‘I Applied to 2,843 Roles’: The Rise of AI-Powered Job Application Bots

Great Coverage again by 404 Media putting light on both the plight of applicants and of hiring managers.

Zuckerberg’s Law

Zuckerberg’s law and platform dynamics: In the 2000s Zuck formulated “Zuckerberg’s Law”. It was printed in the little Orange Book everyone who joined the company would get during onboarding and goes something like:

“When friction to share information is reduced to zero, sharing trends toward infinite - and people sharing about themselves grows by approximately 2x every year and that is good for Facebook”.

While this seems a bit banal and leaves out some key adverse effects (“when content goes towards infinite, noise drown out quality signal” for example), it’s actually rather insightful.

I would rephrase it as follows:

“When content supply starts massively exceeding demand, the formation of AI driven platforms is inevitable to manage information and match supply to demand”

All of FAANG is AI platforms since the 2010s.

Why is more supply good for them?

Because platforms get paid by both participants in the marketplace and competition on the supplier side manifests itself in direct revenue - ads.

When many more products vie for the eyeballs controlled by the platform, ad auction prices reflect that.

Platforms have also changed since the early days. With little competition in their space, their ability to extract value from both demand and supply side has exploded. Early Facebook showed you 7 items from your friends for every ad, the ratio is now 10 items Meta wants you to see (ads, new products) for every item you want to see.

What is missing, completely, is “quality”, It has proven unnecessary as a selective criteria - because there’s no real competition and user habits are so entrenched, there’s almost no risk of churn.

The role of GenAI

The second key observation, in context of generative AI is:

“Generative AI decreases friction on digital content creation towards zero”, which, when seen with the above means

“AI platforms either emerge or existing platforms are massively strengthened in importance (“is good for Facebook”) across any exposed surface managing supply and demand”
The signal value of the information on these surfaces will reduce more and more over time.
While initially value for participants is importantly, dependency on the platform develops and platforms can extract value from market participants that rises with the competition.

Implications for Recruiting

Recruiting is such a surface - a market matching applicants with roles that relies on digital content (resumes/cover letters/applications to carry signal:

AI is multiplying the overall information transmission, burying recruiters while degrading the ability to filter signal from noise.
In their pain, most companies are driven towards platforms (workday, linkedin) offering salvation, but these platforms not only have their own metrics to optimize for, they also can’t filter for quality: No AI is able to unwind the tightly knit yarn of AI written applications.

In the end awaits platform dependency and a vastly degraded experience for everyone but C-Level VIPs.

No easy solutions

It will only get worse. Already for years companies have added more and more hurdles to hiring processes in order to filter applicants, often on weak evidence (leet-code interviews, pseudo scientific quadrant mapping, etc) - but now an apocalyptic deluge of content is polluting the inbound at the earliest possible point in time and any artificial friction added will affect … everyone, including the best profiles.

How to fix it: Leadership.

The courage to acknowledge that the box is broken, that things no longer work and that magic fixes (AI) do not work.

The solution has to be outside the box. Questioning the default.

Do you have to hire on a global platform? Are there better channels that add signal that can separate serious applicants from the noise?