Artificial Intelligence : Reflections on the Time We Live In
Introduction
One of the remarkable advantages of growing older is one’s increased credibility in drawing historical analogies between what has been and what could be, either because you’ve personally experienced it, or you’ve lived through the immediate aftermath.
For me, the current rise and evolution of AI technology echo the development of the internet in the 1990s. Both eras are marked by rapid technological progress, significant investment, and vast societal impacts. Simultaneously, these periods are underscored by considerable uncertainty about the future, and issues related to regulation, privacy, security, and a digital divide.
However, this historical parallel with the 1990s is only part of what is happening today. For the first time, we’re discussing the potential for software to pose an existential threat to humanity. The last time humanity grappled with a situation like this was in 1945, with the development of the atomic bomb — perhaps the first existential threat deliberately brought into being by humans themselves.
What Was Foreseen in 1993
When the World Wide Web emerged, one of the initial purchases I made in 1994 was a book titled Navigating the Internet by Mark Gibbs and Richard Smith, first published in 1993. In this book, under the subheading “Where Is the Internet Going,” the following suggestions were made:
- Education: The internet is a fantastic educational resource for both students and teachers.
- Library Issues: Access to library information from school, home and work is rapidly changing the way people think of information — from what is available in one local library to what is available nationally and even internationally.
- Student Access: In many universities, students staff and faculty can gain access to the internet. This is having a profound effect on education and the way we teach and learn.
- Science and Research: The internet distributes information in a way that is infinitely more flexible and timely.
What Really Happened
Now, let’s delve into a few areas where the impact of the internet was underestimated in the predictions made in 1993:
- Social Interactions and Relationships: The rise of social media platforms like Facebook, Twitter, and Instagram was not foreseen. The idea that people could build relationships, sometimes significant ones, without ever meeting in person was not something anyone would have envisaged.
- E-commerce: Online shopping has fundamentally altered the retail landscape, from the consumer perspective as well as for businesses. The sheer scale and reach of online retail platforms such as Amazon were beyond early predictions. Not only has it changed purchasing behaviours, but it has also affected supply chains, logistics, and even global trade.
- Remote Work and Education: While it was predicted that the internet would impact work and education, the extent of this impact was underestimated. The concept of telecommuting, remote learning, and the ability to operate entire businesses digitally have become prevalent.
- Entertainment and Media: No one fully foresaw the transformative impact of the internet on entertainment, particularly with the advent of streaming services like Netflix, YouTube, and Spotify, was not fully foreseen. Traditional forms of media such as print and broadcast have been greatly disrupted, leading to shifts in journalism, advertising, and content creation.
- Privacy and Security: Early discussions about the internet didn’t fully anticipate the significant concerns around privacy and security that would emerge. Issues like data breaches, identity theft, cybercrime, and the ethical handling of personal data have become major societal concerns.
- Misinformation and Propaganda: The role of the internet in the spread of misinformation and the manipulation of public opinion was largely unforeseen, not to mention that it would begin to undermine democratic institutions like free and fair elections.
It is clear that these kinds of changes are typically not predictable based solely on past precedents. In all these domains, the alterations brought about by the internet were not simply additive; rather, they were transformative, establishing new methods of performing tasks.
It’s 2023 and does anyone really think that the predictions we make today about the future of AI are going to be any more accurate than those made about the internet in 1993 ?
One Historical Analogy Is Not Enough
Is that all there is to it? Are we merely reliving the heady days of the 1990s? While this analogy might appear wholly satisfying, no one at the time regarded the introduction of the World Wide Web into our lives as an existential threat to humanity. For such a concept, we must go back to 1945.
The Manhattan Project, launched during World War II, was a highly confidential scientific and engineering initiative focused on developing the atomic bomb. Spearheaded by the United States, with contributions from the United Kingdom and Canada, the project assembled some of the world’s top scientists, including J. Robert Oppenheimer.
We can learn a great deal about that time from a speech given by Robert Oppenheimer. His full speech can be found in the link below.
This is his farewell speech to the Association of Los Alamos Scientists on November 2, 1945. In this speech he spoke about the challenges scientists and the world faced now that atomic weapons were a reality. Rather than apologize, Oppenheimer justified pursuit of an atomic bomb as inevitable, stressing that scientists must expand humankind’s understanding and control of nature. He also argued that new approaches were needed to govern atomic energy.
In this speech, Oppenheimer discusses the profound impact of the development of atomic weapons and the new age it has ushered in. He highlights the need for scientists to confront the issues arising from their work and emphasizes the importance of honesty and insight in facing the challenges ahead.
So what are the parallels? Both atomic weapons and artificial intelligence have brought about significant changes that necessitate deep reflection and consideration. They have raised questions about the relationship between science and society, the limits of human understanding, and the potential risks and benefits of technological advancements.
In his speech Oppenheimer emphasizes the speed at which atomic weapons were developed and their potential to affect all of humanity. Similarly, the rapid advancement of AI and its integration into various aspects of our lives has had far-reaching consequences. To quote from Oppenheimer’s speech:
I think that it hardly needs to be said why the impact is so strong. There are three reasons: one is the extraordinary speed with which things which were right on the frontier of science were translated into terms where they affected many living people, and potentially all people. Another is the fact, quite accidental in many ways, and connected with the speed, that scientists themselves played such a large part, not merely in providing the foundation for atomic weapons, but in actually making them. In this we are certainly closer to it than any other group. The third is that the thing we made — partly because of the technical nature of the problem, partly because we worked hard, partly because we had good breaks — really arrived in the world with such a shattering reality and suddenness that there was no opportunity for the edges to be worn off.
Oppenheimer even calls for open collaboration and the sharing of knowledge, a topic that is the subject of intense debate today with respect to AI:
I think that the talk has been justified, and that the almost unanimous resistance of scientists to the imposition of control and secrecy is a justified position, but I think that the reason for it may lie a little deeper. I think that it comes from the fact that secrecy strikes at the very root of what science is, and what it is for. It is not possible to be a scientist unless you believe that it is good to learn. It is not good to be a scientist, and it is not possible, unless you think that it is of the highest value to share your knowledge, to share it with anyone who is interested. It is not possible to be a scientist unless you believe that the knowledge of the world, and the power which this gives, is a thing which is of intrinsic value to humanity, and that you are using it to help in the spread of knowledge, and are willing to take the consequences.
Finally, Oppenheimer’s speech highlights the urgency for us all to begin to grapple with the consequences of the changes that have been unleashed on the world and that it will take time.
As I have said, I had for a long time the feeling of the most extreme urgency, and I think maybe there was something right about that. There was a period immediately after the first use of the bomb when it seemed most natural that a clear statement of policy, and the initial steps of implementing it, should have been made; and it would be wrong for me not to admit that something may have been lost, and that there may be tragedy in that loss. But I think the plain fact is that in the actual world, and with the actual people in it, it has taken time, and it may take longer, to understand what this is all about.
While the parallels are evident, the rise of AI presents a distinct scenario. The challenges are marked by rapid advances in AI, sparking worries about job displacement, privacy, algorithmic bias, and potential effects on societal structures. The issues raised by AI are multi-faceted and extend beyond anything as simple as the threat of immediate destruction.
Summary
One thing is certain: our collective capacity to predict the future in the face of rapid and large-scale technological changes has likely not improved. The future, undoubtedly, will transcend anything we can envision today. It always does.
If we can’t foresee the future, then we must brace ourselves for swift transformations in how we perceive and engage with virtually everything around us, as everything is poised to change, much like it did following the advent of the internet.
More than this, few today would contest the potential of AI to present a unique existential threat to humanity, as if the climate crisis weren’t already a significant concern. However, this is a scenario we’ve faced before. Over time, the world heeded Oppenheimer’s call; the Atomic Energy Act of 1946 resulted in the creation of the Atomic Energy Commission, which was given authority over all aspects of nuclear power. In response to the deep-seated fears and expectations stirred by the discoveries and diverse applications of nuclear technology, the International Atomic Energy Agency was established in 1957.
If we take history as a guide, it teaches us that the relationship between society and technology is not a one-way street. It’s not just about how technology transforms society, but also about how society shapes technological development. The latter tends to be undervalued in our predictions.