AI in Design

Exploring the current capabilities of AI as a design tool.

Using AI to rapidly prototype web pages in interaction design and user testing. This approach yielded operational web pages laid out in a manner consistent with good UX design and included html, css and java script which was capable of generating charts dynamically using D3.js.
AI as persona development and silicon sampling. In this simple example the model was able to generate convincing personas with a strong backstory and also able to provide financial information. The model was then able to generate a dialogue between the persona and a designer similar to a cognitive walk-through. This simple example also features running a simulated card sorting activity with a sample population.
AI image generation has advanced significantly and is now at a point where very high quality images can be generated through a number of different models. This example looks at using AI image generation to inform mood boards etc . This work was carried out using the free version of mid journey and the possibilities of integrating this usage of AI into a design workflow are obvious and powerful

Many 3-D modelling packages have a script based in put option available, commonly using Python as the script. The attached example shows the modelling of a simple chair using a dedicated script best 3-D modeller.( AI in 3D modelling with OpenSCAD) . It is important to note here that the AI known what a chair looks like and put in the appropriate dimensions.

A similar process can be conducted using Rhino/Grasshopper or Blender with Python script. Here I asked to model 64 spheres in a cube arrangement and apply a subdivision surface modifier.

Use of GPT-4 to summarise work [[AI prototyping of user interfaces]] and [[GPT-4 summary of key texts on Data Visualisation and Interaction Design]]
Combine with Zotero to make reading notes quickly that link directly to the article [[Zotero to Obsidian]] which needs to be written up as a workflow [[Obsidian Zotero literature review workflow]]

Concept development and core idea generation as a higher value skill
Design appreciation as a higher value skill than design production
Developing a deeper understanding of the user need and the value of the human perspective in design will become increasingly important
Extension of designers skills into other areas such as code development made possible by the power of the tool making the designer more technically capable and proceeding where previously the prototype may have been delayed by the need for specialist input
Research skills will likely see an increase in the focus on primary research as secondary research can be highly automated. It is quite likely that research instruments such as surveys and interviews can be tested prior to going to the field.

AI impact on design assessment

Show your workings. It is going to be critical now that the design student show that path to the creation of the design. These should be contemporaneous notes the show it hasn’t been a work back from an AI generated final design
Clarity Clarity on where the AI can and can’t be used in the generation of designs. Denote which AI was used, and the prompts used to generate the response
Authentic assessment . We have a history of ensuring authentic assessment and it is clear that the use of AI as an accelerator in the design industry is already a fact of life. We should therefore enable our undergraduates to work in this changed landscape and the assessment should be authentic to this changed reality
Increased expectations. With routine access to this technology our expectations that these tools should be used in writing copy and therefore tolerance for poor spelling, grammar and construction should be lower. We currently expect our learners to use spell check and words in built grammar checker to inform their writing
Plagiarism. Does the use of AI generated text constitute plagiarism. Can the AI be used in proofing or drafting stage and for the work to be considered the learners work.