Death Star Hanfest: Why GPT-5 Feels Like Lobtailing, Not a Leap

Incremental upgrade, not a revolution

OpenAI’s GPT-5 has arrived amid a storm of expectations and breathless corporate metaphors. Reporters and users who tested the model describe it as a refined product — smoother, more polished, but not a paradigm shift. That mirrors recent arguments comparing model releases to smartphone refresh cycles: mostly iterative improvements focused on user experience rather than tectonic technological change.

Sam Altman leaned into the spectacle. He teased the release with an image of a looming Death Star and praised GPT-5’s supposed PhD-level intelligence and life-saving potential. But bold marketing claims have met skeptical responses: some users pushed back after feeling features they liked in earlier versions were missing, and OpenAI restored a prior personality option after customer complaints. That response suggests GPT-5’s launch was as much about perceived user experience as about demonstrable performance leaps.

Hype versus observable limits

The pattern is familiar. Company leaders and commentators trade dramatic warnings about runaway AI and superintelligence, even as day-to-day interactions with models reveal glaring limitations. The article points out the mismatch starkly: teams warn about existential risks while the newest flagship model still struggles with simple tasks, like counting letters in a word reliably.

That dissonance feeds a broader hype cycle. Massive investment, data-center buildouts, and environmental and economic stakes create pressure to market each release as a milestone. Yet many recent breakthroughs in AI haven’t been about chatbots at all — they’ve been in other subfields that enable immersive experiences, new creative tools, or entirely different interfaces.

A whale, a chatbot, and the limits of confident answers

A small, human-scale example helps illustrate the gap between flash and substance. The author’s wife sent him a video of a whale repeatedly slapping its tail on the water. A quick query to ChatGPT returned a confident explanation: the behavior is called "lobtailing" and here are several possible reasons why whales do it.

That answer was useful but ultimately overconfident. A conventional web search led the author to an evocative essay by Emily Boring, which captures the scientific uncertainty around the behavior. Researchers debate whether tail-slapping is a feeding tactic, a communication signal, or a posture — the action consumes a lot of energy, and that very cost may make it a credible signal of fitness or urgency.

The anecdote shows both the utility and the limitation of chat models: they can summarize and make information accessible quickly, but they can also smooth over nuance and present contested or uncertain knowledge too definitively.

Where the real breakthroughs are appearing

There are undeniably impressive AI products and genuine leaps: earlier iterations of ChatGPT, DALL·E, NotebookLM, and newer tools like Google DeepMind’s Genie 3 (which generates immersive, navigable 3D scenes from text prompts) demonstrate the diversity of innovation. Some of the most interesting advances may be coming outside of plain conversational agents — in multimodal systems, generative visual or spatial tools, and domain-specific applications that create new user experiences.

Still, marketing narratives and the incentives of venture capital and public attention push leaders to frame incremental advances as epochal. That creates a cycle of overclaiming and reactive skepticism: bold metaphors and apocalyptic warnings meet measured, sometimes underwhelmed user experiences.

Lobtailing as a metaphor for attention

The author borrows a biologist’s interpretation of whale tail-slapping: because it is energetically costly, it serves as a credible signal — "pay attention, I am important." In AI, similar energetically expensive displays take the form of extravagant product launches, eye-catching demos, and alarming predictions. They draw attention, justify investment, and broadcast confidence — whether or not they correspond to a genuine, world-changing leap.

That dynamic explains why many high-profile releases feel like "lobtailing": noisy and attention-getting, signaling importance more than delivering clear, structural change. The article encourages skepticism without dismissing real progress. There are exciting innovations coming, but not every flagship model release is the point at which the world is irreversibly altered.

What to take away

Expect incremental improvements from most model updates, and treat dramatic claims with a dose of healthy skepticism. Celebrate the breakthroughs when they arrive — especially those that create new categories of experiences — but be aware of the incentives that drive hype. Sometimes what looks like a revolution is, in truth, a very expensive tail-slap meant to make us watch.