It often seems like there is a general consensus that human content is “real” content, with AI-generated content being some sort of derivative “copy” of what humans do.
Despite the rapid transformation of simple GAN network models into lively generative AI, we still want to see humans as the best creators. But that may not be the case for long, and that consensus may, in some ways, crumble.
I took a look at this recent X post by Aidan McLaughlin of Mediaite fame, who indicated that he would rather have computers designing the buildings.
“The proliferation of AI’s fun, possibly impossible, surprisingly tasteful and deeply soothing architecture is so cool to me,” McLaughlin writes. “HE MUST comfort beauty. I think there is a happy chance that such things will raise the standards of my generation, and my children will grow up with less human tendencies.”
Human inclination?
This is in contrast to so many comments that talk about how LLM products are based only on the human content they reflect and how there should be human content at the core of training systems. Many analysts have suggested that without the original human content, AI becomes a scavenger of its own products, a recursive repeater that ultimately provides a meager supply of bloodless stuff.
In terms of architecture, the examples in McLaughlin’s post look good – a warm set of living spaces, with warm colors that offer a respite from a snowy landscape – and everything looks very well put together, but are the results of HE “better?”
Architecture, of course, is one of those fields where quality is pretty subjective, and taste (along with utility) is the primary guide, but there’s also McLaughlin’s characterization of AI designs as “probably impossible,” meaning that they have not been verified. fully according to the laws of physics and that an AI architect is prone to some hallucinations while designing. As for “deep relaxation,” many people find AI-generated products attractive, seductive, calm, or otherwise attractive, but just as many are nervous when looking at these kinds of images, at least in part because they have concerns about how genAI fits in our world.
But then there’s McLaughlin’s suggestion that AI designs are more beautiful than handwritten ones. To be sure, many of us talk about AI’s visuals as too perfect, or “pretty,” as a sort of hindrance, but what if that ends up being what we prefer?
What’s in a game?
For a good example of AI leading the way in design, let’s look at the digital spaces where people play.
GenAI has become so good at creating its own game content that we’ve developed a term for it: procedural content creation – where game environments are generated by algorithms, rather than manually created by humans.
The result is endless worlds – in the games Minecraft, Roblox or any other kind of digital construct. The Metaverse, it seems, will be designed primarily by the creators of artificial intelligence. Will they have names?
The groundwork has already been laid: the rise of no-code on platforms like Roblox dates back several years.
“Earlier this year, Roblox said it was testing a tool that could speed up the process of building and modifying in-game objects by having AI write code,” Marcus Law reported in AI Magazine in 2023. “The tool allows anyone Roblox players create objects such as buildings, terrain, and avatars; change the appearance and behavior of those things; and give them new interactive properties by typing what they want to achieve in natural language, instead of code complex.”
These ideas are being implemented in bold new ways.
“PCG offers technical artists, designers and programmers the ability to build fast, iterative tools and content of any complexity, from Asset services, such as buildings or biome generation, to entire worlds,” writes a spokesperson at Unreal Engine. describing those ancillary processes.
How long will it be before HE sweeps the man in the noose?
Tools, Architecture and Architects
In a Google search on architect job displacement, a number of articles promise that “AI will not replace architects” although many of these writers admit that lower-level tasks will be automated, reducing the demand for labor. Then there are the actual architects, reporting from the trenches, for example, that they are being asked to use something like Midjourney for rendering. Check out the tool for yourself and what it can do to turn words into visuals.
It’s not hard to find testimonials from architects talking about how they use AI. A Reddit thread gives special insight into how the tools work as back-up support for human developers.
All of this illustrates how technology is quickly revolutionizing this part of the business world.
The Hardware Revolution
As companies prepare to take advantage of these new capabilities, the hardware industry is turning at an unsustainable pace…
I’ve written about the rise of Nvidia and other chipmakers, and the realities of TSMC’s supply chain, and how countries like the US and China are scrambling for geopolitical positioning in 2025.
Simply put, the skills themselves will create huge demands. Whether AI actually designs “better” than humans is a separate question – but the world will observe a change that is unprecedented in all of human history – for the first time, there are entities, albeit digital ones, that are in way smarter than us. As humans, we were smart enough to remove ourselves almost entirely from the food chain. AI subjects are even further removed from those biological processes that involve mortality. If they can live forever, how smart can they get?
I’ll leave you with that question as we all continue to react to these kinds of blazing breakthroughs in the new year.