Authenticity is probably the wrong word as it’s been severely debased and overused for some time now, especially with respect to those marketable things known as experiences, particularly if they take place in foreign countries. Remember the photos of the long queue of people with their identical bucket lists of authentic experiences waiting to get to the summit of Mount Everest? If you don’t, here’s what it looked like.

Authentic is real as opposed to fake and usually regarded as better. Accordingly, our responses towards something that’s authentic are usually more positive than they are towards the obviously inauthentic. History is always authentic if it’s a chronicle – a full account of what happened in the past – but this never happens. The record is always only partial as it chronicles only who and what was thought important at any given time. It’s subjective to begin with and it remains subject to the agendas and emphases of people in the present interpreting it or possibly even teaching it. Occasionally some future historian or researcher may accurately relink events with actual reasons and consequences but, even then, this may only be for them to be forgotten once again. Now and then in this blog I Identify certain architects as misfit architects in an attempt to stop us dismissing their work too quickly and having them drop out of history. That is, if they ever entered it.

As someone who’s just finished teaching an introductory course on the history of architecture to first year students, I’m guilty of condensing about four thousand years of architectural history into fourteen lectures that I thought covered the essentials and prompted students to see history in terms of solutions to problems at various times and places. Some problems aren’t even tied down to one time or place.

Every century or so, architects are encouraged to learn from history but what needs to be learned is never explicit. The last time around was post-modernism which, in the end, was selective and individual interpretations of bits of history in tune with some marketing agenda. But what was the problem post-modernism was invented to solve? Robert “I-like-complexity-and-contradiction-in-architecture” Venturi isn’t exactly blameless but Charles Jencks made his name promoting the notion that the purpose of post-modernism was to redress an imagined lack of meaning with the visuals of the social housing that went before. Result? Social housing became uncool. Corporations were the new clients.

This was the beginning of representations of something being as good as if not better than the real thing. To architects at least, Seaside, FL was presented as an example of new urbanism and promoted as a new way for towns to be. To everyone else it was timeshare and holiday-let city – a place to pretend you were living in the town you never had or even knew you missed. Is inauthentic nostalgia still nostalgia? It doesn’t matter. The aesthetic was designed to evoke meanings more correctly called fantasies. Not too far away from Seaside, Post-modernism and Disneyworld quickly found each other.

But at least the fantasies were overlaid onto tangible objects called buildings that were in some sense authentic carriers of meaning. The next level of inauthenticity was to dispense with buildings as tangible objects, because fantasies could just as effectively be overlaid onto images of virtual buildings. Online architectural media is clogged with visualizations of proposed buildings that most likely will never be realized. We don’t really care.

This subset of building that were never realized used to consist of the drawings of Sant’Elia, Frank Lloyd Wright’s Mile High Skyscraper proposal, Mies van der Rohe’s Brick Country House proposal, and Vladimir Tatlin’s Monument to the Third International as well as, sometimes, the unbuilt work of Buckminster Fuller and that’s about it. Although unbuilt, these proposals appeared in history books because they were indicative of what we were told were the mood and aspirations of the time. We can’t say this about the surfeit of unbuilt proposals we’re obliged to scroll through these days. We consume architectural imagery in different ways now and architectural imagery has obliged to suit us.

1990–2008 Dubai was responsible for an explosion in the amount of virtual architecture in the online landscape. For architects, two-dimensional representations of buildings are all that’s necessary for the purposes of branding and to show people one’s still alive and open for business. As ever, occasional visionary projects with no hope of ever being realized are always proof of one’s visionary credentials. Mars is in retrograde this year but not too long ago we had Foster+Partners, BIG and Stefano Boeri all showing us their fantastic visions for life on Mars. To mis-paraphrase comedian George Burns, if you can fake authenticity then you’ve got it made. It’s much the same as redefining fake as authentic but we’ll talk about AI in architecture in a bit.

Remember the house Frank Lloyd Wright designed in 1951 for Petra Island in Lake Mahopac (only 15 minutes from Manhattan by helicopter, Dezeen helpfully tells us)? The owners of the island discovered their land came with plans for house designed by Frank Llloyd Wright house and promptly had it built. At first, the Frank Lloyd Wright Foundation refused to acknowledge it as a work of Frank Lloyd Wright. It still doesn’t appear on their website although Monona Terrace Community and Convention Center does even though it was designed in 1959, finally built with a much different layout in 1995, and opening in 1997. I don’t know why some of Wright’s posthumous work is acknowledged and some not. Some of the poured masonry at Petra Island is hideous so, in my gut, I feel it this house is an authentic work. If construction drawings show it was built as intended, then Frank Lloyd Wright was clearly no Gio Ponti when it came to putting stone and concrete together.

It seems a bit arbitrary, but The Foundation may have objected because construction of the house was not begun during Wright’s lifetime, but forty-plus years after his death. But how often would FLW have visited the site and overseen construction anyway? Site supervision by the architect should be a guarantee of authenticity but we’ve no way of knowing how much FLW did. Unless we comb company accounts for the number of site supervision hours billed to Frank Lloyd Wright, we can only make an informed guess.

In 1950 Wright completed sixteen houses [three in California, two in Minnesota, Arizona, Indiana, Illinois, and one in Missouri, New Jersey, Illinois, Michigan, New Jersey, New Hampshire, Ohio, Texas and Tennessee], nine in 1951 [two in Michigan, and one in Wisconsin, Illinois, Ohio, Arizona, Mississippi, California, South Carolina and New York], and eleven in 1952 [one each in Michigan, Arizona, Washington, Wyoming, Idaho, Minnesota, Connecticut, California, Florida, Virginia and Oklahoma]. If you’re knocking out houses at a rate from one every three weeks (1950) to one every six weeks (1951) across the US from Wisconsin to California to South Carolina to New York as in 1951 when the above house was designed, then my informed guess is “not much, if any.” If that’s the case, then it matters very little if Frank Lloyd Wright was dead or alive when they were built.

And so we come to AI. An employee of Zaha Hadid Architects has been posting to LinkedIn his attempts to bring Zaha Hadid back to life and continue adding value to the employee-owned company.

ChatGPT-4/Midjourney have obviously been trained on too much Vincent Callebut.

Apart from the fact that the design moves of Zaha Hadid will sooner or later become as outdated as those of Antonio Gaudí or Frank Lloyd Wright or even Bramante for that matter, this endeavor raises important legal as well as ethical questions. Visual artists and writers are understandably angry their work has been data scraped without their consent and used to train AI but it’s already too late to stop that process and, given the requirement for legal proof, probably impossible to reverse. When a founding partner of an architectural firm dies, it’s not unusual for its designers to continue designing in same manner. It’s also not unusual for other practitioners to design in a similar way as some de-facto “school”. Either way, it’s called a legacy.

Perhaps Zaha Hadid wouldn’t have objected to her work being used in this way to generate revenue for the company, but what if that legacy (or similar training data) is used to generate posthumous knock-offs and income for some other company? Either way, it’s not a good look to exploit the dead for commercial gain. Does one party have a stronger moral or legal right, or do both have none? We’re talking levels of inauthenticity here.

How much homage can a market take, especially now that cutting differently-shaped cookies has become that much easier? If an AI work produced in this way is ever marketed to imply that it is “what Zaha Hadid would have designed had she been alive” – and this looks like the way it’s going – then there’s a big problem with authenticity regardless of who claims it. What’s more, if a work that’s been AI-generated comes to be generally accepted as “authentic” then will the resultant profits be split between the company and the estate of Zaha Hadid because Zaha Hadid is effectively still working?

Death is supposed to give meaning to life. In our rush to exploit AI for short-term profiteering, we seem to be willing to give up our humanity quicker and far more easily than AI has prompted us. And all of our own accord.

• • •

What Happens When Architects Die?

Career Case Study #8: Gio Ponti

The Aftermath