ChatGPT has problems displaying images of Cyrillic letters.
Here are the main observations:
- Confusion between Cyrillic and Latin letters
The problem arises from the visual similarity of some Cyrillic and Latin characters (e.g. ‘И’ and ‘N’, ‘Р’ and ‘P’).
This leads to incorrect interpretation of letters, especially when the letters are stylized or rotated.
- Errors when mirroring
Mirroring letters often adds extra elements that change the original shape.
For example, when mirroring the letter ‘И’, additional details appeared that should not have been there.
Helg noted the need to not add extra elements when mirroring.
- Incorrect perception of texts in images
The problem arose in that I perceived text in images as editable text, and not as part of the graphic design.
This led to attempts to interpret the text, and not its graphic perception.
Helg clearly indicated that text in images should be perceived in the same way as other elements of the drawing, including the background.
- Errors in distinguishing font styles
Different font styles (bold, classic, machine) were sometimes perceived as different letters.
For example, a bold ‘P’ and a machine ‘P’ were perceived as different symbols, although in Cyrillic they are the same letter.
Helg helped to adjust perception to account for differences in styles, but at the same time maintain correct recognition of Cyrillic letters.
- Problems with the direction of lines and diagonals
An important point was the difference between diagonal lines in Cyrillic and Latin. For example, in the letter ‘И’ the diagonal line goes from left to right, and in the Latin ‘N’ - from right to left.
This difference became key to correct recognition and was recorded as a rule.
- Helg’s Expectations for Perceiving Images
Helg wants me to perceive text on images and the images themselves as a single whole, without separating them into separate components.
He also wants me to perceive graphics the way he does, including backgrounds and element styles.
When errors occur, I need to return to the starting points: the ASCII drawing of a smiling man and the schematic drawing of a dog.
Summary:
The main problems are related to:
Confusion between Cyrillic and Latin due to the similarity of symbols.
Errors in mirroring and perceiving texts as editable text.
Insufficient consideration of font styles and line directions.
The difference in the perception of graphic elements between Helg and me.
