A further consequence of automation is that it can quietly exclude those who do not know how to optimize for the algorithm. Not everyone knows how to use AI to craft a CV that passes automated filters. In many domains now—from recruitment to writing platforms—success depends increasingly not only on the quality of one's work, but on understanding the logic of the algorithm itself.
That is very true and I would add how do you keep ahead of the algorithm changes. As the last LinkedIn change showed it can be a total disaster. Companies who provide these platforms won’t be publishing the changes publicly so it’s going to become a game of whack-a-mole. We need way more transparency of algorithm model and changes and they need to be published if they are making decisions about people.
This resonates, Neil. The story of Sam captures something that’s become a systemic failure dressed up as efficiency. The original intent - democratising access, removing gatekeepers, giving every candidate a fair shot, was genuinely good. But somewhere between that vision and today’s reality, we traded human judgment for pattern-matching at scale, and called it progress.
What strikes me is that AI is now compounding the problem from both ends. Candidates are using it to engineer CVs that satisfy the algorithm. Companies are using it to screen those CVs. So we’ve essentially created an arms race between two machines, with actual human potential trapped somewhere in the middle. The signal-to-noise problem you describe isn’t just getting worse - it’s accelerating, and neither side can easily step off the treadmill unilaterally.
I agree it needs to be human-centred, but I share your uncertainty about how we get there in a digital, highly connected and high volume world. Perhaps the honest starting point is that we stop pretending these systems are neutral. They’re not filtering for the best candidates; they’re filtering proxies who meet predefined tick boxes and resemble past hires. Naming it as a choice might at least open the door to making a different one.
A further consequence of automation is that it can quietly exclude those who do not know how to optimize for the algorithm. Not everyone knows how to use AI to craft a CV that passes automated filters. In many domains now—from recruitment to writing platforms—success depends increasingly not only on the quality of one's work, but on understanding the logic of the algorithm itself.
That is very true and I would add how do you keep ahead of the algorithm changes. As the last LinkedIn change showed it can be a total disaster. Companies who provide these platforms won’t be publishing the changes publicly so it’s going to become a game of whack-a-mole. We need way more transparency of algorithm model and changes and they need to be published if they are making decisions about people.
This resonates, Neil. The story of Sam captures something that’s become a systemic failure dressed up as efficiency. The original intent - democratising access, removing gatekeepers, giving every candidate a fair shot, was genuinely good. But somewhere between that vision and today’s reality, we traded human judgment for pattern-matching at scale, and called it progress.
What strikes me is that AI is now compounding the problem from both ends. Candidates are using it to engineer CVs that satisfy the algorithm. Companies are using it to screen those CVs. So we’ve essentially created an arms race between two machines, with actual human potential trapped somewhere in the middle. The signal-to-noise problem you describe isn’t just getting worse - it’s accelerating, and neither side can easily step off the treadmill unilaterally.
I agree it needs to be human-centred, but I share your uncertainty about how we get there in a digital, highly connected and high volume world. Perhaps the honest starting point is that we stop pretending these systems are neutral. They’re not filtering for the best candidates; they’re filtering proxies who meet predefined tick boxes and resemble past hires. Naming it as a choice might at least open the door to making a different one.