top of page
Search
Writer's pictureLuke Hurdman

Are People (Artist)-made renders better than AI or text-to-image?

Ai is impressive, and we have been using it in various capacities for a while. Grammarly, Zapier etc.

The intent here is only to start the conversation, not to deter or reduce validity or anyone's passions, but AI does raise a few questions and things to consider from a marketing and talent perspective. Spoiler: There might be a slight bias here.

  • Is it easier to talk with someone locally or learn how to prompt Ai?

  • Does original curated content communicate your project's specific design and intent?

  • Does your AI-generated image look like other images on Instagram?


Firstly, it is essential to understand that most generative AI programs, in a way, are digitally "recycling" existing work available online. In short, AI composes parts from other imagery so anything generated is not entirely an original image. Think of it like a re-mixed rap song. Let's admit it, some AI-generated songs sound very similar. As a precaution, in time, renders could look homogenised and possibly pasteurised.

David Holz's - Midjourney, OpenAi's - Dall-E 2, Gencraft, Dream Studios - Stable Diffusion, Photoshops - Firefly, Alex Mordvintsev's - DeepDream and Shencai's - Promeai, to name a few, have rapidly made inroads into the Architectural & visulisation industry. They all make amazing images, however... They all require time.

The advantage of working with a Visualisation Studio is we are all human and we are all in this together and the human cognitive function is more efficient and easier to communicate with. The human experience comes with local knowledge of standards, architectural and planning rules. As well as the evolving nature of a project. Yes, impressive images take time, requiring considered thought and artistic merit.

A good prompt generates good renders, and an incredible prompt generates incredible renders. Drawing from Tim Fu's (Zaha Hadid Architects) brilliant work and insights and expanding upon - using Ai, the most effort will be put into rationalising the image idea into a unique project context, addressing site, scale, lighting, warmth, mood, weather, contrast, depth of field, and technical camera specifications. The artist has learnt this for you.

This is built on top of a deep knowledge of Art, visual perceptions, colour theory, cultural nuances, local, federal and council laws, and accepted local rules. This isn't easy to prompt with Ai but it will be easier.

We are sure this view will change, and our industry will evolve again. We are looking forward to the evolution and will revisit this again.


Thank you,

Regards,

Luke



19 views0 comments

Recent Posts

See All

How much should renders cost?

The cost of renders isn't an easy answer as it depends entirely on the individual project and sales objectives. However, there are a few...

Commenti


bottom of page