top of page

Solid impressions of an uncertain future

Don Albert

‘Stereotype’ originates from the French word stéréotype, which is derived from the Greek words στερεός (stereos), meaning "firm, solid," and τύπος (typos), meaning “impression;” a ‘solid impression’ therefore. In more popular parlance, the first recorded use of the word ‘stereotype’ as a noun, meaning an image perpetuated without change, occurred in 1850. However, it wasn't until 1922 that the term ‘stereotype’ gained its modern psychological (and more pejorative) sense when American journalist Walter Lippmann used it in his book Public Opinion.


‘Artificial’ late Middle English: from Old French artificiel or Latin artificialis, from artificium ‘handicraft’

‘Intelligence’ comes from the Latin word intelligere, "to understand,”.

The Persistence of Memory - 2023

50cmx 50cm

Archival Print on Metal, 1 of 10, A.O.R

Futurists by definition


The role of drawing and painting in architecture has evolved alongside human civilisation, serving as a fundamental tool for the conceptualisation, communication and documentation of architecture, closely bound with humankind’s understanding of itself. Frank Lloyd Wright called architecture the ‘mother of all the arts’, unsurprisingly for one of modernism’s icons, but not many have argued with him on that point. He also said that civilisation would have no soul without it.


From its prehistoric origins representing animal and human interaction in cave painting to the geomantic, astronomical and mythical symbolism of the western classical period; from the invention of perspective in the Renaissance to the birth of modernism in the Age of Enlightenment; from the abstraction of the modernist era to the fractured collages of deconstruction in the post-modern era, and finally,  to the parametric era where three-dimensional mathematical constructs exist before two dimensional output is rendered — architectural visualisation has danced in step, if not occasionally led, the broader philosophy and aesthetics of the times. It has also used every medium and tool available at the time to do so.


Renaissance polymaths like Michelangelo and Leonardo da Vinci deftly moved between a variety of subject matter and media, using drawing, painting, sculpture and poetry to expand human understanding of biology and physics, often pushing against the confines of religious doctrine.


Since the Renaissance, ‘paper architects’ such as the utopian and revolutionary Enlightenment era French architect ´Etienne-Louis Boullée, and equally radical 20th Century deconstructionists Peter Eisenman and Zaha Hadid have initiated entirely new architectural discourse about form and space through their paintings and drawings, well before built realities of the same became appreciated by the mainstream. Architects draw in order to predict the future, and, when working with data, facts and scientific forecasting, become futurists by definition.

You Can't Take It With You - 2023

50cmx 50cm

Archival Print on Metal, 1 of 10, A.O.R

From fantasy to posthuman collaboration


My first experience of drawing on a computer was on an Apple Macintosh 128k machine using the ‘MacPaint’ raster graphics editor, circa June 1985, at the age of fourteen. The next was with the now cultish Amiga 2000. The type of art produced on these low-bit programmes, ‘pixel art’,  has had a resurgence in recent years with the invention of non-fungible tokens as a viable virtual distribution mechanism. The ‘pixel generation’ of which I am a nostalgic member, has grown up, but still appreciates such artworks produced with relatively primitive technology.


Since the 1980’s, my livelihood has been inspired and enabled by the creative capabilities offered by a range of 2-D, 3-D, video and music production software. Despite this, I have never given up on analogue drawing and painting, for pleasure or profit, and have often blended both analogue and digital methods for painting, in and outside of architectural production.


Indeed from the dawn of desktop publishing until today’s computationally incomparable explorations in Artificial Intelligence (AI), I have grappled with technology and advocated, somewhat evangelically at times, our need to instrumentalise computation in order to gather information, analyse, synthesise, render and ultimately execute all manner of design solutions to complex problems — from the size and ‘simplicity’ of a chair, to the scale and complexity of cities.


Why can’t we only do this in a purely analogue way you may ask? The answer lies in the computer’s superior ability to handle vast quantities of information and make ‘objective’ averaging conclusions; expose trends and patterns in ways that rapidly exceed human capacity, to synthesise new sounds, forms and spaces, and to represent the formerly unseen, and thus unknown, tangibly, for the purpose of decision-making and/or pure aesthetic enjoyment. Computation as a creative tool to engage with, to master, and enjoy — either via the process, or via the output, or both. The computer is also very good at handling types, ‘design models’ and representation modes. Until recently, the computer has been an excellent slave, and a more than capable foot soldier in the furthering of modernity, notwithstanding its socio-ecological consequences.


Indeed, the computation process is not without bias, neither within its original coding, nor its inputs, nor its outputs, nor in the interpretation of them, and at the very least, ‘garbage in equals garbage out’Such is the nature of any artifice and the vagaries of language and meaning itself .


The same applies with analogue processes which are often intuitive and deeply connected to the author’s cultural background and formative years. Analogue artistic practice and literary thinking reveals a range of universal concerns that are shared amongst most cultures, for example: the pursuit of ideal love; the hero’s journey; creation and origin myths; moral dilemmas; nature and the environment, coming of age and so forth. These common tropes exist in both written and oral cultures and have become stereotypes now embedded within AI’s databases, records of which will continue to widen and deepen.


Until the advent of user-friendly “no-code” AI, in other words until very recently, my interest in the digital was mainly as a tool for achieving design outcomes for architecture and music production. It was mostly goal-oriented and not a completely formalistic fascination in-and-of-itself. Indeed, to operate any particular software, until recently, required a thorough understanding of its capabilities, menus, instructions and techniques with a steep learning curve and a ‘pre-launch plan’ of how it would be employed, more or less, in order to achieve some kind of usable output that hoped to meet a brief.

Carbon - 2023

50cmx 50cm

Archival Print on Metal, 1 of 10, A.O.R

No-Code, no rules


With ‘no-code’ AI, all this has changed. The computer used to fit well within human-centric dualistic narratives of control, as long as it ‘behaved’. Now, with the refinement of graphic AI interfaces, ‘it’ increasingly has a mind of its own, and is rapidly ushering an epoch where former human/machine hierarchies of control and indeed sentience itself are being questioned, if not completely destabilised.


Unsettling as this may be for some, for me, with AI, ‘painting' becomes loose again, non-goal oriented. In this new area of creativity, both the frame of reference and the store of raw ingredients, are blown wide-open. The results are unexpected, uncanny and sometimes totally new.


Armed only with one’s imagination and the most cursory grasp of typing, one can now revel in a fantasy land, ‘imagining’ new scenarios and ‘blending’ scenes and inputs to unforeseeable effect, each building off the last in an evolutionary way — one in which traces of the original inputs can be recognised further down the line as a kind of ‘digital DNA’, almost like the receding red hair gene of a great-grandfather you thought had long died out, popping up in a newborn niece.


Like evolution itself, AI generated imagery has a ‘genetic’ undercurrent, building off of historical reference material (or hand sketches) as it goes, but with the ability to morph, mutate and shift, sometimes by instruction, sometimes purely by chance, often with bits of ‘history’ popping up in unexpected places. AI generators like Midjourney and StableDiffusion present multiple options based off a single prompt. Some can be selected for further modification and be blended into others of a different prompt or origin, while others are shelved. What remains in on the screen is no longer inert, it can always be ‘remixed’.


This ‘guided chance’ aspect, as in most algorithmic processes, is only somewhat intuitive when it comes to AI because AI is drawing on such a vast repertoire of data and does so in a way that often defies expectations. AI is driven by language but is completely indifferent to the meaning or morality of it, at the time of writing at least, and in the prompting remains a wide margin of potential disconnect between the user’s intention and the software’s understanding of them.


The algorithmic process involved in any digital design process is one that needs to be practiced in an abstract way in order to achieve originality in design, thus in that sense, I am simultaneously intrigued and perturbed by AI, drawing on as much stereotypical precedent as it does. In an ironic way, it is the history of AI’s hidden sources that fascinate me as much as the uncanny output it often generates. On a scale that makes Boullée’s sublime imaginings of gargantuan public buildings (and humankind’s role in them) rather twee by comparison, AI knows everything and nothing at all. In that sense, it is all too human.


To interact with AI therefore has been surprising, almost as good as having an unhinged conversation with someone who is on the same wavelength as you, but knows far more than you about certain things — joining dots and being at cross purposes all at once. It becomes a  collaboration reminiscent but somehow much less predictable than Warhol and Basquiat’s, each bringing their own baggage onto the same canvas but in totally different ways.


Syphoning from the colossal, yet not fully comprehensive reservoir of the history of art, architecture, film, photography, illustration, anime, comics, advertising, music, high literature and random holiday photographs, AI not only has access to ‘important’ subject matter, but also seems to know how best to represent it stylistically, and provides options to override or change any of it predetermined stylistic options too. In terms of computation, by using predictive methods, it arranges pixels in a sequence defined by probability algorithms supported by cloud computing, based on user prompts and or uploaded imagery.

Microplastic - 2023

50cmx 50cm

Archival Canvas Print, 1 of 10, A.O.R

Whereas software developers are now grappling with providing discrete layering into the 2-D images that AI generates (in order to deliver a more 3-D outcome),  what intrigues me currently is the purely graphic nature of the medium. The AI I am fascinated with in UNDER WATER WORLD is entirely optical and brings ripples of Op Art into the discussion. Since AI has no concern about subject matter or genuine three dimensionality — only its historic references — the stereotypes that inevitably result please us aesthetically based on what we already know.


The images AI can produce therefore, have compositional and lighting qualities which are somewhat canonical, based off historically ‘significant’ and popular tropes of the past. Where newness of object and/or representation and/or style emerge, it is usually via the (deliberate) cross-pollination of human prompting or, sometimes, a complete misunderstanding. The “hand” therefore of the artist can still be very present and I would refute arguments about any ceding of authorship — it’s a collaboration at the very least. This is the aesthetic equivalent of digital audio sampling mixed with genetic splicing. Royalties may need to be split…


For me, for the first time since 1985, experimenting with graphic AI seems like a genuine collaboration with machinery, as opposed to a simple “master/slave” transaction of relatively predictable inputs and outputs. It has transformed my process of creation into something far more out of control and thrilling to me — as exciting as painting a watercolour, or dealing with any tricky paint medium that has to be handled just so, and at times as frustrating as collaborating with humans too, when AI stubbornly refuses to do what you imagine it should.


This shift brought on a painterly revival for me, one in which I could finally conjure the recurring dreams, hallucinations and nightmares that I have experienced over the years. Some of these breach the surface in UNDER WATER WORLD, visions that would otherwise have been impossible to render in the same fashion or speed.

It is all syntax and no semiotics, but not for long…


As far as we know, homo sapiens are but a subset of only five distinct life forms on earth that tries to derive meaning from their existence. This makes us special of course, but only to ourselves. We may not be the only ones in the universe, but here on earth more critically, our ‘species superiority’ has produced a debilitating “intellectual hubris” as it were, fuelling the destruction of our life support system through the avarice of multinational corporatocracy, the choking of our perceptions, and the snuffing of any imagination of how to change our behaviour and systems.


The purpose of language for humans is not just transactional but also for philosophical contemplation and the appreciation of beauty, for collaboration and human bonds to exist. However, the large language models that AI relies on are not yet capable of any of this, and as they are, have the potential to distort truth and create a considerable amount of global pandemonium if unregulated. Indeed, consensus is building that the singularity of technological ‘superintelligence’ is entirely possible, and very soon.


It is important to grasp that homo sapiens are the last remaining hominids from a long and complex evolutionary line. After us it is ‘game over’ for organically intelligent life of the human variety, if anyone cares. Notwithstanding the melodrama attached to such a grandiose sentence, shouldn’t we care about the world we leave to our children, and theirs? Where and how are they going to live in a future so clearly threatened by both climate change and artificial intelligence?

No Room At The Inn - 2023

(Not for sale)

Fires and Floods - The genesis of UNDER WATER WORLD


My ongoing concerns about human settlement in the age of the anthropocene, amplified since the Australian summer bushfires of 2020 that trapped me in my car on the beach at Turros Heads in New South Wales — while researching a story on community responses to those bushfires for Climate Change Cities no less;  microplastic pollution of the oceans; and, the plight of climate change refugees, have all been funnelled into AI prompts that fused repugnant words and concepts into relatively ‘digestible’ images and poetry. Rapidly, a new under water world emerged, with each strain of creature taking a trajectory of its own, each in a fractal, cascading fashion, each prompting further exploration of its world.








A Mappa Mundi



At the outset, I did not set out to create a post-apocalyptic A Mappa Mundi, but somehow, the AI made me drift into uncharted waters. My initial explorations with Midjourney and ChatGPT, the outcomes of which were cross pollinated into one another, started to take on a certain aesthetic and narrative logic of their own. In each case of the poetry that I prompted, ChatGPT resolved to give them a happy ending or a ‘moral of the story’ some of which I had to edit out to keep true to themes. Sometimes it completely misread the context of the prompts.


In terms of creativity the AI phenomenon is undoubtedly post-modern, but by virtue of its immense speed and the super-human range of historical sources it draws from, working with AI is also a profoundly ‘posthuman’ experience. It knocks the human down a notch and yet ironically, it relies on human history and ingenuity in order to do so - the artificial is, after all, ‘man-made’. The foundation of AI is a vast, biased, morass of information, the most common of tropes of which are… rather stereotypical.


This is not a completely unwholesome proposition in terms of critical posthuman thinking around mores, climate change and the survival of life on earth, because by virtue of ingrained literary tropes, ‘everyone resolves their differences and pulls together as one’, but it does strike one as wishful thinking. As we know, stereotyping leads to reductive pigeonholing, othering and allowing the powerful to maintain control so questions arise: in whose hands will AI end?; how can it help the marginalised?; can historical narratives be relied upon when what we need may be a completely different tack?; and most importantly, can AI save humanity from its human impulses?


Given the the way in which they have been crafted, the artificial impressions in UNDER WATER WORLD are in no way capable of competing with the richness and beauty of real undersea life, and they are not meant to. By taking an absurd, abstracted, and “fantasy” approach, indeed with anthropomorphic characters and stereotypical narrative tropes, I am intending to provoke discussion around humanity and the future of species. Hopefully, the resulting characters, scenarios and my AI assisted attempts at world-building around them, provide access to an anti-speciesist debate around humankind’s responsibilities and the our understanding of sentience in an engaging way.



Save Our Souls


My three years of journalism about the climate crisis and the future of human settlement for Climate Change Cities was crucial to the initiation of UNDER WATER WORLD frustrated as I was by the relentless trolling by Australian coal-lobby bots against our articles on social media, and thereafter, by the Covid-19 pandemic which disrupted my personal movements and climate change activism for many (it’s hard to have school strikes for climate when everyone is home schooling), so, I would hope that this work can reach a wider audience and bring focus back to the most pressing issue for life as we know it, i.e. human induced climate change. We need to radically reduce greenhouse gas emissions and prepare for the re-settlement of ourselves and many other species in anticipation of the climate catastrophes being predicted, and for those already underway which have exceeded previous predictions.

The Siren - 2023

80cm x 80cm

Archival Canvas Print, 1 of 10, A.O.R

National responses to the covid-19 pandemic proved just how quickly, resourcefully and effectively governments can act when faced with a crisis on their watch. A raft of mostly successful vaccines were developed in record time thanks to the sophisticated ‘real-time’ collaboration of international scientists across borders while, ironically, many of the newly elected centre-right governments in the western world that denounced such globalist cooperation provided deep-pocketed support for their populations, and for medicine, and crucially, restricted people’s freedom in ways unseen since World War II. The hole in the ozone layer was also sewn up through international cooperation and legislation. This is a glimmer of hope. International cooperation is possible, and it works.


The problem with climate change compared with a pandemic is that most politicians believe that climate catastrophes are beyond the horizon of their watch and therefore any personal compunction around fossil fuel decision-making and their enabling role in corporatocratic avarice is moot. As the climate disasters unfold week after week, displacing 21,5 million people annually on average in both rich and poor nations, this is proving to no longer be a ‘safe’ political or existential stance.


The corporatocracy, those 100 largest multinational corporations who are responsible for over 70% of carbon emissions as reported in the Carbon Majors Report of 2017, could be repurposed for good. Their potential for the distribution of genuine nourishment to the world’s less fortunate will do far more for collective intelligence, ‘progress’, and survival on earth than artificial intelligence ever will. The building industry is responsible for 36% of carbon emissions and as genuine as certain professionals and innovators in the industry may be, no amount of fretting about the thermal efficiency of glass, or the carbon footprint of concrete, or the efficacy of insulation can stem the rampant destruction to the planet wrought by the over consumption of natural resources.


The only thing certain about the future is that if the world does not reach the 2015 Paris Agreement targets for maintaining the global average surface temperature below the 1.5 degrees Celsius threshold, much of the world’s coastal populations will be in severe jeopardy by 2100, to calamitous effect for the entire population of the planet — a humanitarian crisis certainly, but also a political tinder box of unprecedented proportions.


That is the future UNDER WATER WORLD imagines. Already the tipping-point of Earth reaching an average surface temperature of 1.5 degrees above pre-industrial levels is predicted by the World Meteorological Organisation (WMO) to be reached by 2027, if only temporarily. The likelihood of this happening is a 66% chance according to WMO, and should it happen, these spikes will increase in frequency.


In literary tasks, AI’s predictive language models seem to default to happy endings, but in terms of climate modelling it is warning us of the precise opposite. It’s up to us to correct this disastrous course by bringing human agency back into focus — as genuinely intelligent beings — and acting, not as neo-liberal ‘consumers’ curtailing our personal carbon footprint while Rome burns, but as empowered citizens demanding meaningful change at the political level, informed by genuine facts and science. The planet is going to get on just fine without us, if not better, so we do not need to worry about saving the planet, only to save ourselves.

Don Albert

bottom of page