Re:ChatGPT's image generator can be manipulated to produce v...
by anuramat on Thursday June 18, 2026 @12:01AM
(#48580664)
yes, assuming bugs don't exist
Re:ChatGPT's image generator can be manipulated to produce v...
by solid_fuel on Thursday June 18, 2026 @12:08AM
(#48580708)
Wow, you're halfway there. Yes, when user data gets loaded into an executable code page - which are reserved for command data - it is a bug.
That is why LLMs - which intentionally mix user data and command data into the same space - ARE BROKEN BY DESIGN. Do you get it now? It is a bug, and it is a bug which is fundamental to the design of LLMs. There is no way to build one that does not do this.
Re:ChatGPT's image generator can be manipulated to produce v...
by anuramat on Thursday June 18, 2026 @12:31AM
(#48580839)
are all storage devices broken by design as well?
Re:ChatGPT's image generator can be manipulated to produce v...
by solid_fuel on Thursday June 18, 2026 @12:37AM
(#48580877)
Are you somehow under the impression that storage devices and LLMs fill the same purpose? That's a major misunderstanding. Here's a good starting point if you're struggling with the difference between a computation device and a storage device: https://en.wikipedia.org/wiki/Computer
The CPU physically will not run instructions which are in areas of memory which are not marked as executable. This is a foundational principal of computing security.
> In computer security, executable-space protection marks memory regions as non-executable, such that an attempt to execute machine code in these regions will cause an exception. It relies on hardware features such as the NX bit (no-execute bit), or on software emulation when hardware support is unavailable. Software emulation often introduces a performance cost, or overhead (extra processing time or resources), while hardware-based NX bit implementations have no measurable performance impact.
https://en.wikipedia.org/wiki/Executable-space_protection