A little AI fun......

Chat certainly has improved since it's initial offerings.

Looks good, but when did an Apple //e come standard with a "built-in clock/calendar"? Did ChatGPT get confused between the //e and IIgs?

Also, was the II+ sold with a language card (I guess it probably was)? If it wasn't, then strike the 64K RAM as standard for the II+.

Still quite an impressive poster.

Close reading of every small detail is necessary to distinguish slop from content.

Other problems: the IIe has no "expanded sound", and the keyboard is not represented accurately. The case badge on a ][+ is not in the correct place. No Apple II displays available memory on the BASIC prompt. You need to type something like "PRINT PEEK(116)*256+PEEK(115)" to display it (and not all the memory below HIMEM is really available).

But we don't know what prompt(s) were used to elicit this output.

I like it in the spirit of fun as offered.

I think it conflated ProDOS features with hardware features.  The sequence "clock/calendar" appears many times in ProDOS 8 documentation, and "built-in clock driver" appears nearby.  Since the //e would likely be delivered with ProDOS, and the II+ would not be, in some sense the algorithm got that right, except it didn't fold in the fact that the enabling hardware is an add-on.

Stargeezer wrote:

The prompt was simple.  "Make a poster advertising the Apple II+ and Apple IIe in the style of a 1950's scifi movie poster". Yup pretty basic and giving Chat free creativity

Ah, that explains the non-Apple-style branding of the Plus.  The poster's use of a + sign doesn't match Apple's own advertisements because their style guide insisted that the word "Plus" always be fully spelled out.

Perhaps the AI model didn't correlate the "Apple II+" in your prompt with the Apple ][ Plus in Apple's literature, and hence didn't constrain its Apple II+ feature list to the actual limitations of an Apple ][ Plus.  Out of curiosity, try that prompt again with "Plus" spelled out and see if it devises a poster with more accurate pictures and features for the Plus model.

From my experience AI will make a mistake to some degree in almost everything, sometimes it can be fixed with photoshop sometimes not due to the low resolution it uses, aka 1k, it needs to improve resolution....1024x1024 is meh...

Khaibitgfx, have you used a resolution-enhancing plugin like Genuine Fractals? They can do some amazing things (but can't restore missing detail, obviously).

It is technically based on L-Systems, not "fractals", but I guess they needed a more memorable name for it.

I haven't used any plug-ins, Genuine Fractals is an upscaler, I have Topaz already, it upscales and is over 20 gigs on the hard drive.

Native creative resolution which is a MUST in any professional circle is currently is unavailable. 

So asking AI...

Yes, AI generation platforms will eventually reach a point where their outputs become indistinguishable from true native high resolution, as machine learning models transition from approximating pixels to intelligently synthesizing true high-definition detail.

The Evolution of AI UpscalingCurrent Tech:

Technologies like DLSS and FSR currently take a lower-resolution base image and use machine learning to intelligently reconstruct and upscale it in real-time.The Goal: The ultimate aim for tools is to generate or upscale lower-resolution assets so flawlessly that the human eye cannot tell it was ever compressed, avoiding the shimmering or plastic smoothing sometimes caused by older models.

Why "Eventually"?

Compute Constraints: Running advanced generative super-resolution is computationally expensive and requires highly specialized microprocessing units (MPUs) to run efficiently.

Model Training: AI image and video models are primarily trained on smaller, fixed latents and then scaled up via latent diffusion or VAEs. The models require more training on massive, natively captured high-resolution datasets before pixel-perfect recreation becomes the baseline.

That's all folks...

I wonder what Arthur C. Clarke would think if he was here today...