Video: Did Nano Banana Pro Kill Photography or Just Photo Editing?

Is the era of trusting our eyes finally over? While the photography community has been bracing for the impact of artificial intelligence, few predicted just how quickly the technology would master the subtle nuances of optical physics and authentic lighting.

This confronting reality is the focus of a new investigation by Andy Hutchinson, a British expat, photographer, and YouTuber based on the South Coast of NSW.

In his latest video, Hutchinson explores the startling capabilities of Google’s new generative imaging model, Nano Banana Pro (part of the Gemini 3 suite). Comparing it against Midjourney v5 and his own portfolio, his discoveries paint a picture of a technology that has not just improved, but fundamentally broken the barrier between reality and simulation.

A quantum leap in realism?

The video demonstrates a terrifying leap in fidelity. Where 2023's Midjourney produced 'graphic art' interpretations of prompts like an elderly man or a forest sunset, Nano Banana Pro generates images that are optically indistinguishable from high-end photography. It mimics specific camera behaviours – like lens flare and depth of field – so perfectly that even expert eyes are fooled.

More impressive, however, is its editing capability. Hutchinson shows how the model can take an existing photograph and 'relight' it – turning a flat midday shot into a golden-hour masterpiece – without altering the physical structure of the scene. Every brick on the Sydney Opera House remains perfectly in place, just lit differently. 

The tool also excels at geo-specific generation. From Jervis Bay to Patagonia, the AI can conjure geographically accurate landscapes without the photographer ever leaving their desk. It can remove crowds from tourist hotspots, clean up Google Street View screenshots into high-res travel photos, and even insert new subjects (like a surfer) with physically accurate reflections and water physics.

The end of the 'decisive moment'?

The findings lead to a sobering conclusion: the 'trust contract' inherent in photography may be dead. For over a century, a photograph was proof that someone stood in a specific place at a specific time and witnessed an event. Nano Banana Pro erases that proof. When an AI can generate a 'dawn shoot' in Patagonia while the creator sits in their pyjamas, the value of effort, patience, and the luck of good weather evaporates. Furthermore, is the temptation to use AI to turn an already good photo into an 'even better' one too much for some to handle?

Commercially, this may signal the end for stock photography and many influencer niches. Why pay for a travel shoot when you can generate the perfect, crowd-free destination shot instantly? There is already a lot of online hate for AI content, and interestingly, Instagram seems to prioritise views for low-fi timely videos over those with high production values. People still love 'real' content – the experience.

Ultimately, we are entering a 'new normal' where the image is divorced from reality. Photography must now pivot from being about the result to being about the experience. We will continue to take photos not to prove we were there, but simply because we enjoy the act of seeing (and doing). The output is now cheap commodity – the human experience remains the only thing of value.

You can follow Andy on his YouTube Channel, or see his work on his website.