle of proteins in the nucleus of a microbe or a group of workplace workers. Without making any distinctions, the AI ​​categorizes its targets in such a way as to make them more likely to be neutralized.

It was inevitable. Humanity has proven itself smart enough to develop artificial intelligence algorithms capable of fully decoding the proteins of a dangerous virus – without any human intervention – in order to manufacture appropriate antibiotics to kill it. Was there any doubt that multinationals like Amazon would seize this opportunity to define and shrink workplaces where AI predicts there is a greater likelihood of unionization?

Economists believe that the forces of supply and demand work covertly but reliably to ensure that technological change benefits us. This fairy tale allows them to look away from the reality of the brutal class struggle going on right under their noses, a class struggle that is destroying the lives of millions and rendering the overall economy unable to generate (at least without unsustainable levels of debt) enough. Demand for goods that technology can produce.

Warren Buffett, whose success is largely due to shattering the illusions of economists, once aptly said that class war is a fact and that it is easily won by his class. This was before algorithmic digital machines replaced store attendants, dictating the pace of work and the total monitoring system that makes the iconic scenes of Charlie Chaplin’s modern day seem like a worker’s paradise. Moreover, AI now enables multinational corporations to stifle the only institution capable of giving workers even a modicum of power in a world where they have almost nothing: the trade unions.

The class war that Buffett spoke of will force AI cloud capital into every sector against the interests of the global precariat which will now be systematically losing. Whatever one’s policies or aspirations, it should be clear to all of us that this economy is unsustainable, disgusting and therefore doomed.

* The article above is her introduction monthly column By Yannis Varoufakis in Project Syndicate originally titled “New AI Sporebusters Could Corrupt Unions Too”.