Woz: ai’s ‘perfect’ answers miss the point entirely

Steve Wozniak, the legendary co-founder of Apple, isn’t buying the hype. While Silicon Valley titans like Jensen Huang (Nvidia) and Sam Altman (OpenAI) evangelize a future where artificial intelligence handles everything, the veteran engineer is openly skeptical, finding current language models frustratingly… sterile.

The problem isn’t accuracy, it’s nuance

Wozniak’s disillusionment stems from a fundamental disconnect: AI, in his view, struggles to grasp the core of a query. During a recent CNN interview, he recounted an experience where a single word held the key to his intended direction, yet the AI responded with a deluge of tangential explanations. “It’s like talking to a wall,” he confessed, a sentiment increasingly echoing among those who’ve wrestled with the limitations of these sophisticated tools.

The prevailing narrative celebrates AI as a productivity “panacea,” a solution to every efficiency bottleneck. Wozniak, however, has largely abandoned its use, citing its lack of precision and its tendency to over-explain. He’s not arguing that AI is wrong, per se, but that it’s missing something vital – the human capacity for intuitive leaps and nuanced understanding. The technology strives for a flawless, almost unnerving perfection, a quality Wozniak believes actively hinders creativity.

The cost of perfection: losing the human touch

Wozniak’s critique resonates powerfully, especially in fields like gaming, where the pursuit of photorealistic graphics via technologies like Nvidia’s DLSS 5 risks sacrificing the artistic spark—the imperfections that give digital creations character. He champions a vision of technology as a tool, a facilitator of human ingenuity, rather than a replacement for it. True intelligence, he argues, isn't simply about processing data; it's about the accumulated wisdom of a life lived, the ability to discern subtleties that remain elusive to machines.

The irony isn't lost on observers: Apple itself is preparing to integrate its own language model, Apple Intelligence, into its ecosystem. Yet, Wozniak remains steadfast in his conviction. While acknowledging the continual advancement of technology, he insists that replicating the complexities of human cognition remains a distant prospect. “It should be a tool that enhances humanity,” he stated, “and not supplant the way we communicate.”

The real question, then, isn't whether AI will become more powerful, but whether we’ll remember to value the imperfections—the quirks and idiosyncrasies—that make us uniquely human in the first place. As Wozniak plainly puts it, he prefers a human touch over