The Complicated Relationship Between Technology and Luddites

When AI first strode purposefully into the market, corporate executives perceived its commercial capacity to displace personnel. At the same time, executives downplayed AI’s potential to reduce staff count. While you still need all your employees, why antagonize them with future scenarios in which their services are no longer required?

In taking a calculated, clinical stance, corporate executives never lost sight of the pot of gold at the end of the rainbow. In their eyes, AI would help them save money by increasing efficiency, boosting productivity, and inexorably reducing headcount where it wasn’t commercially warranted.

As I’ve mentioned before, automated mechanisms don’t get sick, don’t require benefits, don’t take vacations, don’t need breaks, never go on strikes, and can work around the clock. They’re perfect employees, provided they can do their jobs adequately.

So, the official top-down narrative from the corporate penthouse was all about how AI would help employees become more productive, about how the technology would make knowledge workers more efficient while saving them from rote drudgery. Well, yes, to a point that is true, but it was never the end game.

Sure, AI might one day lead to breakthroughs that were previously unattainable, but well before that day arrived, AI is intended to displace personnel and perform work historically done by human beings.

For many CEOs and CFOs, the animating vision of AI was its capacity to reduce payroll and to cut costs. Words can be used to flatter or deceive — and sometimes to tell the truth — but actions are what communicate irrefutable intent. By all means, listen to what is said, but always pay closer attention to what is done.

AI is a neutral technology, like all technologies. It doesn’t have a will or an intent. People supply the will and intent, and they decide how technologies will be applied in the workplace and the world beyond. Specifically, the people who make these decisions are the ones whose offices reside along what we call mahogany row. This is where you’ll find the executive team when it reports to the office, which is more often these days than it was during the immediate post-COVID period.

AI and Software Developers

In past missives here, I’ve discussed the areas where AI is most likely to be applied, specifically in processes that are relatively closed and do not involve intensive interaction with other people. AI is not well suited for use in roles that involve intensive interpersonal communication, which tends to be inherently complex, contingent, and real-time — unpredictable, almost infinite, in its intonations, permutations, and verbal and non-verbal variations.

As for software development? Well, AI has a role there, and the corporate chieftains of the technology industry clearly apprehended a major cost-cutting opportunity.

In an article published today in Business Insider, the AI-related aspirations of technology executives are highlighted. Notably, there is this excerpt:

. . . Microsoft CTO Kevin Scott expects that within five years, 95% of all code will be AI-generated. "Very little is going to be line-by-line human-written code," he said last month on the 20VC podcast.

Still, he emphasized that humans would remain essential in shaping the high-level structure, goals, and design of software.

When we see a major technology CTO mention 95% of anything, we should take notice. That is a big number, and it carries, if realized, profound ramifications. Don’t worry, though, because we’ll still need software architects to guide and instruct the AI bots. Admittedly, that particular caveat is likely of modest consolation to most developers employed in the software industry.

Of course, by necessity we live in the present. AI isn’t as far along in software development today as Scott expects it to be five years from now. So, where are we today, and where will we be in the more immediate future? The article provides a few data points for our consideration.

Satya Nadella, the CEO of Microsoft, said Tuesday that between 20% and 30% of the code for some of the company's projects is written by AI.
Speaking to Mark Zuckerberg at Meta's LlamaCon conference, Nadella said that the precise figure depends on the programming language.
He added that Microsoft is leaning on AI for more than just code generation. "The agents we have for reviewing code — that usage has increased," he said, in a sign the company is weaving AI into the full software development cycle.
Nadella isn't the only Big Tech CEO leaning on AI for coding.
Last week, Google CEO Sundar Pichai said during Alphabet's earnings call that more than 30% of new code at the company is written by AI — up from 25% in October.
At LlamaCon, Nadella asked Zuckerberg how much of Meta's code was created by AI, but the social media boss couldn't give a precise figure.
Instead, Zuckerberg gave a prediction for the AI agents Meta's building to help write and test code for its Llama models: "Our bet is that in the next year, probably, maybe half the development is going be done by AI as opposed to people, and then that will just increase from there."

Automation and the Profit Motive

We can deduce, then, that AI is already involved in a significant share of the software-development lifecycle at the companies mentioned above. We can also reasonably conclude that other major technology companies are on similar AI trajectories. If the working assumptions of these companies are validated by measurable results — satisfactory code quality, increased productivity, and lower costs — the momentum and spread of AI will accelerate and grow. That’s how things typically advance, first in gradual increments and then more ubiquitously.

The article includes other observations. In a February post-earnings call, Salesforce CEO Marc Benioff said his company would pause engineer hiring in 2025 as a result of AI increasing engineering productivity by 30%. Payment company Stripe reported in January than it had laid off 300 engineers.

Technology companies, as well as those in other industries, relentlessly pursue efficiency gains and productivity increases. Much-maligned middle managers, for example, are an increasingly endangered species in most organizations.

Despite what corporate chieftains might say about their business’ broader social aims and commitments to philanthropy, companies are invariably incentivized by the profit motive. They are in business to generate profits and to increase revenue. As an employee, if you’re perceived to be impeding rather than facilitating those objectives, you are literally working on borrowed time. People lose sight of this fact, so I’m restating it here as public service.

Have no illusions about the relationship between employer and employee. It is a commercial relationship predicated on the business value of the employee to the employer. Similarly, I suppose, employees take their own considerations, commercial and otherwise, into account when deciding whether to take a job, remain with their current employer, or seek employment elsewhere. Employee motivations, however, are generally more varied than employer incentives.

Learning from the Luddites

Now, allow me to go back in time, specifically in the context of a non-fiction book that I read last year. The book was Brian Merchant’s Blood in the Machine, which explored the history of the much-maligned Luddites and drew analogies with information technology and our current era.

When we use the word luddite today, we wield it as a blunt pejorative weapon connoting a stubborn technological backwardness verging on the antediluvian. That’s not a fair characterization of the original Luddites, who were more concerned, in the time before social safety nets, with keeping their jobs to that so that they could feed their families, avoid starvation, and keep roofs over their heads. They weren’t against technology as much as they were for self-preservation. Yes, they were guilty of damaging and destroying the machinery of production, but they did not so from desperation, as a means of protesting against oppressive immiseration.

On the rare occasions that the original Luddites come to mind today, we tend to think of them as quixotic vandals, vainly tilting at machinery whose rise was inevitable. What we forget, and what the book makes plain, is that the Luddites were actually responding to the ramifications, often severe, of how technology was wielded by a business and political class that was unconcerned with the plight of those who were displaced by automated production.

Then as now, the technology itself was inert and neutral. People, invariably those wielding influence and power, determined how the technology would be applied back then in the factories of England. There are analogies that we can draw today, as AI contributes to a wave of intelligent automation that has been gathering force now for decades. Major displacement of jobs has yet to occur, but we should be thinking of how we’ll respond to it if and when it does result. If we don’t anticipate the consequences of change, we’ll have deal reactively with the disruptions and repercussions that ensue.

Automation is likely to continue making incursions in workplaces everywhere.

Make no mistake, the technology industry is not immune; perhaps, it is even in the crosshairs more than most industries. What we should be considering is how our societies will adapt to the changes so as to avoid the economic and social dislocations that occasioned so much pain and unrest during the industrial revolution.

Early returns suggest that dislocations are already happening. Those who don’t prepare for change risk being consumed by it.