Welcome to this week’s edition of Architecture Insights.

Adobe Creative Suite has been long involved in keeping up with the changes. Today we will explore a few prominent features of the Photoshop AI tool ‘generative fill’ for architectural renderings.

As always, here is this week’s latest news in the AI world that impacts architects and designers.

News & Updates

1. Google releases Gemini, a remake of their Chatbot and AI assistant, Bard.

Summary: Google's previous GPT, named Bard, has been renamed Gemini. It comes with a free upgrade to their new edition named ‘Pro 1.0’, and a paid version named ‘Ultra 1.0’ for $20USD a month. Currently, Gemini is the largest competitor of ChatGPT. Both the free and paid version are available to all users.

2. DALL-E 3 will start adding watermarks to images.

Summary: Starting on the 12th, images will have a visible watermark in the top left corner. This helps users verify the origin of images through specific verification websites. This also represents a significant move towards promoting transparency and authenticity in digital content to combat misinformation.

3. EvolveLAB introduces ‘Glyph Copilot’, an AI GPT assistant for the Revit plug-in.

Summary: This upgrade will use GPT-4 to automate tasks like drawing dimensions, tagging, and view creation. Below is an example of generating dimensions in seconds with the use of a personalized AI project chatbot.

A full blog post by EvolveLAB breaking down each new feature is also available.

Photoshop AI (Generative Fill)

We want to showcase three prominent features of Photoshop’s generative fill that are useful for editing and upgrading site images and perspectives.

A few things to note; ensure your Photoshop is updated to the latest version or at least version 24.6 to use the generative fill tool. 

You may also need to navigate to the 'Beta apps' in your creative cloud apps manager and install 'Photoshop (Beta)', this is required if you have not done so already.

The AI features in Photoshop will be released first through the Beta version.

You can also access a limited version of Photoshop with AI features on the web at Adobe Photoshop.

Extending the image canvas

By using the marquee tool → then selecting the image outline → followed by the select inverse feature → then hitting generate (with or without a prompt). Photoshop will do the rest to expand your image.

Below we took the first image and extended it to show more of the surrounding area.

Original Image:

Expanding Image:

Removing parts of an image

Use the lasso or marquee tool to highlight the area you want to remove. In the generative fill toolbar select ‘Contract Selection’, leave the prompt box empty and hit generate.

Image variations

Here we can change and explore variations of our main subject, in this case, the building.

Select the area you want to change and in the prompt box type your desired outcome. In our case, we highlighted the central building and in our prompt box wrote: Bright lights and parametric design.

The result:

Now we want to add a phone booth to the image. Use the lasso or rectangular marquee tool to outline your area and describe the object you want to add in the prompt box.

Image w/ phone booth:

Photoshop will create 3 different variations which you can toggle from on the toolbar or properties manager. It will also create a layer of any new objects giving you the choice to manipulate it manually..

AI Image of the week

Thank you for reading this week’s issue, check past issues here. Share this newsletter with colleagues, friends, or anyone interested in the combined world of architecture and artificial intelligence.

Feedback is always appreciated, let us know how we did by responding to this email.

Until next Friday,

A.I.