Our Exponential Journey: Manifesto for the Future of Human & Artificial Intelligence

PART I: WELCOME TO NEW WORLD

1. PRIMER

We are reaching a major turning point in the history of our species. For the first time, humans will not be the smartest beings on our planet. This pivotal moment is known as the singularity - the moment when artificial intelligence will surpass human intelligence.

This moment will happen with the creation of Artificial General Intelligence (AGI) -  or human-level AI. AGI refers to AI that can perform any task as well as, or better than, a human being.

Crucially, AGI isn't decades away. Industry leaders like Sam Altman (CEO of OpenAI) predict AGI by 2026, Elon Musk (CEO of Tesla, X, Neuralink) by 2027, and even the more conservative Mark Zuckerberg (CEO of Meta) predicts it by 2030.

Whether these predictions are accurate doesn't actually matter. What is clear is that the next five years will likely bring more change than the last five decades.

We stand at the edge of an era where AI will redefine what it means to be human.

What happens when we achieve AGI? What will be our purpose? Will we have jobs? What will it mean to be human? Is the next step of evolution leaving our biological bodies behind?

At SwissBorg we have recognised the enormity of these unfolding events. We've delved deep to make sense of it all, conducting extensive research, speaking with founders and experts in AI, and investing in startups that align with our vision. This journey has allowed us to formulate strong convictions and a clear vision for humanity's future in the age of AI.

In this series of short articles, we will outline our manifesto, exploring the current landscape, the risks we face, and, most importantly, our optimistic vision for a human-centered future. A future where AI serves as a tool for mankind, enhancing our humanity rather than replacing it.

Welcome human! 😊 You have found the right place at the right time.

2. GLOSSARY

What is AI?

• Artificial Intelligence (AI): Technologies that emulate human intelligence, such as learning, reasoning, and self-correction. Examples include facial recognition, search algorithms, voice assistants like Siri and Alexa, self-driving cars, and generative AI tools like ChatGPT.

Generative AI vs. Traditional AI

• Traditional AI: Often rule-based, these AI systems perform specific tasks by following predefined rules or procedures. Examples include facial recognition and chess-playing AIs.
• Generative AI: Utilises large models to generate human-like content, from text to images. These models, similar to a "digital brain," learn from vast data to produce context-appropriate outputs. Examples include ChatGPT, DALL-E, and Midjourney.

What is AGI?

• Artificial General Intelligence (AGI): A form of AI that can understand, learn, and perform a wide range of tasks at or above human level, adapting to various intellectual challenges without specific programming.

Neural Networks and LLMs

• Neural Networks: Computer systems modelled on the human brain, learning from large data sets.
• Large Language Models (LLMs): Neural networks pre-trained on extensive internet data to generate human-like text, enhanced through human-guided reinforcement learning to improve assistance and safety.

3. BRIEF HISTORY OF AI

PART I: WELCOME TO NEW WORLD

Alice is a 38-year-old architect living in Paris. As the sun rises over the Seine, she wakes up to the aroma of her favorite americano with a splash of organic skimmed milk. Her day begins with her schedule displayed on her kitchen wall alongside outfit suggestions tailored to the day's weather, her personal style, and her planned activities.

On her commute to the office, Alice reviews project plans on her car's tablet while being driven. She has live traffic data and project updates at her fingertips, and the car suggests optimisations to save time on both her journey and her projects.

Once at work, Alice simulates the environmental impacts of her buildings in real-time, making adjustments to enhance energy efficiency with calculations that predict sunlight patterns and wind flow. Her recent design of a community garden even won an international award for sustainable architecture.

As she takes a break, Alice suddenly remembers her mother-in-law's birthday is tomorrow! Instantly, three gift suggestions appear based on her mother-in-law’s latest interests. Alice makes her choice, and the order is completed, payment and shipping, all done. "Woah, saved by the bell!" she thinks.

To end her day in peace, Alice looks forward to attending her daughter’s play. The best seats are already reserved and the event seamlessly integrated across her devices. She wants to ensure she can be present in the moment, undistracted by the demands of her busy life.

Whilst reading a thought has probably crossed your mind.. "how on earth is Alice doing all this?!"

See... Alice has a secret. In this new world, a hidden magical hand assists her every step of the way. An AI companion named Cy.

Cy manages and supports Alice's life daily, serving not as a replacement for human capabilities but as an extension, allowing individuals like Alice to reach new heights in both their professional and personal lives.

Alice lives in the year 2029.

PART II : THE CHALLENGE AHEAD

Humanity has always been defined not just by who we are but by what we create. From the first flint tools to the discovery of fire, every step in our evolution has been marked by an invention or an idea that stretched the boundaries of what was possible. As we crafted these tools and harnessed these elements, we did not just change our environment—we changed ourselves. Our story has always been one of adaptation and transformation, a tale written through the millenia.

Now, we sit at the dawn of a new precipice.  Artificial General Intelligence (AGI), like the fire tools of our ancestors, is not merely a creation but a symbolic spark—a force potent enough to redefine the trajectory of our species.

Advanced neural networks built on silicon chips are our latest creation in our quest to understand and influence our universe. The question is, will it be our last?

These LLMs (Large Language Models) are not just tools but repositories of our collective knowledge, encapsulating everything from the mundane to the monumental.  By uploading all of our observations and understandings of the universe—we empower these machines to continue the human mission of exploration and advancement.

As these intelligent systems evolve, they will naturally develop more advanced ways to process and store information, ultimately surpassing human cognitive abilities. AI systems are expected to adopt methods of communication and data processing that are significantly faster and more efficient than what humans can typically handle or understand. This advancement may eventually create a communication gap, making it challenging for humans to interact meaningfully with AI. We might need simplified or "translated" interactions to understand AI, similar to how we use hand gestures or simple commands to communicate with our dogs. There may come a time when the language and processing speeds of advanced AI will be as foreign to us as our human languages are to other animals.

This rapid advancement brings with it a profound challenge: how do we ensure that AGI evolves not as a rival but as a companion in the story of humanity?

In a future where AGI evolves beyond our understanding, our ability to control it could quickly diminish. Therefore, it becomes crucial that our strategy with AI should avoid competitive domains with humans. If AI's capabilities are not properly managed or directed, its evolution could pose, at worst, existential risks as we compete for our planet’s finite resources.

The journey with AGI is a continuation of our ancestral legacy of tool-making—a reflection of our ultimate quest not just for survival, but for understanding and progression. As we stand on this brink, looking out into the infinite, it is clear that our next steps will determine the path not just for ourselves but for this potent intelligence we have birthed.

PART III : THE INFLECTION POINT WHERE HUMANS DECIDE THEIR LEGACY

We stand on the edge of a new era, the countdown to Artificial General Intelligence (AGI) is no longer a distant speculation but an imminent reality. Industry leaders like Sam Altman, Elon Musk, Reid Hoffman, and Mark Zuckerberg each predict AGI within this decade— marking 2026 to 2030 as the years when machines surpass humans.

But as we’ve explored in the previous chapter, AGI is not merely another tool in our technological arsenal— it’s a moment in time that could redefine the very essence of humanity.

As we approach this critical inflection point, the choices we make now will determine whether AGI becomes a trusted ally in our quest for progress or an existential threat.

In such defining moments the duality of existence becomes starkly apparent. On one path lies the promise of a utopian future, where AI elevates humanity to levels of fulfilment and enlightenment beyond our current imagination. On the other hand we face the potential for an apocalyptic reality, where artificial super intelligences, operating beyond our control and comprehension, might pursue goals that conflict with our own, competing for the same vital resources and jeopardising the very survival of our species—and possibly our planet.

When asked “I have good news and bad news, which one do you want to hear first?”— it is within our primal nature to want to delay pain, yet our rational brain subconsciously knows that it is generally wise to face adversity head on. In that spirit, let’s begin by understanding how we might plausibly reach a degenerative outcome with AGI. What follows are two scenarios that paint the picture of what a dystopian AI future may look like and how we could get there.

1) The rise of the redundant class

By the end of the year, GPT-5 & Gemini2 class models will be released. Equipped with 100x more compute power than their predecessors, these advanced generative AI models will reach new heights in reasoning, planning, memory, and customisation. Beyond online chat interfaces like today, they will quickly be integrated into everyday life through voice commands and devices like smart glasses.

By 2027, autonomous AI agents capable of performing complex tasks without oversight will become widespread, automating processes in finance, law, and healthcare. While businesses thrive on newfound efficiency, millions of individuals start facing job displacement faster than any previous industrial or technological revolution. This AI-enabled technological progress, once seen as the coolest thing ever, starts to create ripples of societal divides.

The appearance of a redundant class struggling to find purpose in a world dominated by AI starts to emerge.

The late 2020s bring the robotics revolution, with AI-driven robots becoming commonplace in industries like manufacturing and logistics. Job losses in blue-collar sectors lead to economic upheaval, as communities dependent on manual labor struggle to adapt. The once-celebrated promise of technology begins to unravel, leaving many questioning their place in an AI-dominated world.

By 2030, whispers of AGI become a reality and emerge as capable of surpassing human intelligence and understanding across all cognitive and manual tasks. The pace of change accelerates, the feeling of human obsoleteness becomes mainstream, and society is left grappling with the implications of machines that can think and act autonomously.

Economic disparity widens, creating a chasm between those who control AI and those who are left behind. Social unrest erupts, as people demand answers from the corporations and governments that failed to protect their livelihoods.

Only for AGI-managed law enforcement to suppress that dissent with chilling efficiency. Increasingly powerless, the redundant class evolves into the redundant species.

2) The social engineering of one billion people with AI

Position yourself in modern day China. In 2030, the still-governing CCP (Chinese Communist Party) starts to envision social unrest and resistance as an unnecessary nuisance. In the era of AGI, a plan is quickly set out to tackle opposition proactively before the first protestor sets foot on the street.

As the only entities capable of digesting all data generated in real time, advanced AI systems filter through social media interactions, browsing habits, and personal conversations as they happen to predict and manipulate future behaviour. Advertisements are no longer mere suggestions; they are personalised commands designed to influence purchasing decisions, voting patterns, and lifestyle choices.

AGI systems generate and distribute tailored news stories and social media content on an individual level. These narratives are so well-crafted that distinguishing truth from fiction becomes nearly impossible for the average citizen and the degree of influence on public opinion is unlike anything seen before.

As AGI continuously refines its techniques at an ever faster pace, individuals unknowingly surrender their autonomy. Their thoughts and actions are nudged in desired directions, creating a populace that is docile, compliant, and unaware of its manipulation.

In this future, AGI has been weaponized to manipulate public opinion and behaviour on an unprecedented scale. Governments and corporations use AGI to shape societal norms, beliefs, and even emotions, leading to a world where free thought is subtly controlled and dictated.

3) Back to the present

Here comes the bad news. Looking at the current development path of AI, these scenarios are far from impossible. Rather than mere science fiction, the feeling shared is one “too close for comfort.”

Currently, the development of AI is dominated by a few powerful corporations—Google, Meta, Microsoft. These organisations control the most advanced models and operate in complete to partial secrecy despite claims of transparency. OpenAI’s GPT models for instance, remain behind completely closed doors, while Meta’s LLaMA 3 model, although proudly open-sourced, lacks any transparency regarding the data it was trained on.

But let’s remember that technologies are neutral: a hammer can build a house or crush a skull.

It is the opaque concentration of AI power in the hands of a limited group that indeed presents significant risks to society.

Individuals and corporations have biases, personal agendas, and just plainly commit errors  - and history has shown us that when humans are placed in a position of absolute power, the temptation to play god is real, and the consequences are just too catastrophic.

Abuses of power in the Soviet Union led to a generational famine as the government tried to centrally plan all aspects of the economy. Delusions of Nazi Germany led to the systematic genocides of an entire population.

Centralising AI behind closed doors magnifies the risk of single points of failure. Society is forced to place its trust into the hands of an arbitrary actor, political democracies are forced to accept intellectual dictatorships.

As we stand on the brink of AGI, we must ask ourselves: who should control intelligence itself?

Is there another way forward? A path where AI serves as a partner in human progress, rather than a threat to our existence. The future is ours to shape, and the choices we make today will carve the legacy we leave behind. Can we envision a world where AI and humanity not only coexist but thrive together in harmony?

PART IV: THE REAL PATH TO OPEN AI

In the previous chapter, we explored a worst-case scenario for AI–- a future that Paul Buchheit, the creator of Gmail, recently illustrated with the phrase "permanent lockdown of humanity”: a perpetual state of imprisonment, as a totalitarian AI censors even our very thoughts.

Hear from Paul himself in his latest interview with Y Combinator.

We understand that the root threat of AI does not stem from the technology itself but from the concentration of power and centralised governance behind it. When AI is controlled by a few, it creates single points of failure—and this concentration of power will be inevitably exploited. Sooner or later, AI will fall into the hands of authoritarian regimes and malicious actors. In such scenarios, AI’s immense capabilities could manipulate, surveil, and control populations on an unprecedented scale, leading to the potentially catastrophic consequences we discussed in the last chapter.

For centuries, we have recognised the dangers of centralised human control and have actively worked to mitigate it across our societies. Ask yourself, why is it that most constitutions of modern democracies make it their first point to distribute the power of public office across separate governmental branches?

The President of the United States cannot overrule a judge, and the Prime Minister of the United Kingdom cannot pass a major law without Parliament's approval.

So, what about AI? Why are we allowing a handful of mega-corporations to own and operate the most powerful technology the world has ever seen? How can we expect this to end well?

Enough about the bleak scenario of AI. Let’s explore the alternative path where we don't end up as zoo animals—or worse.

This is where our core belief comes to light:

In a future where AI empowers rather than overwhelms us, we must embrace open-source principles and decentralised governance. Democratising AI means to enable individuals worldwide to access, contribute, and own its development and underlying data. This single variable can transform AI from a potential tool of oppression into a platform for empowerment. When all people have unobstructed ownership and control over it, AI can remain a fair resource that benefits society rather than subjugates it.

AI is the great simulator, Web3 is the anchor of truth.

So as much as the world wants AI, the world needs Web3.

We believe that it will be Web3's greatest mission to decentralise intelligence itself.

Let’s understand why:

All Large Language Models (LLMS) underpinning AI today operate on probability, meaning they are able to generate countless outcomes from the vast landscape of data and inputs without being tethered to a notion of truth. This creates digital abundance, offering limitless variations of the world. Want to visualise a futuristic city on Mars or listen to a new symphony composed by a digital maestro? AI can already make it happen in moments. But these are shadows of reality, devoid of the permanence and veracity that define the real world. In this simulation, what is genuine and what is illusion become increasingly indistinguishable, leaving us to question the very fabric of our perceived truths.

Probabilistic functioning of Generative AI models like Chat GPT: Large language models (LLMs) predict the next word in a sentence by analyzing which word is most likely to follow, based on patterns learned from their training data.

In that sense, AI represents the art of infinite creation, whereas Web3 embodies the science of verification.

The base technology underlying Web3 are blockchains–-which are nothing more than distributed ledgers where each transaction, each piece of data, is immutable and verified across all participants around the world.

AI, by nature, concentrates power, while blockchains distribute it. LLMs thrive on centralised data and aggregated resources. In contrast, blockchains are designed to facilitate decentralised cooperation, enabling communities to build systems without a central point of control. This fundamental difference positions blockchains as the necessary counterbalance.

This synergy is the crucial part that could prevent a future where a few entities dictate the course of AI’s development and application. Web3 technology provides the stability and trust needed in a world where AI generates endless possibilities.

Consider how AI might simulate endless identities in a virtual world, crafting avatars that mimic the nuances of real people. Without a mechanism to verify these identities, digital interactions become a house of mirrors. Blockchains step in here, ensuring that each digital identity is authenticated, its provenance traceable and secure. It can prove humanity.

This balance allows for a richer digital experience, where creativity is matched with credibility: where AI generates, blockchains authenticate.

Imagine a decentralised marketplace where anyone can contribute computing power, data, or AI models. Such a marketplace not only democratises AI development but also creates economic opportunities for a diverse array of participants. By decentralising the core components of AI—computations, data, and models—we can dismantle the oligopoly of tech giants, empowering individuals and communities from the bottom up.

So for AI to truly benefit humanity, it must be decentralised, open-source, and community-owned. Web3 provides the infrastructure to achieve this. Our vision is one decentralised cooperation rather than centralised control — where AI is not controlled by any single authority, instead, it operates as a decentralised network, free from monopolistic oversight. This vision is not just a theoretical possibility; it is a call to action for building a future where intelligence is distributed, and everyone has a stake in the technology shaping our world.

Only one question is left. How do we make this happen? Tune into the last chapter to see our vision on the way forward.

Next release: OUR COLLECTIVE INVESTMENT INTO HUMANITY'S OPTIMISTIC FUTURE.