D656 Leveraging AI and Technology in Design, catalog number DES 3500, is the three CU course in the WGU School of Business design sequence covering the use of artificial intelligence and other emerging technology within the design process. The interesting question in this course is not what the tools can generate. It is which parts of a design process can be accelerated without removing the step that made the outcome good.
Which step is being replaced
A design process has distinct stages, and they have very different exposure to automation. Research synthesis can be accelerated but not delegated, because the judgment about what matters is the work. Ideation can genuinely be widened, since volume and variation are exactly what these tools produce cheaply. Visual production is the most automatable stage of all. Evaluation cannot be automated in any meaningful sense, because it depends on watching a real person fail.
DES 3500 rubrics tend to reward submissions organised around that analysis rather than around tool capability. The failure mode is a document that celebrates the speed of generation without noticing that speed applied to the wrong stage produces a great deal of well rendered work aimed at a problem nobody checked.
The second scored theme is what gets lost. Generated concepts converge, because they are drawn from what already exists, and a design process fed entirely by them tends towards the average of its training material. Personas invented rather than researched are plausible and empty. A submission that names these costs specifically reads as considered, while one that lists only benefits reads as promotional.
Third, provenance and rights. Assets, likenesses, styles and text all carry questions about origin and permission, and a design course is the right place to take them seriously rather than assume them away.
Turning scored aspects into a section plan
Scoring detail lives in your Course of Study rather than the catalog. Count the aspects and use them as headings in the rubric's own language. Each is judged independently against a three point scale and each needs a 2, so an impressive set of generated artefacts cannot carry a thin ethics section.
Where you used these tools to produce part of the submission, document it exactly as your task directions require, and treat that documentation as part of the deliverable rather than as a disclaimer.
The word budget, worked. Assume six scored aspects and around 1,700 words. Reserve 130 words to name the design problem and the process stages under discussion, and 110 to close. That leaves near 1,460, about 240 per aspect. Then take 50 words from each of two capability aspects and give 100 to the aspect covering limitations and 100 to the aspect covering rights and disclosure. Those two cannot be filled with output samples and are where thin work becomes visible.
A structure that fits this deliverable
Where your directions specify a structure, follow it. Where the shape is open, this ordering supports the reasoning the aspects look for.
| Section | What belongs in it | How it gets read |
|---|---|---|
| Process baseline | Your design process stage by stage, with time and effort attached | Nothing can be shown to have improved without this |
| Stage analysis | Where automation helps, where it assists, where it must not be used | The core reasoning; stage level beats tool level every time |
| Applied example | One stage worked both ways, with the outputs compared | Direct evidence beats description of capability |
| Quality control | How generated material is checked before it influences a decision | Generated research findings are the highest risk item |
| Limitations observed | Convergence, plausible but empty output, and anything you caught | Specific observed failures are the most convincing material available |
| Rights and disclosure | Asset provenance, licence terms, and what is disclosed to whom | Expected in a design context; its absence is noticed |
| Recommendation | Where you would and would not adopt this, with reasons | Turns the analysis into a position |
Run one stage both ways and show the results side by side. A single honest comparison does more for this submission than any amount of description, and it supplies evidence for several aspects at once.
Evidence craft on a fast moving subject
Sourcing is hard here because most available writing is either promotional or reflexively dismissive.
- Date every capability claim, since what a system could do at a point in time is a fact and a general claim is not.
- Prefer independent evaluation to vendor material, and label vendor claims clearly when you use them.
- Show your own outputs, including the poor ones, rather than describing what the tools produce.
- Record the prompts or inputs you used, so a reader can judge whether the output reflects the tool or the instruction.
- Check licence terms for anything you generate or incorporate, and state what they permit.
- Use APA throughout with access dates, and disclose your own use exactly as the directions require.
Report a case where the generated output was confidently wrong or subtly off, and say how you caught it. Nothing else demonstrates the critical stance the course is looking for as economically.
What separates Competent from a submission sent back
Aspects score independently, and returns concentrate on limitations and rights.
- The analysis is organised by process stage rather than by product.
- At least one stage is worked both ways with the outputs compared.
- Limitations are observed personally rather than quoted from commentary.
- Rights, licensing and disclosure are addressed concretely.
- The recommendation states where you would not use these tools, not only where you would.
Performance assessment work can be revised and resubmitted with no grade penalty, so a return costs calendar. Terms are six months at a flat rate, which makes a clean first submission the objective in a course that is otherwise one of the lighter ones in the sequence.
Six mistakes that cost time in D656
- Writing a tool review. It dates immediately and answers few aspects.
- Automating evaluation. Nothing substitutes for watching a person struggle with your design.
- Generated personas presented as research. They are plausible, empty, and a serious methodological error in a design course.
- No baseline process. Improvement cannot be claimed without knowing what the stage cost before.
- Ignoring convergence. Generated concepts trend towards the average of existing work, which is a real design cost.
- Silence on rights. Provenance and licensing are part of professional practice, not an optional appendix.
Where acceleration helps and where it hollows out the work
The most useful analysis you can produce in D656 is a stage by stage judgment with reasons, and it holds up regardless of which tools exist when you write it.
Research gathering can be assisted usefully: transcription, tagging, first pass clustering of notes. What cannot be delegated is deciding which observations matter, because that decision is the design insight rather than an administrative step preceding it. A process that automates synthesis produces themes nobody found surprising, which is the signature of research that changed nothing.
Ideation is where automation genuinely earns its place. The method depends on volume and variation, and cheap volume is exactly what these systems supply. The discipline required is to treat generated concepts as raw material for selection rather than as candidates for adoption, and to keep a human generated set alongside for comparison, since the two tend to differ in interesting ways.
Production is straightforwardly accelerated, with the caveats about rights and about house style. This is where most of the honest time saving sits, and saying so plainly is better than claiming transformative gains across the whole process.
Evaluation is the stage to defend. Synthetic feedback on a design, however articulate, is a prediction generated from prior text rather than an observation of a person failing to complete a task. A submission that automates testing has removed the only step in the process that could have told it the concept was wrong, and naming that clearly is one of the strongest arguments available in this course.
How support works on this course
Send the rubric from your Course of Study and the task directions. What comes back is aspect mapped: a process baseline, a stage by stage analysis rather than a tool tour, a worked comparison, quality control and rights sections, and a recommendation with limits. Plus a walkthrough so the position is yours to argue.
D656 sits shortly before the design capstone and the judgment it builds is directly useful there. Terms run six months at a flat rate, so clearing the lighter courses quickly protects the time the capstone will need.
Questions students ask about D656
Is D656 the same course as DES 3500?
Am I allowed to use these tools in my own coursework?
Can you write this submission for me?
Where D656 sits in WGU's programs
The July 2026 catalog places this code in 1 current WGU program. Open a program page for the complete standard path and term positions. The live Degree Plan remains authoritative after transfer credit, substitutions, and mentor planning.
The assessments, one by one
The public catalog does not publish this course's PA/OA identity or task count. WGU Tutors publishes at most one PA manual per course and only from a WGU-controlled public rubric. Until that source exists, PA help begins from the student's real Course of Study and OA support remains preparation only.