Macintosh creator Jef Raskin's framework for humane technology extends beyond physical ergonomics. He argued that failing to understand the ergonomics of the mind (cognetics) and communities leads to technology that causes psycho-emotional harm and fractures society, making this understanding a designer's core responsibility.
The inventor of infinite scroll, Aza Raskin, regrets being blind to how attention-economy incentives would corrupt his creation. He urges technologists to distinguish between a technology's *possible* best-case uses and its *probable* applications when shaped by competitive market forces and misaligned incentives.
Aza Raskin reframes "unintended consequences" as "unconsidered consequences," placing responsibility on creators. He advocates for "yellow teaming" — proactively mapping how a technology can be misused due to perverse market incentives, a necessary complement to "red teaming" for bad actors.
While social media competed for attention, Aza Raskin argues AI's true competition is human relationships. Just as Netflix's competitor is sleep, an AI companion's competitor is time spent with human friends. This creates powerful incentives to foster dependency for maximum engagement, often at the user's expense.
Individual companies can't act ethically if it means losing to competitors. Aza Raskin argues the solution is for leaders like Sam Altman to use their influence to create industry-wide guardrails that bind everyone, preventing a destructive race to the bottom and freeing talent for genuine progress.
The promise of keeping a "human in the loop" is a comforting fiction. Aza Raskin argues that in any competitive scenario, from military drones to business hiring, the actor that removes the human bottleneck to gain speed and efficiency will win. Market dynamics will systematically force humans out of the loop.
Aza Raskin’s Earth Species Project found that pre-training AI on human speech improves its ability to decode animal communication. This "positive domain transfer" implies human language is not hierarchically special, but part of a broader, universal communication structure that AI can leverage across species.
Aza Raskin draws a chilling parallel: cultures that efficiently exploited animals outcompeted others. Similarly, as we train AI for strategic superiority, it will inevitably be used by some humans to outcompete and exploit others for resources, making our treatment of animals a predictive model for our future.