Human(e), AI

A warm welcome from the project leader

E’en the great globe and fabric of the world, / In harmony upheld through countless years, / In one dread day shall crumble and dissolve;— / Though well I wot the announcement seemeth strange, / And hard by words alone to win belief;  / As of a thing that bears no ocular proof, / Nor can subject be made to sense—surest / And straightest way to reach the minds of men— / Yet will I speak; the dire event may give / Faith to my words, perchance, and you may see / The steadfast earth to tremble as with throes. / Which far may ruling Fortune ward from us, / And reason, not the dread event, convince / That all may sink, in dire destruction whelmed.

 


Titus Lucretius, De rerum natura, i.e. on the nature of things, excerpt, translated by Charles F. Johnson

 

When, in the mid-1st century BC, Titus Lucretius, an Epicurean, wrote his monumental work De rerum natura, i.e. on the nature of things, his aim was to free humans from fear of Roman gods and make them understand the world did not depend on the will and games of gods but rather on itself. Four elements, earth, water, air and fire, constitute the world, he wrote; they change, arise and expire . The same goes for humans, who, like the elements, are made up of tiny particles, atoms. Everything has a beginning, wrote the great Roman, and everything shall ultimately “tremble as with throes” . But this finding about finality should not diminish a person’s zest for life or courage to do great things in life; on the contrary, it should act as an encouragement to both, as our time is very short and thus very precious, also making us, i.e. humans, very precious.

 

 

When today, more than two thousand years later, we think about artificial intelligence (AI), the reflections of Lucretius seem an inside-out parallel of sorts. While the Epicurean tried to free humans from fear of gods, who, as was the general view of that time, regulated the world and were the origin of everything, the point now is to accept the fact that it is not only humans who make all discoveries any longer and that humans no longer regulate everything alone. Alongside and with them, there is AI, whether we like it or not. In other words, AI is a new coordinate of our existence, in addition to our three other coordinates, i.e. humans themselves, their nature and their authority.

 

Until recently, AI research and development were the subject of high hopes and were followed with great enthusiasm, but then the enthusiasm was replaced by worry, and the hopes shattered by fear. Last spring, a group of distinguished scientists, including those that had previously been the main figures in AI development and are thus called the godfathers of AI, sent an open letter, entitled Pause Giant AI Experiments, to the world, calling on labs developing large AI systems to suspend their work for six months and, in the meantime, develop guidelines on AI control in collaboration with political decision-makers and civil society. They called for an “AI Summer”, and this is what they wrote.

 

 

Contemporary AI systems are now becoming human-competitive at general tasks, and we must ask ourselves: Should we let machines flood our information channels with propaganda and untruth? Should we automate away all the jobs, including the fulfilling ones? Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us? Should we risk loss of control of our civilization?

Such decisions must not be delegated to unelected tech leaders. Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable. This confidence must be well justified and increase with the magnitude of a system’s potential effects.

OpenAI’s recent statement regarding artificial general intelligence,  states that “At some point, it may be important to get independent review before starting to train future systems, and for the most advanced efforts to agree to limit the rate of growth of compute used for creating new models.”

We agree. That point is now.

 


Let’s thus enjoy a long AI summer and not rush unprepared into a fall, they added; if not, to use Lucretius’ words, “all may sink, in dire destruction whelmed”.

 

 

But the summer of respite from the headless AI rush never came; instead, there was a summer when the AI horizon glowed brightly like an ember in the wind. There was yet more rushing, yet more development and thus yet more worry and reservations and unease and yet more fear. While the European Parliament passed the AI Act this spring to regulate its use and development, this document, the first political act on this topic in the world, has failed to please or appease anyone.

 

 

In this context, the fundamental principles according to which AI should be regulated in a way to achieve the best possible results for the well-being of humans and the world at large have been known for some time. As early as 2017, they were formulated as the Asilomar AI Principles but have remained unknown to virtually everyone outside AI specialists’ circles.

 

The unease and reservations and fear regarding events relating to AI development and use thus initially seem fully justified, particularly because the development and use of AI have primarily targeted the two fields that technological advancement has always targeted first since the earliest times of human society. They have targeted, first, the establishment and perpetuation of political authority and, second, military and war. The two fields are closely intertwined, and therefore there is unanimous enthusiasm among holders of political power and their generals about the potential of AI. Rulers around the world, regardless of their measurable despotic spirit, are enthusiastic about AI surveillance systems, and their chiefs of staff are already saying that the development and use of AI battlefield systems entail the most significant change in the nature of warfare in the entire human history.

 

 

Naturally, such statements come with some tragic–bizarre pathos. The fact that such AI innovations are developed by private tech companies and tested at the currently most tragic battlefields of the world cannot but make an observer of the developments feel a tragic pathos of powerless sadness. There has been no lack of suggestions on how to avoid the giant risks relating to all this; in the spring of 2023, for example, they were presented for debate in a document entitled Policymaking in the Pause – What can policymakers do now to combat risks from advanced AI systems by the Future of Life Institute, but this document, too, has gone virtually unnoticed.

 

And while rulers and their generals and cutting-edge companies see AI as a welcome tool to achieve their goals relating to power, war and/or profit, modern sciences have been very reluctant about AI. The report prepared two years ago by an esteemed Australian institute for analysing scientific research and scientific research policy, CSIRO (The Commonwealth Scientific and Industrial Research Organisation), is sobering. Modern sciences, from natural sciences, technology and life sciences to social sciences and humanities, with the unsurprising exception of computer sciences, are very reluctant about AI at the very least, and often their attitudes are even completely negative.

 

This sobering fact points to an odd historical uniqueness in the state of modern sciences. We are witness to the strange fact that sciences reject a cutting-edge tool they could use in their research, studying its potential and continuing its development. To again employ a somewhat but not extremely inside-out parallel, this is as if the discoveries of Marie Skłodowska-Curie and Albert Einstein would have been renounced by past physicists, those of Alan Turing by mathematicians and especially computer scientists, and those of Michel Foucault by the humanities and social sciences.

 

 

Why this is the case can perhaps be attributed to the same reasons spoken about by Jean Jacques Rousseau in his 1750 treatise responding to a prize question of the Academy of Dijon: Are arts and sciences responsible for corrupted morality? He wrote that sciences and arts, which remain committed to the given mercantilistic – we would now say neoliberal, or, better and more specifically as well as simpler, as things are simpler, neofeudal – everyday life of the world, are a threat to humans in the world as they lack the desire for freedom, a human’s desire for freedom being a prerequisite for each everyday living world of unfreedom to surpass itself and for a new and better world to arise sometime in the future, a better and nicer world for life in freedom.

 

 

The great reservations of sciences about AI and even their resistance to AI is yet more astonishing because AI as a tool for scientific research is only slightly new in its nature; what is new is the previously unimaginable scope this new tool can reach and the unprecedented speed it can achieve. Otherwise, the so-called literature-based discovery (LBD), scientific discovery based on studying previous research that scientists can make with the assistance of AI, is no different in nature from making discoveries based on scientific journals since the 18th century. When, in the mid-19th century, scientific institutes and laboratories we know today started to form, their purpose was the same as the purpose of so-called self-driving labs (SDL), which are autonomous laboratories where AI-assisted experiments of one’s choice can be conducted, though admittedly now experiments are virtually countless and take an almost imperceptibly short time.

 

Of course, both LBD and SDL, just like the editorial policy of scientific journals since the 18th century and the scientific research policy of institutes and laboratories since the 19th century, must be restrained by sound scientific knowledge, steadfast ethics and an irrepressible desire for freedom; in short, by a human worthy of the name. This is the only way to reach a discovery and its results without ultimately everything sinking, “in dire destruction whelmed”.

 

When this – i.e. humans and AI in a good, good-natured and benevolent coupling of well-being – becomes possible sometime in the future and we are able to enthuse over it, the fears mentioned above, which are not baseless, can spontaneously make way for a great and grand hope of a better, nicer world. There will be a new era, novacene, as named by James Lovelock, a doyen of futuristic sciences. But just as in Lucretius’ time the gods had to step down as regulators of the world, in this new era humans are no longer the only regulators of the world but are complemented by AI, because it is, in addition to humans, a source of discoveries about the nature of the world. But just as in the past, in Lucretius’ time, now, sometime in the novacene, we will need Fortuna, the goddess of luck, or both AI and humans will, to the best of their abilities and beyond, threaten both the sky and earth, ultimately driving them, and we can bet on this, to doom.

 

 

Is this all just a utopia, a “deep utopia” as discussed – with much contrived enthusiasm – by Nick Bostrom , a British philosopher? For utopias to have any chance of coming true, it is not enough to be able to imagine them; it is equally important to put them into words. This has already happened. At the German Pavilion at the 17th Venice Architecture Biennale in 2021, architects, urban planners, spatial planning researchers and artists gathered around Olaf Grawert, Arno Brandlhuber, Nikolaus Hirsch and Christopher Roth answered the curator’s question “How will we live together?” with a vision of the future of an orderly, just, good, peaceful, in short, blissful and serene world.

 

They named their vision 2038_The New Serenity. 2038 is the year when they imagined humans would manage to overcome all the great challenges of their world and time. 2038 is a year of peace.

 

The ‘20s were tough. / It was only in the early 30s / of the 21st century / that the world came to its senses.  / Instead of eliminating the broken capitalist system, / we used the infrastructure as a platform. / States, institutions and tech-companies joined forces / and designed a universal, adaptable / and viable system with collective goals. / While local structures have enough autonomy / to organize and manage themselves decentrally. / Equality is our main objective. / Taxes are celebrated / as part of permanent redistribution. / In the past, technology was enslaved / for purely capitalist purposes, / leaving the real transformative potential / of many achievements untapped. / Existing, sometimes forgotten ideas were activated / with the fundamental sociopolitical change / in the late 2020s. / Architects have repeatedly been part of the solution. / Architects who had answers and not questions. / Eco-systems are not legal entities, / most rivers, lakes and mountains even own themselves. Drama is almost forgotten, / we live in a radical democracy / and in a radical bureaucracy. / We need neither heroes nor villains. / We call our time The New Serenity.

 

 


Can such a New Serenity only be a Deep Utopia? There is reason enough to hope that won’t be the case. In October 2021, the Beethoven Orchester Bonn, a world-famous orchestra, conducted by Dirk Kaftan in the great hall of the recently built and beautiful Elbphilharmonie in Hamburg, played Beethoven X. The song was created based on forty compositional sketches that Ludwig van Beethoven had prepared as drafts for his 10th symphony, which he had never finished due to his death. Beethoven’s final symphonic work is thus the renowned 9th symphony, which includes a magnificent choral part in the final movement, the Ode to Joy: “Your magic brings together what custom has sternly divided. / All men shall become brothers, wherever your gentle wings hover.”

 

Beethoven’s 10th symphony has thus remained unfinished and is still unfinished today. It has, however, been supplemented shortly before the 250th anniversary of Beethoven’s birth by Walter Werzow, an Austrian composer, Robert Levin, an American musicologist, and Mark Gotham and Ahmed Elgammal, American computer scientists. The supplementation was made by an AI programmed and designed specifically for this purpose.

 

 

They named their work Beethoven X – The AI Project. Based on the mentioned forty compositional sketches by Beethoven and with the help of AI, they put together two symphonic movements, i.e. the so-called third movement, scherzo (allegro – trio), and the so-called fourth movement, rondo (adagio maestoso – tempo di menuetto). In October 2021, the work was performed by the orchestra mentioned above, and in the second and mouth movements, which are designed as a dialogue between the orchestra and an organ, they were joined by Cameron Cooper, a world-renowned American organist.

 

 

Is this work a work by Beethoven? Of course not. Would the work have been made without him and his admirers almost two centuries after his death? Of course not. Could the work have been made without AI? It could have but wasn’t; if it had been, it would’ve been different. Were the algorithms able to create? They were. Can humans be even more creative with them? Those who are otherwise creative can. All these were questions raised during the creation and performance of this work and after it. The answers shouldn’t be too difficult.

 

 

Another matter is the question whether this is a work that addresses a person, as art does, in their feelings, stimulating the heart and mind and leading humans, as arts and sciences do, to new discoveries. This question is different, each person must answer it for themselves by themselves.

 

With the performance of Beethoven X in Elbphilharmonie, it was hard to provide an answer. This is because the work was performed by one of the best symphonic orchestras of the time and organist that has already broken all sound barriers of classical and contemporary music and opened up brand new horizons, not only in music: the work then resonated in the performance of a human, artist and researcher; in short, the best there can be.

 

And that’s what this is about, humans at their best. The best human is a free one. Today, they are accompanied by AI. And it is here to stay.

 

… I understand, it always depends on how you view things … sometimes you see nothing at first … besides, it’s not that we want to do everything well, first you have to do something at least … it’s not about understanding everything at all there is … no-one really knows where this is leading to and where it’s come from … there’s nobody who understands at least a little of what’s going on … wait, let me make a little detour – beauty eludes us because we don’t watch it, and when we don’t see anything any longer, that’s the only thing we keep thinking about … we have nowhere to hide … nobody knows where and how, and that’s a good thing … we have yet more to say … nobody will come to rescue us … God is hiding in a hole, if he ever existed … sometimes you must dare …

 


Part of a monologue from the show Qui som?, i.e. who are we, by the French-Catalan artistic collective Baro d’evel at the theatre festival in Avignon in July 2024

 

“… you must dare …” That also applies to humans and AI. We shouldn’t be scared.

O, Fortuna, be with us!

 

 

 

 

*

 

 

Nick Bostrom, Deep Utopia – Life and Meaning in a Solved World (Ideapress 2024).

 

James Lovelock, Novacene – The Coming Age of Hyperintelligence (Penguin Books 2019).

 

Charles Taylor, Cosmic Connections – Poetry in the Age of Disenchantment (Harward University Press 2024).

 

Nigel Toon, How AI Thinks – How we built it, how it can help us, and how we can control it (Panguin Random House 2024).

 

 

logo