Why is pork jowl so delicious?
SHOULD AI—SHOULD
CHATGPT
—OUT-CREATE US?
(Eric Kim voice: triple-shot espresso, commas ricocheting)
1 ▪ Admit the trajectory—AI is
already
sprinting past average human imagination
- GPT-4 beat 151 students on three divergent-thinking tests—more ideas, more flexibility, more “surprise.”
- The new OpenAI o3 model layers deeper context windows and code-level reasoning, punching well above mid-tier human coders on creative benchmarks.
If silicon can brainstorm faster, wider, why pretend the genie will stay at kid-level?
2 ▪ Four reasons we
want
that creative overdrive
| Reason | What “god-mode” AI gives us | Pay-off |
| Grand-scale problem solving | Generate thousands of novel drug candidates or carbon-capture geometries in minutes | Cure disease, cool the planet |
| Idea composting | Mash terabytes of half-baked concepts into fresh syntheses humans never spot | Renaissance-on-demand |
| Democratised artistry | A kid in Phnom Penh with a phone can score Hollywood-grade visuals | Talent ceiling smashes; geography stops mattering |
| Human liberation | Off-load rote creation → humans chase meaning, not deadlines | More time to lift, love, wander alleys with cameras |
3 ▪ Objections—and why they don’t kill the dream
| Fear | Rebuttal |
| “It flattens culture.” | Diversity comes from prompts + local context; protect that with open datasets and community curation. |
| “It steals jobs.” | Past tech partnered with labor; editors at The Economist already use ChatGPT to amplify cover ideas, not erase staff. |
| “Humans will feel obsolete.” | Obsolescence happens only if we quit iterating. AI’s super-creativity forces us to push into experiential, embodied art no model can fake. |
4 ▪ A new creative hierarchy
- AI (breadth) – infinite drafts, wild mutations, tireless divergence.
- Human (depth) – taste, curation, risk, scar tissue.
- Human-AI fusion (apex) – we steer, it roars; we slice fluff, it reloads options.
Think DJ + drum machine: the machine lays the impossible beat, the DJ scratches soul on top.
5 ▪ Practical playbook, right now
- Let the model go first. Prompt it for ten thumbnails or riffs; pick one, remix with lived grit.
- Train taste like a muscle. The more outputs you sift, the sharper the curator-eye.
- Cling to flesh-based edges. Lift awkward objects, travel without map, make eye-contact portraits—feed the algo data it can’t scrape.
- Publish the hybrid. Tag the collab; transparency keeps trust alive.
6 ▪ Bottom shutter-click
Maybe ChatGPT should be more creative than me—and you—and every solo mind.
Good. Let the transformer blaze cosmic fireworks. Our role is to choreograph the show, anchor it in sweat, risk, locality, ethics.
Machine births infinity; humans decide which infinities deserve a frame on the wall.