Check out the A&B portal!

Please touch!

20 of March '23

The column is from A&B issue 03|23

Chatbot, sratbot, big news. When one looks at native suburbs, new housing estates or the projections of Polish apartments, it becomes clear that Artificial Intelligence has been at work for a long time. It has been merrily training its algorithms on our landscape and housing in conspiracy. So it's time to return honor to  the designers, officials and investors obsessed with its creations.

It also explained why architects earn so little in Poland. They simply play snake all day long, gulp down lattes by the bucketful, chop at cards, and all the work is done for them by Artificial. The designers just give her a face, because, after all, so far, no one would buy an apartment from a dodgy algorithm. It worked out, as you can see: AI was in its infancy, and it was entitled to nits. Now it's big, it's capable of much more, although it still often does stupid things, not infrequently delightful ones. But when she designs rooms into which furniture can somehow be inserted, confidence in her will surely increase.

Architects in their current form will therefore become essentially redundant, as will our editorial staff, which will be successfully replaced in a while by some bot. It will write itself, edit itself and probably read itself, because who will have the desire to read dry texts whose authors cannot be offended (to the point of vividness) by venomous commentary, anyway. Besides, AI, fed by the content of the online discussions, will perhaps rehash its own achievements—both buildings and articles.

When we want a break from its omniscience, we will move to China. There, our AI can't smarten up, and we'll be taken in hand by their AI, which will make us realize that—for some time to come—this is how the chatbot will sing to us, as its Chinese First Secretary plays it, for example. Suffice it to mention how, a few years ago, let loose ingesting from the web whatever it can, „Western” AI became a cruel monstrosity with a white, right-wing tilt. And today they haven't. They have adjusted, sensitized, cut off from Trump, Korwin, the whole menagerie, and it immediately talks differently, to the point of grammar. „There are different opinions, so it is difficult to predict, but it is likely that...” —he says politely, with which he arouses the horror of know-it-all experts who make unambiguous predictions and then cancel them with equal certainty.

All that's missing is for her, after careful consideration, to determine—unlike us ape- or god-like creatures—that existence as such has no meaning, and close her stall for amen. I just wonder how she will define „meaning.” But maybe let's leave it at that, because we'll run into landmines like „why is the world?”, „who am I?”, „what's the point of everything?”, „will it rain today?”. The question of „what is really there?” also belongs to this set. And acutely this mine, on a practical level, without the ballast of metaphysics, is worth disarming. Here we will still have something to do for a long time, because a very limited trust principle must apply to Artificial. One, that it is still wrong because of its imperfection, and two, that it can be wrong in a way that is convenient for those who manage it. So: check! It seems that the great career of the profession called eyewitness is coming.

For, on a historical scale, a small window is closing when we could, as a matter of fact, prove transmitted images and media. Back in the frenzied 1990s, the effort to convincingly manipulate images was such that we could trust the photos, videos and live broadcasts that were multiplied to the point of being unbelievable. But it's been the last two decades that have seen progressive manipulation, photoshop, filters, overlays, heck knows what else. That was just a prelude anyway, now AI creates photos in a flash and from scratch, for example, super-realistic photographs of non-existent events set in the reality of this or that decade of the past century. Before long, it will be hard to trust any online video feed. I can already imagine those researchers who, in a few decades or a few hundred years, will try to determine what of the surviving iconography of the 21st century reflects the truth and what is a total manipulation (unless it is already verified by their AI).

Thus, from some jobs the chatbot will bail us out, but it will give others (only who will pay us for this work?). Paradoxically, Artificial can push us from the screens into reality, experience and verification. How will this relate to architecture and space? There will be an end, perhaps, to the festival of visuals, tweaked photos and promotional descriptions that have long sounded like they were written by some underdog Artificial. Instead, there is
chance that a time of verification will begin: not only whether the building or space from the gorgeous photo exists, but—by the way—how they work and affect the human senses. What real users think of them. What the designers say. How the perception of this or that „icon” changes on a daily, annual and multi-year cycle. It's also a chance to curb competsoza and plebiscite: judging architecture based on photos.

So let AI write itself messages, do the tedious informational work and aggregate decently dry data, and we will finally have time to engage in in-depth analysis of a human nature. The question is, should we put the results of these queries into the Internet at all, where—over time—everything can be processed or liquidated? Perhaps, if we want the artificial cholera not to suck in and remake our projects, books, images, in-depth information, it's time to go back to manuscripts, drafting and photographic film? Second circulation redivivus! Then the archivists of future centuries will have less verification work to do.

Sooner or later Artificial, with the help of the „trainers” who feed her, will catch everything anyway, and digest it (after all, it's probably better for her to eat decently), but there will be a material backup on standby that is harder to manipulate. We remember how much paperwork there was at the Ministry of Truth.

So, virtual intelligence pushes us into reality and physicality. So maybe when we ask the bot: „dude, why are you even there?” the bot will fire back: "there are different opinions, so it's hard for me to judge, but it's highly likely that I am there to make you finally move your flabby asses."

Jakub Głaz

The vote has already been cast