Do Open AI reduce the "sampling_steps": for image generation?

In DALL-E, “sampling steps” refer to the number of iterations the model goes through to generate an image from the input prompt. Higher sampling steps can lead to more detailed and refined images, but they also require more computational resources and time to produce.

Here’s a brief overview of what it means:

  1. Higher Sampling Steps: More iterations in the image generation process, leading to potentially higher quality and more detailed images. It allows the model to better refine the image and reduce noise.
  2. Lower Sampling Steps: Fewer iterations, resulting in faster image generation but potentially less detail and higher noise levels.

For example, specifying 60 sampling steps means the model will iterate 60 times to generate and refine the image, aiming for a higher level of detail and accuracy based on the input prompt.

on of this images is older version, other 2 are I generate today, same prompt, its clear that they reduce the sampleing steps, so means, they are getting more money and providing bad images.