Walk into almost any corporate boardroom today and you will meet someone who thinks generative AI has made graphic designers obsolete. Because a chatbot can spit out a colorful picture or a ten-slide presentation in thirty seconds, people with zero training or field experience in art or visual communication assume the hard work of design is over.
I understand the temptation or delusion, yet nothing could be further from the truth.
Before I ran technology and communications firms, I trained at a fine art college. Over the decades, I worked in the trenches as an art director, creative director, and print and multimedia production manager. I was also an early Apple Mac desktop publishing (DTP) pioneer and evangelist back when digital layout tools first transformed print shops and design studios. When you have spent decades worrying about ink traps, vector curves, typography kerning, and bezier curves, you look at current AI visual tools and immediately see the cracks beneath the glossy veneer.
Generative AI is terrific at words, logic, and structure. When it comes to actual graphic design and presentation decks, however, it falls flat on its face.
The Illusion of AI “Art” and “Design”
The biggest mistake non-designers make is confusing a pretty picture with functional design. Commercial design is not about generating a neat image once; it is about building flexible, accurate visual assets that survive real-world revisions.
Most standalone AI image generators output locked pixel data. They hand you a flat bitmap made of a fixed grid of pixels. Unlike true vector graphics constructed from mathematical paths, flat raster images fall apart the moment you try to scale them for a large trade show banner or a crisp print brochure. Even worse, you cannot click into a flat AI image to select, move, or recolor an individual shape. If a client asks you to shift a logo slightly to the left or change the shade of blue on a single icon, you are stuck.
Some people point out that dedicated vector illustration software now includes built-in generative features, yet that point is completely moot. Working professionals already use those professional tools and know how to manipulate anchor points, paths, and layers properly. Amateurs neither own those industry-standard applications nor know how to use them, leaving them trapped with uneditable bitmaps from consumer prompts.
In professional production, non-destructive editing is everything. Fixing a tiny flaw in a flat AI graphic often takes longer than building the entire illustration from scratch in Adobe Illustrator or Figma.
Furthermore, generative tools lack design system logic. Real brand work relies on master templates, linked color palettes, and strict typographic rules. AI treats every prompt as an isolated visual guess rather than part of a cohesive, rule-based document. It produces discrete visual elements with no connected context.
Why AI Slide Generators Disappoint
That exact same flaw ruins AI-generated presentation decks. Many executives paste a rough prompt into an AI slide maker, hoping for a pitch-ready deck, and end up with a formatting nightmare. I know. I have tried.
Presentation design requires spatial geometry, visual contrast, and breathing room. Text-based AI models cannot actually see a slide canvas. Because they lack genuine spatial awareness, they cannot judge how much physical room a sentence occupies. The result is predictable: awkward line breaks, text overflowing out of boxes, or cramped walls of tiny type.
Worse still, AI slide builders trap your message inside fragmented text containers. Instead of building responsive, grouped layouts, they scatter disconnected text lines across the page. If you edit one headline or adjust your page margins, you have to manually realign every other box on the slide.
Because the AI generates copy for isolated boxes one by one, it frequently loses the broader thread of your presentation. You end up with repetitive phrasing across slides and generic templates that slap a title box over three bullet boxes regardless of what your message actually requires.
Where AI Actually Shines: Textual Ideation and Outlining
Does this mean you should throw generative AI out of your creative workflow? Not at all. You simply need to use it for what it was built to do: process language.
Large language models predict the next word based on continuous text flow. Because they train heavily on structured essays, books, and reports, they naturally grasp hierarchy, argument structure, and narrative pacing. When you ask AI to brainstorm angles or draft a slide-by-slide outline, it maintains a clear logical thread from start to finish. Point B actually follows Point A.
Editing a plain text outline takes seconds, whereas fixing broken typography and misaligned containers across fifteen AI-generated slides wastes hours.
A Workflow That Respects Both Craft and Speed
If you want presentations and marketing assets that command respect, stop asking AI to be your art director and start treating it like a sharp junior copywriter.
First, use AI to build your structural outline. Ask it to map out your presentation slide by slide, including a clear action title, two or three supporting points, and brief speaker notes for each section.
Second, edit the narrative yourself. Read the outline from top to bottom, cut the fluff, and sharpen the voice so it sounds like a human leader speaking directly to an audience.
Third, take that approved text into proper design software. Build your visuals using real vector assets, clean grids, and deliberate white space tailored to the exact message on each slide.
Technology should enhance human craftsmanship, not replace it with clumsy shortcuts. Let AI do the heavy lifting of textual scaffolding, and let people who actually understand design do the visual storytelling.
Need a branding, strategy, or communications expert? Talk to us today!
###
Dr Seamus Phan – Global C-suite Publicist & Strategist (Biochemist, Cybersecurity & Webdev pioneer, Author, Journalist) with nearly 40 years of professional field experience. Some articles are reproduced at McGallen & Bolden, where he is CTO and Head of Content. Visit him on LinkedIn.