AI experimentation
AI is rapidly becoming a powerful co-pilot in the design process, particularly in accelerating the leap from concept to prototype. I’ve found significant value in using tools like Figma Make to breathe life into flat designs, alongside AI's ability to generate prototypes populated with realistic, personalised user data. It also serves as an advanced editorial tool - acting as a sophisticated "find and replace" to ensure content is tailored precisely to the different points of view and needs of various user groups.
However, the technology still requires a steady hand to remain effective. While AI can draft styleguides or design systems, it often veers off-brand, requiring repetitive prompting to avoid certain treatments or maintain visual consistency. For me, the current strength of AI lies in automating the heavy lifting of data and basic structure, allowing more time to focus on the nuanced craft and strategic flow that AI cannot yet replicate.
Figma Make
Ive explored how Figma Make can be utilised to recreate prototypes and making them feel ‘live’.
By feeding in flat designs from my work on the Economist Espresso App, I was able to prompt it to populate current news from The Economist website and summarise the content to fit as daily short form articles but still read as if written in the same journalistic style. The new quiz feature was also contextual based on todays news articles. Using AI also brought the reading habits/leaderboard stats page to life with more realistic content.
Claude Figma MCP
Ive used the Claude plugin to help co design new features for my work on Adam Bray Dermatology website. I was able to connect Claude to my Figma design file and co-design a self diagnosis file upload feature to check moles.
Using the components and variables I already set up, Claude was able to interpret the designs and make 3 variants of the feature showing depending on the mole type.
GitHub Copilot in VS Code
Working in the public sector using the .GOV Prototype Kit, Ive experimented with using copilot to help me tidy up messy code, find label inconsistencies and help with dynamic routing.
I found the tool to be very advanced and designers would equally need to have a very good technical understanding of coding languages in order to effectively utilise the copilot.
AI experimentation
AI is rapidly becoming a powerful co-pilot in the design process, particularly in accelerating the leap from concept to prototype. I’ve found significant value in using tools like Figma Make to breathe life into flat designs, alongside AI's ability to generate prototypes populated with realistic, personalised user data. It also serves as an advanced editorial tool - acting as a sophisticated "find and replace" to ensure content is tailored precisely to the different points of view and needs of various user groups.
However, the technology still requires a steady hand to remain effective. While AI can draft styleguides or design systems, it often veers off-brand, requiring repetitive prompting to avoid certain treatments or maintain visual consistency. For me, the current strength of AI lies in automating the heavy lifting of data and basic structure, allowing more time to focus on the nuanced craft and strategic flow that AI cannot yet replicate.
Figma Make
Ive explored how Figma Make can be utilised to recreate prototypes and making them feel ‘live’.
By feeding in flat designs from my work on the Economist Espresso App, I was able to prompt it to populate current news from The Economist website and summarise the content to fit as daily short form articles but still read as if written in the same journalistic style. The new quiz feature was also contextual based on todays news articles. Using AI also brought the reading habits/leaderboard stats page to life with more realistic content.
Claude Figma MCP
Ive used the Claude plugin to help co design new features for my work on Adam Bray Dermatology website. I was able to connect Claude to my Figma design file and co-design a self diagnosis file upload feature to check moles.
Using the components and variables I already set up, Claude was able to interpret the designs and make 3 variants of the feature showing depending on the mole type.
GitHub Copilot in VS Code
Working in the public sector using the .GOV Prototype Kit, Ive experimented with using copilot to help me tidy up messy code, find label inconsistencies and help with dynamic routing.
I found the tool to be very advanced and designers would equally need to have a very good technical understanding of coding languages in order to effectively utilise the copilot.
AI experimentation
AI is rapidly becoming a powerful co-pilot in the design process, particularly in accelerating the leap from concept to prototype. I’ve found significant value in using tools like Figma Make to breathe life into flat designs, alongside AI's ability to generate prototypes populated with realistic, personalised user data. It also serves as an advanced editorial tool - acting as a sophisticated "find and replace" to ensure content is tailored precisely to the different points of view and needs of various user groups.
However, the technology still requires a steady hand to remain effective. While AI can draft styleguides or design systems, it often veers off-brand, requiring repetitive prompting to avoid certain treatments or maintain visual consistency. For me, the current strength of AI lies in automating the heavy lifting of data and basic structure, allowing more time to focus on the nuanced craft and strategic flow that AI cannot yet replicate.
Figma Make
Ive explored how Figma Make can be utilised to recreate prototypes and making them feel ‘live’.
By feeding in flat designs from my work on the Economist Espresso App, I was able to prompt it to populate current news from The Economist website and summarise the content to fit as daily short form articles but still read as if written in the same journalistic style. The new quiz feature was also contextual based on todays news articles. Using AI also brought the reading habits/leaderboard stats page to life with more realistic content.
Claude Figma MCP
Ive used the Claude plugin to help co design new features for my work on Adam Bray Dermatology website. I was able to connect Claude to my Figma design file and co-design a self diagnosis file upload feature to check moles.
Using the components and variables I already set up, Claude was able to interpret the designs and make 3 variants of the feature showing depending on the mole type.
GitHub Copilot in VS Code
Working in the public sector using the .GOV Prototype Kit, Ive experimented with using copilot to help me tidy up messy code, find label inconsistencies and help with dynamic routing.
I found the tool to be very advanced and designers would equally need to have a very good technical understanding of coding languages in order to effectively utilise the copilot.