Inside Computer Science

Course project for SOCY405 Sociology of Science (2023) at FLAME University, Pune, taught by Professor Sinjini Mukherjee.

Introduction

Over the past year, we have seen an insurmountable boom of public interest in Artificial Intelligence. New technologies and applications in the field of Computer Science, such as the Internet-breaking ChatGPT, have made AI a buzzword and a household term across the world. AI is not a brand-new concept: its conceptualisation began well over 73 years ago, when Alan Turing (1950) first dared to ask the question “Can machines think?” Computer Scientists know AI to be a fairly young domain which underwent much growth and development in the past decade, but one whose basics are still built on principles laid down in the previous century. While public perception of AI was initially limited to the futuristic sci-fi worlds of movies like The Terminator, its usage in popular applications from the past decade has created a sense of digital literacy – almost every popular app is powered by AI: from social media, to shopping, to movies, to music, to games etc. However, it was the introduction of ChatGPT by OpenAI that promoted AI from a futuristic concept or a behind-the-scenes player to today’s centre of attention.

By offering to the general public for the first time a live, tangible way to interact with state-of-the-art technology, society saw the roots of a scientific revolution potentially as large as the invention of the Internet (or even larger). The extraordinary capabilities of AI in applications that till now were not possible served as a reminder that the days of futuristic science fiction are indeed here. As with any other technology that threatens to disrupt the status-quo (and being the largest one in a while to do so), AI spawned an extremely passionate debate around its development and deployment. While some people swear by the scientific prowess of AI and all the good it can do at a species-level, citing it as the next step in evolution, critics of AI are unconfident in humanity’s ability to adapt to such technology in a safe and meaningful way. While the best-case scenario may lead to social good, progress and development all around the world, the worst case may just be the eventual decline, death and destruction of humanity.

In this paper, we will examine how a product of a seemingly apolitical scientific domain became a hot topic in worldwide politics. We wish to understand what led the field of Computer Science into the direction of pursuing Artificial Intelligence at a domain-level, how the decisions to do so were made, and how society external to the field became such a huge source of influence on the discoveries and creations associated with the field. By interviewing a number of Computer Science professors as well as the next generation of Computer Scientists, we wish to gain a deeper understanding on exactly how much influence society has on scientific research and development in the field of Computer Science.

ChatGPT, Sam Altman, and a Pause on AI

The motivation behind this research, as mentioned above, lies within the recently escalating debate about AI – most notably, the recent fiasco at OpenAI. OpenAI, the tech mogul behind ChatGPT, started off in 2015 as a non-profit organisation dedicated to achieving Artificial General Intelligence (AGI) which “benefits all of humanity”. The pursuit of this goal involves steady and cautiously optimistic research and development in the field of Artificial Intelligence, with oversight from a diverse company board involving both supporters and critics of AI development. However, over the years and under the leadership of CEO Sam Altman, concerns have risen from both within and outside the company regarding socially irresponsible methodologies of training AI models – prioritising efficiency over quality and not taking the time to assess and evaluate the negative risks to society. Coupled with an allegedly premature public deployment of ChatGPT and a partnership with Microsoft, critics argue that OpenAI has begun drifting from its original mission of safe and socially useful AI, to fast-paced and reckless development driven by goals of publicisation and commercialisation.

According to Helen Toner, a well-known critic of AI development and ex-board member at OpenAI, although tools like ChatGPT are highly capable of meaningful social utility, their public release should require a much more stringent evaluation process than what was followed (Imbrie, Daniels, and Toner 2023). Advanced AIs are no longer just products of Computer Science, but rather highly disruptive technologies with large and all-pervading effects on society. Careless public releases of such tools without adequate evaluation of their risks can have unseen ramifications, and could fundamentally alter human society – for better or worse. Thus, the threat that Sam Altman’s fast-paced knock-things-over development attitude has posed has cumulated into a civil war: between those who wish to develop and deploy safe and socially responsible Artificial Intelligence with the potential to benefit all of humanity and with minimal risks to society; and those who wish to push the boundaries and capabilities of AI as far as possible, as fast as possible.

Out of concern over Altman’s policies, motivations, seeming abandonment of the company’s mission, and lack of mindfulness regarding the social consequences of AI, the board fired Altman on November 17th, 2023 (Lostri, Rozenshtein, and Sharma 2023). Unfortunately, this move ended up backfiring. Almost all of the Computer Scientists at OpenAI, in solidarity with Altman and his ambitions, threatened to quit unless he is re-hired as CEO. Along with this, Microsoft – known for its cutthroat and profit-first attitude – came to be interested in Altman. The board feared Sam Altman leading large-scale AI development in a for-profit environment with zero safety checks or oversight, and thus, succumbed to the pressure: re-hiring Altman as CEO followed by resignations of some board members including Helen Toner (Lostri, Rozenshtein, and Sharma 2023). Although Computer Scientists rejoiced in being able to advance their research without having to worry about the boring and unrelated subjects like political and social impact, only time can tell what the effects of unthought and untested AI on society will be.

The debate at OpenAI is simply the most recent in a long list. Just earlier this year, an open letter signed by multiple political and tech personalities (including Elon Musk, Steve Wozniak, Andrew Yang and Rachel Bronson) called for a six month pause on the development and training of AI more advanced than OpenAI’s GPT-4 (Metz and Schmidt 2023). Citing “profound risks to society and humanity” from unplanned and unprepared AI development and deployment, the letter says, “Powerful AI systems should be developed only once we are confident that their effects will be positive and their risks will be manageable”. It adds, “AI labs locked in an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control” (Future of Life Institute 2023). According to the letter, a six-month “AI Summer” – voluntary or government-mandated – is necessary for humanity to catch up and adapt to an AI-centric world, develop policies and protocols, examine and evaluate the effects of such tools on society, and ensure our continued survival and prosperity.

Kuhn, Scientific Paradigms and Scientific Revolutions

The societal view of Computer Science is one of rapid expansion and excitement – one in which what was impossible yesterday becomes possible today – and for good reason. Computational power over the past 70 years has seen a trillionfold increase: an Apple Watch today is thousands of times more powerful than the Apollo Guidance Computer that first took humanity to the moon (Routley 2017). However, based on my interviews, the full picture of Computer Science research is less of the fantastical jumps in innovation it seems to be, and more of steady and sceptical acceptance of ideas over a long period of time. In Computer Science, research must often be differentiated from technology – while technology refers to the tangible outputs produced by applying the learnings of Computer Science, research is much more theoretical, and deals with what we do not know.

Keeping this boundary in mind, the professors have all noticed that over the decades that ‘the next big thing’ is never immediate. Computer Science, benefitting from both its young age as well as the technologically advanced society it exists within, does stay one step ahead of the natural sciences in terms of the pace of development and research. Still, researchers do not refer to this development as fast – at least not as fast as what the layman may think of it to be. Referring to Thomas Kuhn (1969) and his ideas of Scientific Paradigms and Scientific Revolutions, we observe that revolutions in Computer Science, although frequent, are not as rapid as they may appear to society. In fact, as one professor puts it, “Most scientific revolutions happen in hindsight.”

The normal response to any brand-new piece of research is a hefty amount of criticism – especially when the research appears to be a shift from the current paradigms of Computer Science. There is no ‘hype’ phenomenon in scientific research, and most researchers tend to approach new works with a healthy amount of cautious scepticism. Thus, a lot of the subsequent research work in the field will focus on pointing out flaws in this work, followed by the reparation, strengthening and improvement of the research. It is only after numerous alternating rounds of criticism and improvement that the research work will be validated by the community, and only then will it create a shift in the paradigms of ordinary science. For this reason, as one professor points out, the most respected and awarded researchers happen to be quite elderly in age – since it is only after years or even decades that their work is appreciated for being as revolutionary as it is.

It is only after a research work survives the tests of scrutiny and establishes its value to Computer Science that a shift in the paradigms of scientific research occurs. But when it does, it is almost always a major shift, pervading throughout the domain. Despite the seemingly unending number of applications and possibilities in Computer Science, once a scientific revolution is established, we see the creation of a new paradigm under which each and every research question – old or new, solved or unsolved – gets re-examined and re-evaluated. For example, with the rise of Deep Learning in the past two decades, not only has there been a rising focus of research in the field of Deep Learning itself, but it has also already been applied to every other Computer Science application (no matter how major or obscure) – from finance to healthcare to law to entertainment. As a result, today, one will find that Deep Learning is no longer an unheard solution to any and every possible research question in Computer Science.

As one professor recounts, even AI has had a similar story. The ‘AI Winter’ in the 80s and the 90s was a time of domain-wide disappointment in Artificial Intelligence. Due to research failures and reduced funding, AI was seen as a lost cause by the scientific community – except for a handful of researchers. It was the fruits of this research, during a time of widespread dismissal, that allowed AI to survive and become the worldwide phenomenon it is today. As remarked by another professor, scientific research – exploring the unknown and answering the unanswered – is by definition a gamble. Within this, while ordinary science focuses on looking at old questions from the safety of new but proven paradigms, it is the out-of-paradigm research that is the biggest gamble of them all: with the biggest risks of failure, but also the biggest rewards.

As one of the professors notes, “it may not always end up working out, and there have been a lot of such failures which only become apparent in hindsight, but if it does work out then you have something truly extraordinary in your hands.”

Merton, Societal Influence and the Scientific Culture

In his book The Sociology of Science, Robert K. Merton (1973) describes the ‘scientific ethos’ as a set of four Mertonian Norms: Communism (the common and collaborative ownership of scientific knowledge), Universalism (universal validity of scientific knowledge), Disinterestedness (lack of personal interests or agendas of researchers outside of investigating the scientific truth) and Organised Scepticism (institutionalised criticism and scrutiny of scientific research). He argues that together, these norms strive to create an apolitical, neutral and unbiased domain of scientific investigation. For the scientific community which prioritises scientific truth-seeking over anything else, these norms are the status-quo: internalised, self-imposed and self-policed. The motivations of researchers lie in a self-sustaining system of awarding scientific merit, credibility and fame – as well as from the personal curiosity and drive that brings them into the field of science in the first place. Computer Science, being a relatively brand-new field of knowledge, embodies this scientific ethos as well as any other field.

Merton also lists a number of possibilities of societal influences upon any such scientific field with an ethos that seeks to remain neutral. For a field with an ethos as strong as that of Computer Science, yet an impact on society as large as that of Computer Science, this remains all the truer. Once again, the differentiation between theory and technology becomes useful, as while theory remains detached and independent in its scientific curiosity from the understanding and influence of society, it is technology – the visible, tangible and comprehensible products of Computer Science – that allows society to latch on and influence the field and its direction (for better or worse). It must not be forgotten however, just as the technology of today was the theory of yesterday, the theory of today will be the technology of tomorrow.

Merton identifies two main sources of social influence on science, which dictate its ‘position in the modern world’: the express approval of scientific pursuits that seek to satisfy important values, and the outright rejection of pursuits whose ethos seems incompatible with the sentiments of society. These two forces are nowhere more visible than in the AI Race previously mentioned. On one hand, the shift in technology has spawned a wave of research into applications of existing AI. Business interests and governmental policies often lead to a disproportionate funding of application-based research, and a significant lack of interest and understanding of Computer Science theory: something that Merton defines as ‘anti-intellectualism’. As one professor explains, power in Computer Science comes in the form of grants – the ability to fund research pursuits.

Businesses and governments invest in a range of research efforts, not for an immediate or a guaranteed return, but for a long-term payoff. The outcomes of science are unknown, and, by definition, not guaranteed. So, what was often observed is that a single entity or organisation would be willing to fund a huge range of scientific pursuits, with the hopes that one success can pay for a thousand failures. As another professor puts it, “what is the return on investment for Isaac Newton?” However, the AI race has proven to disrupt this balance, as the survivability of such powerful entities has come to depend on putting out applications of AI as fast and as efficiently as possible. Despite the unknown, long-term and theoretical nature of science, we can observe today that most research studies happen to be practical, short-term and safe in their objectives. Researchers, now more than ever, are discouraged from breaking out of the current paradigms of Computer Science. After all, this is where the money is.

On the other hand, we can see that the possibilities of such AI applications have deeply upset the sentiments of society, threatening to disrupt the status-quo at unimaginable scales. Manifesting from this we can see anti-science movements by society targeted towards the field of Computer Science – perhaps nowhere more apparent than the previously proposed pause on AI. Referring to Large-Language Models (LLMs) – the technology behind ChatGPT and similar AIs – a long-time professor in this field notes, “there was a very positive perception until LLMs came into the picture. The flip side of LLMs is scaring people.” In fact, even a pause on technology is not a novel idea, and Merton describes a similar pause proposed by Sir Josiah Stamp in 1934, “in order that man may have a breathing spell in which to adjust his social and economic structure to the constantly changing environment with which he is presented by the “embarrassing fecundity of technology.""

We can observe that both of these sources of influence, despite their seemingly opposite interests, happen to target the technology-side of Computer Science much more than the theory-side. Merton describes this as a self-fulfilling prophecy: technology is the side of science which allows society to interact and communicate with it, but it is also the side which tends to be the most under social pressure and scrutiny. He writes, “readiness to accept the authority of science rests, to a considerable extent, upon its daily demonstration of power.” Insofar, technology becomes the medium that allows the recognition and appreciation of Computer Science research. However, similar to how Merton describes, while the theoretical nature of pure science fails to capture the minds and hearts of society as much as technology, it also manages to retain its autonomy for the same reason: social influences do not consider pure science to be extremely useful or worthy of interest. Surprisingly, it is this very disintellectualism movement that shifts attention away from theory to technology, that also protects theory from the disapproval of society.

For this reason, as one professor notes, we can find that while practical applications will be governed by market interest, theoretical research still very much belongs to academia.

When talking about whether power is able to dictate the direction of research (as initially suspected), and whether, consequently, a monopoly in power is possible and can be translated to a monopoly in science, all of our professors adamantly stated that this is not the case. Given the large number of players in this global field of open knowledge, an aggregated monopoly is an impossible concept – the competition is just too fierce. Even more so is the case with the individual researchers, with each of them pursuing one of an infinite number of research possibilities. In Computer Science, there is no organised control of the entire discipline, but rather a collection of independent and individual entities pursuing their own scientific research. As a result, as one professor puts it, “power is not the deciding factor.”

Relating to Andrew Pickering (1995) and his idea of The Mangle of Practice, Computer Science is – as much as any other science – a field which seeks to learn how to accommodate the resistances of non-human agents, who have their own agency. In this way, although social power can allow an organisation to explore a given direction of science, it is still in no way the make-or-break factor. Scientific exploration means that researchers bow down not to the power of the sponsor, but to the agency of the objects that must be researched. Due to the unpredictable nature of science (even Computer Science), even the largest corporate organisations could fail in their efforts, and even the smallest research team could be successful in theirs. A professor states, “no planned research works.”

As a result, although a huge range of vested interests (private or social) seek to invest in, influence or shun research in Computer Science, the sheer size of the discipline – combined with its conviction in its ethos – means even the most powerful players in the world struggle to do so.

Latour, Modernity and Society

Although the Mertonian nature of Computer Science allows us to paint a picture of the past and the present state of AI research, the next question on our minds is: what happens now? Our modern world prides itself on the compartmentalised nature of life – one where every function of society has a role, and together, like cogs in a machine, they create modern society as we know it. In this modern world, science is said to be an isolated field that deals with nature and non-human agents, separate from the realm of politics that deals with human affairs, and the realm of discourse that deals with ideas, thoughts and representation. Science has a clear and defined objective: to investigate the scientific truth of the world, and to wrestle it for humanity’s benefit.

Computer Science, for the better part of the past century, has been quietly and obediently involved in marching towards this goal and bringing to society the fruits of its labour – technology – which the businessmen and the bureaucrats from the political realm install and distribute to society in the form of infrastructure. However, with the rapid development and adoption of technology in the past few decades, our once cleanly organised society has been forced to adapt to a growingly decentralised and fast-evolving world. The lines of the various functions of society, once emboldened, are now blurred at best. After all, when today’s technology can easily become outdated tomorrow, society cannot afford to keep up the laborious organisation it once had. But where does this leave our Computer Scientists, who until now knew nothing but their work and its relation to society, but all of a sudden are held responsible for the development of the entire world?

In his book, We Have Never Been Modern, Bruno Latour (1993) explores this notion of a neatly divided ‘modern’ world – and how it is nothing but a myth. He explains that a modern world, considering itself as superior to the pre-moderns, does so by flaunting a modular society with detached, isolated and specialised functions that interact with each other. The moderns believe that conceiving such a separated yet functional society, made up of individual, independent and specialised domains that come together like neatly-fitting jigsaw pieces, requires a certain intellectual calibre acquired across the duration of multiple civilisations. This aligns with Merton’s idea that the emergence of Science, as a distinct, neutral, independent and self-sustaining field of truth-seeking, is only a trait of a high-order society.

While modernity insists that the remnants of such a classification, referred to by Latour as ‘hybrids’, are simply exceptions, corners and grey-areas, Latour argues that these hybrids are in fact symptoms and manifestations of the underlying interconnectedness of our supposedly modern society. Today, no bigger example of a hybrid may exist than Artificial Intelligence. Despite being a product of Science, today’s discussion revolves around its pervasive impact on politics (humans in a human setting, which it directly affects) as well as discourse (the very meaning of representation, language and symbolism – which it puts into question).

Latour remarks that understanding scientific knowledge is inseparable from understanding the societal context in which it occurs. Within Computer Science, this can be found in the motivations behind researchers for joining this field – while all of the interviewed professors recalled shifting to Computer Science during the technological boom during the late 20th century, all of the interviewed students recalled a long-time interest in Computer Science originating from a childhood spent during this technological boom. The social contexts in which these researchers emerged were essential in shaping the ‘why’ behind their research endeavours. At a larger scale, this can be observed with the 21st century phenomenon known as ‘The Fourth Industrial Revolution’ – in a world of globalised capitalism with limitless potential of technological development, the final objective appears to be upgrading human life by blurring the lines between the physical, digital and biological.

While working towards this goal, one professor observes, the features and functionalities of today’s products are direct responses to the needs and desires of yesterday, in what can be considered technological hedonism. With a motivation to automate as much human work as possible, this aligns with Latour’s notion, “Science is politics pursued by other means.” However, for a long time, our blind and mindless quest of development has kept us from asking whether or not this modern society is prepared to handle the technology it craves, and now it has reached a point where the answer is uncertain. The modern world lives comfortably in what Latour describes as a constitution of various dualisms: human – non-human, science – politics, nature – society. Artificial Intelligence, with all of its unforeseen and incomprehensible effects on society, not only lies outside of this constitution – it breaks it.

An interdisciplinary respect between the barriers of science and politics is not even the least to be done to solve this issue: as long as this division remains, we fail to address the existence of this problem in the first place. Learning the true impacts that technology can have on society begins with accepting modernity as a myth and destroying the antique notion of a separable field of science and politics. Furthermore, drawing ideas from Latour’s Actor-Network Theory and the Parliament of Things, we must be willing to accept and consider non-human entities (such as AI technologies) as valid and significant actants, with an agency of their own, and a tangible impact on society. Instead of looking at technology’s impact from a scientific perspective as we have been doing for so long, this involves looking at technology from a completely different, societal perspective – requiring us to first blur the line between science and politics.

But how do we get there? Well, our professors believe that the Orientals have the answer. The word ‘modern’ typically brings up pictures of the West, which sees such divisions in society as a reductionist concept. But for the Orientals, these divisions are not an attempt at asserting dominance over the vast range of functions in a society, but a way to submit to their authority. While Western perceptions of society are based on the notion of modernity, and usually work with a magnified look at the details, the Orientals, by culture and by tradition, tend to look at the bigger picture of things – to view individual pieces as parts of a whole. As one professor explains with an example, “even our languages are by nature very precise.” Mathematics in Ancient India was not separate from language, but rather expressed in Sanskrit itself; and one will find that for the word ‘love’ in English, Urdu has hundreds of different shades: ‘ishq’, ‘mohobbat’, ‘pyaar’, ‘junoon’ etc.

The classification of society, then, becomes just a way to observe the different facets of the same larger picture. As another professor adds, these divisions keep human arrogance in check, as a single mind cannot possibly grasp every detail of such a multifaceted world. What becomes important then is not to understand the inner workings of each of these different divisions, but to foster respect between them, and to understand how they interact, collaborate and impact each other. In order to reach a solution regarding AI, society does not need a dismissal of divisions entirely, but rather a dismissal of the notion that these divisions are true, absolute and natural. Only when policymakers learn to collaborate with scientists, and when scientists learn that they and their works are not in fact isolated from the larger society, will we be able to reach a common ground regarding Artificial Intelligence.

Conclusion, Social Responsibility and the Next Generation

As Merton breaks down the scientific ethos, he repeatedly emphasises that scientists are institutionally expected to be disinterested in any goal other than the pursuit of scientific knowledge. Disinterestedness (a norm, but not a law) claims that neither the personal rewards nor the social implications of a research work are to sway a scientist from pursuing or abandoning a given piece of research – the objective truth, no matter what it is, should be the priority. While this works great when science has limited interaction with the larger ‘modern’ society, a majority of anti-science movements occur when the impact of science on society is too large to still refer to science as isolated or separate. Merton notes that scientists, remaining adamant in their neutrality, argue that “an inadequate social structure has led to the perversion of his discoveries.”

As one professor puts it, blaming Computer Science for the destructive potential of their work is like “blaming metallurgy for the sword that kills, even though the same metallurgy gave us the surgeon’s needle.” He further adds, “the arguments of social utility should not be brought into computing in any way, as doing so WILL introduce a selection bias.” Merton argues that even though the purpose of science is to bring social benefit in the long run, the pursuit of science in no way guarantees that the outcome will be beneficial. The question he poses is: “can a good tree bring forth evil fruit?” Well, according to our professors, “this is not a scientific problem, it’s a human problem.”

The implications of science, despite concerning society due to its social impact, are still thought of scientifically, and rarely socially. The divisions of our supposed modernity have become so ingrained that it becomes hard for us to even view the effects of Artificial Intelligence as a social problem entirely – only as a scientific problem with social implications. Talking to Computer Science students about their views on social responsibility, they argue, “the potential bad should never hinder progress, but we should remain more mindful about how we use technology at a society-level.” Relating to the concepts of non-human agents, once we start looking at science as much more than just the research-wing of society, and at its products as not just tools but significant non-human actants that can influence society, only then can we begin to understand the true scale of our post-modern social problems.

Something both professors and students agree upon is the urgent need for a better overall Computer Science education, at both a national and a global level. Almost every college-level Computer Science education involves a significant amount of time teaching and training students in the domain of Artificial Intelligence, but next to none provide them with the opportunity to learn about their social implications. The interviewed students claim that an unrestricted, diverse and interdisciplinary liberal arts education has provided them with not just exposure to other fields, but also the opportunity to look at Computer Science from an outside perspective, giving them the chance to relate the two. As the professors claim, this holistic and liberal education puts such students significantly ahead of their mainstream counterparts in being able to drive change in not just Computer Science, but also in how it interacts with society.

The debates of today are new problems – too complex and pervasive to fit into one category – that require new ways of thinking to come up with new solutions. Although our society remains far too ingrained in its ways to break out of its primitively ‘modern’ worldview, it is by creating a responsible next generation of Computer Scientists – those who can look at society as a whole – that we may define new ways of thinking and dealing with these problems.

References

1. Future of Life Institute. 2023. “Pause Giant AI Experiments: An Open Letter - Future of Life Institute.” Future of Life Institute. Retrieved (https://futureoflife.org/open-letter/pause-giant-ai-experiments/).
2. Imbrie, Andrew, Owen Daniels, and Helen Toner. 2023. “Decoding Intentions: Artificial Intelligence and Costly Signals.” Center for Security and Emerging Technology. Retrieved (https://cset.georgetown.edu/publication/decoding-intentions/).
3. Kuhn, Thomas S. 1969. The Structure of Scientific Revolutions. University of Chicago Press.
4. Latour, Bruno. 1993. We Have Never Been Modern. Harvard University Press.
5. Lostri, Eugenia, Alan Z. Rozenshtein, and Chinmayi Sharma. 2023. “The Chaos at OpenAI Is a Death Knell for AI Self-Regulation.” Lawfare. Retrieved (https://www.lawfaremedia.org/article/the-chaos-at-openai-is-a-death-knell-for-ai-self-regulation).
6. Merton, Robert K. 1973. The Sociology of Science: Theoretical and Empirical Investigations. University of Chicago Press.
7. Metz, Cade, and Gregory Schmidt. 2023. “Elon Musk and Others Call for Pause on A.I., Citing ‘Risks to Society.’” The New York Times, March 29.
8. Pickering, Andrew. 1995. The Mangle of Practice: Time, Agency, and Science. University of Chicago Press.
9. Routley, Nick. 2017. “Visualizing the Trillion-Fold Increase in Computing Power.” Visual Capitalist. Retrieved (https://www.visualcapitalist.com/visualizing-trillion-fold-increase-computing-power/).
10. Turing, Alan. 1950. “I.—COMPUTING MACHINERY AND INTELLIGENCE.” Mind LIX(236):433–60. doi: 10.1093/mind/lix.236.433.