3 ms·

I asked gpt4 to "draw a unicorn using svg", and it gave me way better results: <svg xmlns="http://www.w3.org/2000/svg http://www.w3.org/2000/svg" viewBox="0 0

• ukd1 3y ago
  telling it to "make it better" results in...actually better: <svg xmlns="http://www.w3.org/2000/svg http://www.w3.org/2000/svg" viewBox="0 0 100 100" width="100" height="100"> <!-- Body --> <ellipse cx="50" cy="55" rx="20" ry="28" fill="#FFC0CB" /> <!-- Legs --> <rect x="32" y="70" width="6" height="20" fill="#FFC0CB" /> <rect x="62" y="70" width="6" height="20" fill="#FFC0CB" /> <rect x="40" y="70" width="6" height="20" fill="#FFC0CB" /> <rect x="54" y="70" width="6" height="20" fill="#FFC0CB" /> <!-- Head --> <circle cx="43" cy="32" r="10" fill="#FFC0CB" /> <!-- Eyes --> <circle cx="39" cy="29" r="1.5" fill="black" /> <circle cx="47" cy="29" r="1.5" fill="black" /> <!-- Horn --> <polygon points="43,16 42,25 38,22" fill="#FFD700" /> <!-- Ears --> <polygon points="37,22 35,29 38,27" fill="#FFC0CB" /> <polygon points="49,22 51,29 48,27" fill="#FFC0CB" /> <!-- Mane --> <path d="M32 35 Q40 28 45 40 T55 50" stroke="purple" stroke-width="2" fill="none" /> <!-- Tail --> <path d="M73 65 Q75 75 80 70 T90 80" stroke="purple" stroke-width="2" fill="none" /> </svg>
• an-allen 3y ago
  surely this purely has to do with the examples in the training corpus. I reckon SVG will have more illustrative content and as a result a more accurate representation. I reckon with the right examples in the prompt to take advantage of in-context learning, it could be pretty accurate.
• raldi 3y ago
  Can you post screenshots of these results?
  • raizer88 3y ago
    copypaste it here https://www.svgviewer.dev/ https://www.svgviewer.dev/