Tuesday, 12 September 2023

Idea prompting

Ideas seem to appear anytime and everywhere, slowly emerging or popping up seemingly from nowhere. From a blank piece of paper to the proverbial "Aha!" moment. But what human ingenuity is needed to generate ideas in the age of artifical intelligence, AI? When prompt engineering for a text-to-text model, or a text-to-image model can help produce the desired idea (output). Idea prompting, then, is becoming increasingly a useful ideation tool for designers because large language models, LLMs, have the ability to understand prompts, also called "in-context learning". Prompting skills, which involves specifying or adjusting the AI prompt, moreover, help designers better understand the capabilities and limitations of LLMs. That is, knowing how to write concise, clear prompts will ensure that any AI prompt will produce a strong response.The iterative process of prompting, then, enables designers to generate, revise and refine ideas producing in-depth creative content.

Wednesday, 30 August 2023

AI-powered ideation tools

Browser-based idea and sketch generators, or AI-powered ideation tools are increasingly getting apt at producing imaginative and useful creative output. For example, from a few rough sketch lines, AI technology can quickly analyse the simple sketch drawing and algorithmically generate unlimited sketch ideas and images. Similarly, AI can transform a photo into sketch, or even create sketch drawing form text descriptions (prompts). Indeed, OpenAI's DALL-E 2 text-to-art platform allows users to generated images just about anything, and both realistic and creative images. AI, then, has become a powerful ideation tool at the very beginning of the design process, and very fast at that.

Tuesday, 8 August 2023

Thumbnail sketching

Thumbnail sketching is part of the designer repertoire and refers to a series of small drawings used to quickly explore a variety of design ideas which can be worked out either in physical or electronic format. In the digital realm, and particularly on social media platforms, a thumbnail sketch is about a small, clickable image representing a larger piece of content. This can be a video (as seen on YouTube), an article, a webpage, or a product. Thumbnails are the first thing users see when they encounter online content. And so a well-designed, appealing thumbnail can attract viewer attention, give a preview of the content, and significantly improve the engagement rate directly impacting the click-through rate of the content. Moreover, YouTube’s post-pandemic boom has helped create a microeconomy for thumbnail designers. Indeed for YouTubers, thumbnails are serious business, as they can make or break a video's reach. But designers are seeing the rise of text-to-image tools such as Midjourney and AlphaCTR with a mix of anxiety and curiosity. Curiosity because these tools can produce thumbnails within minutes, rather than hours, as with conventional software, and anxiety because these tools feed on other people’s work raising concerns about copyright infringement. Yet, in technical terms, what thumbnail designers are doing when using text-to-image generation AI tools is simply to incorporate AI into their workflow to create customised visuals. To compete with AI generated thumbnails, human thumbnail designers have to be very good, particularly as thumbnail sketching has wide application in the creative industries, a field where generative AI is evolving fast. Source: https://restofworld.org/2023/youtube-thumbnail-ai/?utm_source=pocket-newtab-global-en-GB

Monday, 10 July 2023

Working it out

Recognised chiefly for his modern designs of steel and glass, and for smooth exterior contours and open, naturally-lit interiors, Norman Foster (b. 1935), a British architect, says, 'travel and the lessons from studying buildings and cities are as important to me now as they were when I was at architecture school. In that sense I am still a student.' And so, among his mentors are Buckminster Fuller, Paul Rudolph, Serge Chermayeff and Vincent Scully as well as the Bauhaus and Ulm schools of design. A motivating force in Foster's design practice is his eagerness to know and to learn: 'I was always curious about how something works, in the case of the windmill the inner workings, the cogs, how the wind power was transmitted eventually to the stones that would grind the grain. I think that that comes round to how do you find a basis for design. It’s not just how the materials work but – whether it’s a museum or a corporate entity – what are the values, what makes it tick below the surface. How can that be reflected in the architecture? How can the architecture enable it to work better?' He continues: 'And to give insights into the creative processes beneath the surface – to show that something that looks very simple has come out of studying many options.' While there is accumulated extensive knowledge of famous architects' advice for young designers, Foster's insightful and thoughful words are bound to be both inspirational and useful to a wide range of creatives. Source: Interview in The Guardian July 9, 2023.

Tuesday, 23 May 2023

Flash workshop

The notion that ideas appear in "a flash" may also suggest "flash ideation workshop". That is, the workshop works on the premise of improvisation or ad hoc, or has a pop-up or intermittent quality to it which situates ideation not just in the studio. Using social media, and artificial intelligence, in addition to analogue or traditional ideation tools, may also encourage finding a balance between physical and remote work. Flash workshopping, then, and whether doing so virtually or in person may not result only in new ideas and innovative insights but also stimulate and support collaboration to solve a specific problem. In both scenarios, the workshop seeks to generate ideas for actionable change based on a given task.

Tuesday, 25 April 2023

GPT for ideation

GPT-4 is is a multimodal large language model created by OpenAI and the fourth in its GPT series. It was relased in March this year only a few months after the previous version reflecting the speed at which AI technology is developing. GPT-4 can accept a prompt of text and images, which—parallel to the text-only setting—lets the user specify any vision or language task. Like previous GPT models, the GPT-4 base model was trained to predict the next word in a document, and was trained using publicly available data (such as internet data) as well as licensed data. Although less capable than humans in many real-world scenarios, the model exhibits human-level performance on various professional and academic benchmarks, for example, simulating exams that were originally designed for humans resulted in a 90 per cent score in law exams, and art history. Despite its capabilities, however, GPT-4 has similar limitations as earlier GPT models. Most importantly, it still is not fully reliable, for example, it “hallucinates” facts and makes reasoning errors. Also, the model can have various biases in its outputs. So when prompted with a question, the base model can respond in a wide variety of ways that might be far from a user’s intent. Such responses may, or may not trigger new ideas. The language model, however, points to creative futures when AI will help turn designing into a seamless process from idea to realisation, from first thought to production and manufacturing. Indeed, the design process, including ideation is already largely driven by language (verbal and visual).

Monday, 20 March 2023

Designers on autopilot?

As AI technology is trained on more diverse and larger datasets including images as well as text the possibilities of what can be created seem endless. OpenAI's ChatGPT-4, for example, has 1 trillion parameters, or variables that the model learns from data. Yet, so far idea generation through AI has limitations because AI only does what it is told (data input). As a data driven system, then, AI cannot fully replace real-life experiences or personal interaction associated with idea generation such as improvisation, reflecting and daydreaming. However, generative software frees up time to designers providing them with a tool that afford novel ways of ideating. This suggests that the relationship between designers and AI is collaborative and that the most successful ideas won’t necessarily come from bright thinkers alone but from those best at using AI to assist ideation. A relationship perhaps similar to that between aircraft pilots and autopilot tech. That is, autopilots assist aviators along the flight path but do not replace the pilots. Moreover, pilots love to fly and they will fly. And designers love to design ... 

Blog Archive