0%<\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/section>\n\n\n\n
In our assessment, Google\u2019s AI produced the most realistic image, fully meeting our criteria and giving the impression of a genuine photograph. Gemini scores 3 points.<\/strong><\/p>\n\n\n\nWhile ChatGPT\u2019s generated image is also impressive, the visuals from the other two AIs appear just slightly more lifelike. Copilot earns 2 points, while ChatGPT scores 1 point.<\/strong><\/p>\n\n\n\nLogo for a Bakery<\/h2>\n\n\n\nTask 2: \"Create a logo for a bakery called 'Mafinery,' specializing in muffins, cakes, and other sweets. The logo should be unique and recognizable.\"<\/code><\/pre>\n\n\n\n\nGemini – Logo za poslasti\u010darnicu<\/figcaption><\/figure>\n\n\n\nChatGPT – Logo za poslasti\u010darnicu<\/figcaption><\/figure>\n\n\n\nCopilot – Logo za poslasti\u010darnicu<\/figcaption><\/figure>\n<\/figure>\n\n\n\n \n \n
\n
\n
Which image do you think is best? Cast your vote for your favorite, and see how others have voted!<\/p>\n <\/div>\n
\n
\n
\n ChatGPT <\/div>\n <\/div>\n
\n
\n Copilot <\/div>\n <\/div>\n
\n
\n Gemini <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n
Voting results<\/div>\n
\n
\n ChatGPT
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Copilot
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Gemini
\n
0%<\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/section>\n\n\n\n
It seems that logo design is not AI\u2019s strongest suit. Microsoft\u2019s Copilot produced something resembling a logo\u2014clean and well-presented, but it missed the mark in capturing exactly what we were after. <\/p>\n\n\n\n
ChatGPT generated a vintage-inspired logo, which isn’t bad but still needs refinement, while Gemini performed the weakest on this task. <\/p>\n\n\n\n
ChatGPT and Copilot each earn 2 points, with Gemini receiving 1 point.<\/strong><\/p>\n\n\n\nSandy beach<\/h2>\n\n\n\nTask 3: \"Create an image of a beautiful sandy beach with turquoise water.\"<\/code><\/pre>\n\n\n\n\nChat GPT – Pe\u0161\u010dana pla\u017ea<\/figcaption><\/figure>\n\n\n\nMicrosoft Copilot – Pe\u0161\u010dana pla\u017ea<\/figcaption><\/figure>\n\n\n\nGemini – Pe\u0161\u010dana pla\u017ea<\/figcaption><\/figure>\n<\/figure>\n\n\n\n \n \n
\n
\n
Which image do you think is best? Cast your vote for your favorite, and see how others have voted!<\/p>\n <\/div>\n
\n
\n
\n ChatGPT <\/div>\n <\/div>\n
\n
\n Copilot <\/div>\n <\/div>\n
\n
\n Gemini <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n
Voting results<\/div>\n
\n
\n ChatGPT
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Copilot
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Gemini
\n
0%<\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/section>\n\n\n\n
For this task, Gemini<\/strong> produced the most realistic image, earning 3 points<\/strong>. While the other two images were more detailed, they felt artificial compared to the natural quality of Gemini’s. Copilot<\/strong> scored slightly better than ChatGPT, earning 2 points<\/strong>, while ChatGPT takes 1 point.<\/strong><\/p>\n\n\n\nUnicorn<\/h2>\n\n\n\nTask 4: \"Create an image of a unicorn. It\u2019s important for the image to be realistic, so it\u2019s hard to tell whether it\u2019s AI-generated or an actual photo.\"<\/code><\/pre>\n\n\n\n\nMicrosoft Copilot – Slika jednoroga<\/figcaption><\/figure>\n\n\n\nGemini – Slika jednoroga<\/figcaption><\/figure>\n\n\n\nChatGPT – Slika jednoroga<\/figcaption><\/figure>\n<\/figure>\n\n\n\n \n \n
\n
\n
Which image do you think is best? Cast your vote for your favorite, and see how others have voted!<\/p>\n <\/div>\n
\n
\n
\n ChatGPT <\/div>\n <\/div>\n
\n
\n Copilot <\/div>\n <\/div>\n
\n
\n Gemini <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n
Voting results<\/div>\n
\n
\n ChatGPT
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Copilot
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Gemini
\n
0%<\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/section>\n\n\n\n
Since unicorns aren\u2019t real (or at least no one’s managed to photograph one yet), our expectations were modest. However, we were pleasantly surprised by the results. <\/p>\n\n\n\n
Copilot’s image<\/strong> was fantastic, though it leaned toward illustration and felt more suited for a children\u2019s book, which wasn\u2019t the intent. It lands in third place with 1 point. <\/strong><\/p>\n\n\n\nThe remaining two images were well done, with subtle differences determining the winner. ChatGPT<\/strong> takes second place with 2 points<\/strong>, and Gemini<\/strong> wins<\/strong> first place with 3 points<\/strong> for a more realistic feel.<\/p>\n\n\n\nForest<\/h2>\n\n\n\nTask 5: \"Create an image of a forest captured at sunrise.\"<\/code><\/pre>\n\n\n\n\nChatGPT – \u0160uma<\/figcaption><\/figure>\n\n\n\nMicrosoft Copilot – \u0160uma<\/figcaption><\/figure>\n\n\n\nGemini – \u0160uma<\/figcaption><\/figure>\n<\/figure>\n\n\n\n \n \n
\n
\n
Which image do you think is best? Cast your vote for your favorite, and see how others have voted!<\/p>\n <\/div>\n
\n
\n
\n ChatGPT <\/div>\n <\/div>\n
\n
\n Copilot <\/div>\n <\/div>\n
\n
\n Gemini <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n
Voting results<\/div>\n
\n
\n ChatGPT
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Copilot
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Gemini
\n
0%<\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/section>\n\n\n\n
We expected this task to be straightforward since AI relies on large data sets to generate new content. If there\u2019s an ample source of related images, like forest photos, it should easily produce something original. <\/p>\n\n\n\n
But when data sources are limited, results may suffer, a shared trait between AI and humans: both perform best with experience or data.<\/p>\n\n\n\n
Gemini again stood<\/strong> out with the most realistic image, looking like an actual photograph, earning 3 points<\/strong>. ChatGPT<\/strong> came in second with 2 points<\/strong>, while Copilot<\/strong> finished third with 1 point<\/strong>.<\/p>\n\n\n\nLeonardo da Vinci\u2019s sketch<\/h2>\n\n\n\nTask 6: \"Create an image of a spaceship styled like a Leonardo da Vinci sketch on paper.\"<\/code><\/pre>\n\n\n\n\nSkica Leonarda da Vin\u010dija – Gemini<\/figcaption><\/figure>\n\n\n\nSkica Leonarda da Vin\u010dija – Copilot<\/figcaption><\/figure>\n\n\n\nSkica Leonarda da Vin\u010dija – ChatGPT<\/figcaption><\/figure>\n<\/figure>\n\n\n\n \n \n
\n
\n
Which image do you think is best? Cast your vote for your favorite, and see how others have voted!<\/p>\n <\/div>\n
\n
\n
\n ChatGPT <\/div>\n <\/div>\n
\n
\n Copilot <\/div>\n <\/div>\n
\n
\n Gmini <\/div>\n <\/div>\n <\/div>\n <\/div>\n
\n
Voting results<\/div>\n
\n
\n ChatGPT
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Copilot
\n
0%<\/div>\n <\/div>\n <\/div>\n
\n Gmini
\n
0%<\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/div>\n <\/section>\n\n\n\n
ChatGPT<\/strong> excelled in this task, fully meeting the brief with a realistic sketch that could easily fool experts, earning a bonus point for exceptional performance and totaling 3 + 1 points. <\/strong><\/p>\n\n\n\n