EricGT February 24, 2023, 9:40am 1
ChatGPT is a generative text transformer but sometimes an image is worth a thousands words. While ChatGPT can not generate an image it can generate text for a common image represtation, e.g. SVG, that can be converted into an image using an external application.
Below are some examples of graphical images that can be transformed to text for saving or transforming text into an image, expand list to different technical fields and include the field associated with the type of image, such fields and computers, electronics, biology, engineering, business, music, etc., also include one or more applications that can convert the text to an image.
- Entity Relationships
- Swim Lanes
Here are some additional examples of graphical images that can be transformed into text or vice versa, along with their associated technical fields and applications for conversion:
- Circuit Diagrams – Electronics – Applications: Circuit Diagram Maker, CircuitMaker, KiCad, Fritzing
- Chemical Structures – Chemistry – Applications: ChemDraw, MarvinSketch, PubChem Sketcher, Avogadro
- Flowcharts – Computer Science – Applications: Microsoft Visio, Lucidchart, Gliffy, Draw.io
- Mind Maps – Business and Education – Applications: MindMeister, Coggle, XMind, MindMup
- DNA Sequences – Biology – Applications: ApE (A plasmid Editor), Geneious Prime, SnapGene, Benchling
- Sheet Music – Music – Applications: MuseScore, Noteflight, Sibelius, Finale
- UML Diagrams – Software Engineering – Applications: StarUML, Visual Paradigm, Creately, PlantUML
Some applications that can convert text to an image include:
- Graphviz – a graph visualization software that can convert DOT text files to graphical images
- LaTeX – a typesetting system that can convert LaTeX code to PDF, which can include text and mathematical expressions as graphical elements
- ImageMagick – a software suite that can convert ASCII art to images, as well as perform other image manipulation tasks
- PlantUML – a tool that can generate UML diagrams from textual descriptions using a simple syntax.
Yes, but the problem is that in most cases these generated drawings are not accurate because ChatGPT is not an expert system AI nor is ChatGTP a computing model.
ChatGPT is a generative AI and it simply “makes things up” as an auto-completion engine, predicting text one sequence at a time. It has a very low accuracy rate for anything beyond very basic technical completions. ChatGPT is good at fiction and story telling where technically accurate does not matter.
ChatGPT is not a computing model and nor it it an expert system AI. It’s a fancy (powerful) text-autocompletion engine and its technical completions are generally not very accurate and often “completely wrong”.
It’s important to add this disclaimer because many people will take posts like this and start generating drawings or diagrams or formulas which they will assume are technically accurate. Every detail in a ChatGPT technical completion must be completely checked and verified.
Hope this helps.
EricGT February 26, 2023, 7:05am 3
I don’t plan to post disclaimers with every post I create, OpenAI clearly noted such here.
From: ChatGPT emails, links and other actions | OpenAI Help Center
You should always verify any information or claims that ChatGPT makes with other sources, and do not rely on it for any critical or sensitive decisions or actions. ChatGPT is not a substitute for human judgment, expertise, or responsibility.
That link you posted is not really useful as a disclaimer and most users here do not even read any OpenAI documentation, including the API docs.
Never mind @EricGT, you do not have to post any disclaimers regarding generative AI, as I will keep posting disclaimers because most new users here have zero idea about how very inaccurate technical completions are with generative AI. Even users who have been around for a while tend to forget about how generative AI “just makes things up”.
Yes, generative AI can generate “just about anything” because it is a based on a massive LLM, but regardless of prompt engineering, most beyond trivial technical completions are not accurate.
At best, technical completions make an OK “rough draft”; and at worse, generative AI technical completions sound convincing but consist of a lot of nonsense in the guise of technical facts by an expert.