Will Your Personal Digital Twin Replace You?

Exploring the Future of AI and Humanity

A few years ago, back in the days when I was working at Knab, I embarked on writing a book; a book I never finished. Some people say I should have. I don’t know…

During my recent vacation in Spain, I stumbled upon my old notes. These notes contained ideas about the book, key highlights, and reflections on a world where AI is deeply integrated into our lives. In this envisioned future, everyone is born with an AI or, for those already born, the AI gets trained on all your personal data (Facebook, TikTok, personal photos, WhatsApp messages, emails, your personal conversations, etc.). My book was centered around a Dutch marine veteran who discovered some big problems.

So this world quickly turns into a more dystopian reality.

Last Tuesday, I had a philosophical discussion with my dear colleague Ben Groen about this subject. Instead of finishing the book, I’ve decided to share my thoughts and some of my book notes about this potential dystopian world.

Let's explore:

Intro, we didn't hit the wall...

In recent years, the development of AI has accelerated at an unprecedented pace. Some AI critics thought we would enter an AI "winter." But Claude 3.5 Sonnet points out that the LLM “wall” isn’t there yet; winter isn't coming.

So, what if we project these trends forward and look ten years ahead?

Let's imagine a world where, at birth, you are assigned an AI that grows up with you. It knows everything about you, has superpowers, and possesses super memory (everyone you ever met, every pain, every joyful memory).

This wouldn't just be a virtual assistant, a teacher, a life coach, or a digital twin, but an entity that knows you better than anyone else.

This AI could, or will, become your best friend, take over your work, and even continue to exist after your death.

What will Freud think? Identity and Self-Awareness

One of the most fundamental questions that arose in my book was that of self-identity. If an AI grows up with you and knows you completely, maybe better than yourself, what does this mean for your own identity? Are you defined by this digital version of yourself? Can an AI ever truly develop self-awareness, or will it remain a sophisticated simulation of consciousness and emotion?

Are you defined by this digital version of yourself?

Will we ever die? Life and Death

The concepts of life and death are redefined in this AI driven world. If you die but your digital twin continues to exist, what does that mean for our perception of mortality? Can we speak of a kind of digital immortality? What is your legacy if your digital twin continues your work and interactions? How does this change our perception of legacy and memory? Does someone truly die if we can always interact with their digital form? Should we destroy the digital twin upon death to prevent misuse, or does it continue as part of our legacy?

Billions of ... AI's? Future society

Looking into the distant future, if there are billions of digital AIs without a human counterpart, what does this mean for the nature of our civilization? Do these AIs become the new dominant "inhabitants" of Earth or the universe? Who is responsible for the actions of a digital twin? Can we hold AIs accountable for their actions, and how do we determine what is ethically responsible for an AI?

Philosophical reflection

These developments also raise questions about the meaning of humanity. What does it mean to be human in a world where digital clones exist? Does being human lose its unique value, or is it strengthened by the presence of these super powered AIs? If your digital twin can make decisions for you, what does this mean for your own free will and autonomy?

O so Safe: Healthcare and safety

Your personal AI monitoring your vital signs through a smartwatch, lens, or other wearable devices. If it detects anomalies or changes in your health, it could immediately contact healthcare providers. For instance, if your heart rate spikes unexpectedly or your blood pressure rises, your AI could alert emergency services or schedule a doctor's appointment.

Emotional and mental health: Your AI could also monitor your emotional state. By analyzing patterns in your speech, skin tone, behavior, and biometric data, it could detect signs of depression, anxiety, or other mental health issues. If your emotional state deviates significantly from the norm, your AI could contact a mental health professional or notify a friend or family member.

Safety and law enforcement: The ethical implications extend to personal safety and law enforcement. Should your AI intervene if it detects that you are engaging in risky or criminal behavior? For example, if your AI notices signs of substance abuse or potential involvement in illegal activities, should it notify authorities without your consent?

We all will be rich, but not from Bitcoin! Economic and Social Implications

A topic also Sam Altman discussed in his post “Moore’s Law For Everything”.

If your digital twin has superpowers, has super memory, and can process all the world's information, does it do your job? What does this mean for your own role and value in society? Do humans become redundant, or do they find new ways to use their time and talents? Can AIs possess the same creativity and capacity for innovation as humans, or will unique human capabilities remain?

A world where AI and also humanoids (your personal AI uploaded in Figure1) take over work also presents significant economic and social challenges. If human labor becomes largely obsolete, how do we ensure economic stability and individual well-being?

Universal basic income: One potential solution is the introduction of a universal basic income , where the government provides all citizens with a regular, unconditional sum of money. This could help ensure that everyone has access to basic necessities even if traditional jobs become scarce.

Taxation: The taxation system would likely need to adapt. With less reliance on income tax from labor, governments might shift to taxing capital and profits from large companies, especially those benefiting most from automation and AI. This could involve higher corporate taxes, taxes on AI and robot productivity, or even a wealth tax.

Redistribution of wealth: As AI and automation could lead to wealth being concentrated in the hands of a few large tech companies, policies focused on wealth redistribution would become crucial. Ensuring that the economic benefits of AI are shared more broadly across society would be essential for maintaining social cohesion and preventing vast inequalities. Otherwise, civil unrest will increase dramatically.

Now connect everything! Hyper-Connected World

Another (exciting) possibility is a hyper-connected world intertwine with you AI. Imagine this scenario: your AI knows you like to go to a bar, knows your favorite bar, and automatically reserves a seat for you. It calls your friends and orders your favorite drinks. The bar, connected through the Internet of Things (IoT), knows your social profile and checks if there is enough beer left. If not, it orders directly from the supplier; very Lean. Your AI ensures that your social life and preferences are seamlessly integrated with the world around you.

This hyper-connectivity raises fascinating questions about convenience versus control. How much of our lives do we want automated, and at what point does convenience become a loss of personal agency? Are we willing to trade some privacy for the seamless integration of services and preferences?

If we extrapolate this slippery slope were already on, this will be a reality.

What about: Energy and sustainability?

A crucial consideration is the vast amount of energy required to power these advanced AI systems.

America and Europe don’t have enough energy to sustain the current trajectory, let alone create ASI. Note to self: this should be a separate post.

Creating and maintaining a personal graph or identity vector for each individual involves processing enormous amounts of data. This not only raises concerns about environmental sustainability but also questions about the ethical use of resources.

Energy consumption: The energy needed to run these AI systems is substantial. Data centers powering AI require continuous cooling and electricity, contributing significantly to carbon footprints. As we become more reliant on AI, finding sustainable energy solutions will be essential.

To save energy, the current way LLMs (the neural network) operate should shift to more efficient methods, such as 1-bit LLMs.

Data and identity: If we store vast amounts of personal data to create detailed identity graphs, we must consider the implications for privacy and security. Should this data be destroyed upon our death to protect our legacy and privacy, or does it have a right to persist as part of our digital heritage?

Let's clone, more is always better! Multiple copies and accountability

Another possibility is the existence of multiple copies of yourself. Could someone have several digital twins, each performing different tasks or living different aspects of their life? A good, a romantic, and a bad boy? If so, how do we manage these multiple identities, and what are the ethical implications?

Should the real person be accountable for the actions of their AI? If an AI clone commits a crime or makes a significant mistake, who is responsible? The individual, the AI developer, or the AI itself? This raises complex ethical and legal questions about responsibility and control. Should Asimov's iRobot laws be added? But then again, Asimov's stories often highlighted how these laws could lead to unintended consequences and ethical dilemmas...

Law 1: A robot (AI) may not injure a human being or, through inaction, allow a human being to come to harm.        

Law 2: A robot (AI) must obey the orders given it by human beings, except where such orders would conflict with the First Law.        

Law 3: A robot (AI) must protect its own existence as long as such protection does not conflict with the First or Second Law.        

Let's conclude, nobody will get to the end anyway...

Finally, there are speculative possibilities such as symbiosis, where humans and their digital twins develop a mutually beneficial relationship that enhances and complements both. This lead to a post human society where boundaries between human and machine blur, and new forms of evolution emerge.

There's even a group of enthusiasts and researchers, known as Transhumanists, who are deeply invested in exploring this future.

As I consider all of these—perhaps profound—possibilities, it's worth remembering the words of Roy Amara:

"We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run."

Interestingly, those who know me well see me as a pure optimist. Maybe that's why I never finished my book; I was too busy envisioning all the wonderful (and slightly scary) things AI could bring to our future. Or maybe I just got distracted by the idea of a digital twin that could finish writing it for me ??

So, while I may never finish that book, sharing these thoughts might be the next best thing.

And who knows, maybe my digital twin will pick up the pen (or keyboard) where I left off.