I was reading Mitchell Hashimoto’s excellent article, Everyone Should Know SIMD, when I came across this disclaimer:
I hate that I have to do this for every post now, but I also want to note this was completely hand-written with no AI assistance.
It made me wonder whether a new distinction is beginning to emerge.
Imagine Mitchell writes the article himself, then asks a generative model to produce Zig examples illustrating the concepts. Suppose the model produces exactly the same examples he would have written. Has the quality of the article declined?
No. The resulting content is identical.
Yet we increasingly treat “handwritten” or “made without AI” as a signal of quality, authenticity or intellectual effort. The language resembles the distinction between handmade and mass-produced goods: the method of production begins to affect how we value the result, even when no observable difference exists in the finished product.
Perhaps there is a perception that generative-AI-assisted work is inherently less original. By describing something as handmade, the author may be offering an implicit assurance: this work contains more original thought than something generated by a machine.
But there may also be a deeper argument about the relationship between effort and value—and an imprecise sense of fairness when substantial effort is hidden from view.
This recalls the familiar story of a car owner who calls a mechanic to repair a dent. The mechanic examines the damage, strikes the panel once with a hammer in exactly the right place and restores it almost instantly. The customer objects to the £750 bill because the repair took only a second. The mechanic explains that the charge is £50 for the visit and £700 for the 15 years spent learning precisely where—and how hard—to strike.
The customer is not paying for the second of visible work. They are paying for the accumulated judgement that made one second sufficient.
Perhaps there is now also a market emerging for the truly human.
According to Herodotus, the Egyptian pharaoh Psamtik I ordered two infants to be raised without exposure to human speech, hoping that their first words would reveal humanity’s original language. The experiment treated human influence as contamination: remove culture, and something more fundamental might emerge.
Ramez Naam’s Crux reverses that idea. Su-Yong Shu’s consciousness has been uploaded into a quantum-computing cluster, but begins to deteriorate because computation alone cannot reproduce the biological and sensory processes her mind requires. She survives only by connecting through Nexus to the living brain of her daughter, Ling.
In one case, human input obscures the truth. In the other, it is the condition for continued existence.
Is the growing interest in proof of effort and human authorship evidence of a similar intuition: that human contribution is objectively “more better”, to borrow a phrase from my son when he was a toddler?
Perhaps we sense that human input carries something that cannot be evaluated entirely through the observable output: experience, intention, responsibility, originality or connection to a life actually lived.
This raises an uncomfortable question for software.
Would you pay more for software written entirely by humans than for software produced with substantial AI assistance, even when its reliability, maintainability, performance and security were indistinguishable?
If so, what exactly would you be paying for?
Perhaps authorship itself is becoming part of the product. Perhaps people value evidence of human effort, judgement and craft independently of measurable quality. Or perhaps “handwritten” is merely serving as a temporary proxy for properties that AI-generated work often lacks: coherence, originality, accountability and attention to detail.
That distinction matters. One position values the process intrinsically. The other values it only because the process currently predicts a better result.
As generated and human-written software become harder to distinguish, we may need to decide whether software is judged solely by what it does—or also by how, and by whom, it was made.