The experiment compared ChatGPT’s editing results with Adobe Lightroom and Photoshop, revealing that while traditional apps rely on manual adjustments, AI editing depends entirely on descriptive language. Vague commands like “brighten this photo” often produced subpar results. In contrast, precise, context-rich prompts consistently delivered professional-quality edits.
Three categories of prompts stood out:
- Detailed instructions: Requests such as “brighten the background while keeping skin tones natural” or “remove shadows from the lower left corner” gave ChatGPT clear direction, resulting in balanced, accurate edits.
- Contextual style references: Adding cues like “cinematic look with warm tones” or “portrait with studio lighting” helped the AI align with specific aesthetics, reducing randomness and enhancing creativity.
- Step-by-step prompts: Breaking edits into smaller stages—first adjusting brightness, then refining sharpness, then applying filters—produced more controlled outcomes than trying to do everything at once.
The testing also showed that overly broad prompts like “make this photo better” led to inconsistent results, while prompts lacking subject or environmental detail confused the AI. The key takeaway: precision, context, and sequencing are essential for effective AI photo editing.
This experiment underscores a broader trend in AI adoption. Success depends not only on the technology but also on how humans communicate with it. As generative AI becomes more integrated into creative workflows, prompt engineering is emerging as a critical skill—just as important as mastering Lightroom sliders or Photoshop layers.
For photographers, designers, and casual users alike, this means learning to “speak the language” of AI editing. With the right prompts, ChatGPT can deliver results that rival traditional software, offering speed and accessibility without sacrificing quality.
