Humanity stands at a crossroads. The recent breakthrough in artificial intelligence, specifically Large Language Models (LLM) trained on sufficiently large and curated datasets, has enabled natural conversation between humans and computers. Moreover, since they are trained with datasets that encapsulates human knowledge, LLMs also have access to an expansive array of information across various fields. With this breakthrough and the pace of innovation that is ongoing, people are now speculating whether the advent of Artificial General Intelligence (AGI)
Throughout history, humans have demonstrated a natural inclination to delegate tasks and responsibilities. This practice has enabled us to enhance efficiency and productivity, and it remains an integral part of human nature. As technology advances, it is now possible to delegate those tasks to a bare chunk of metal instead of a living human being.
Although it is going to be harsh to say that it is good for the society because surely it will remove many jobs, but at least with machine we could just pressure a bare metal instead of a living human being with a beautiful soul. From this point of time, humans will be required to befriend technology as part of their life with the reason they could do everything much faster than us. But remember, they could only do well in a specifically given order. Jobs that requires many manual labor and muscle would not be replaced easily, such as building a home, installing electronics appliance, etc. Also jobs that required human affections such as teachers, nurse, and doctors would never ever be replaced.
While machines and AI can enhance productivity and efficiency, they cannot replicate the essence of human creativity. Human beings possess a unique capacity to create something from nothing. The journey of creation, the process of going from zero to one, encompasses the depth of human experiences and aspirations. As we delegate routine tasks to machines, we gain the freedom of time.
However, after all these things we discuss earlier, the bigger question still remains.
If we finally had our free time, what will we do with it? Is freedom itself could become the bare necessities required for us to be happy? Or the journey to achieve itself is already enough? The quest for meaning to resolve one of life biggest question will still be intact. But then again, any attempts to make the world a better place or to make people's life easier even in just the slightest manner should always be appreciated and hold into the highest regards.
Thank You for sharing the thought-provoking questions @Audi Pratama, I have some interesting follow-ups for you.
1. I'm very much aligned with your statement "..they could only do well in a specifically given order..". There's already a paper that's discuss about this: https://arxiv.org/pdf/2212.09251.pdf - you might find the section "Evaluating Sandbagging" interesting
2. "...they cannot replicate the essence of human creativity..." very much aligned as well, I'm thinking content that is fully generated by human (validated) will increase in value over time.
3. "If we finally had our free time, what will we do with it?" - Yes, this is the interesting part. My proposal was to pair the "innate drive to achieve something" (urgency) as goal along with freedom.
If you watch star trek then there's hypothetical scenario in the future where money is not required anymore because there's a technology that could generate any (most?) materials at will and they have the capability to harness enough energy to run the society. Therefore people choose to work following their passion instead with self-actualization as probably the highest motivation.
TL;DR: The distinctions between goals/values, needs, and resources are essential, and freedom itself isn't value-agnostic. Economic value definitions could fundamentally change with the automation of human capital.
I think it's crucial to differentiate between inherent non-negotiable needs (the essentials required for survival) and purpose (arbitrary values or goals one wishes to achieve). Additionally, resources (imo) can be defined as necessary and/or sufficient means to approach these goals or values. While these concepts (needs, purpose, resources) are distinct, they're not mutually exclusive. In Indonesia, where I presume both of us reside, poverty, malnutrition, water scarcity, and limited internet access are pressing issues that deserve our attention. Furthermore, global financial inefficiencies and impending ecological collapse warrant our concern.
In my view, "time" alone isn't a resource; attention (not used for earning a living) might be a more fitting concept. Finances or money serve as **convergence instrumental goals** for almost any **terminal goal**. As instrumental goals, expertise, health, and access may can be gained through money. We can narrow down generic resources to money and attention. Assuming that we require 8 hours of free time and 10 million Rupiah per month (basic living cost in Indonesia) to pursue any arbitrary goal in life, up to this level, i think it's essential to determine our life's purpose rigorously.
Research suggests a strong correlation between life satisfaction and income per capita. However, when our needs are met, we might blindly follow our innate biological needs without restraint keep running in hedonic treadmill cause habituation and compare to other (https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1010316). Addressing the problem of axiology in life is of utmost importance.
Regarding your concept of freedom, i guess it's difficult to bootstrap the notion of freedom and treat it as a value without a clear understanding of goals/values. it is like, Establishing an action space as a reward function along with a few hidden specific values/goals (in reinforcement learning). i don't think treating freedom as a "proxy goal for a better life" make problems any concreate.
It's also worth mentioning that the inevitable ecological collapse means that using LLMs to boost productivity isn't a standalone solution. like, single gpt-3.5-turbo response alone can cost 1000x than single google search. Even with access to the most advanced GPTs, the value of one's work or enterprise cannot exceed the market's upper limit unless market manipulation occurs. Moreover, the book **Bullshit Jobs** addresses the existence of jobs that add no value to the economy.
In my opinion, we live in fascinating times. All I need is decent internet access and basic living costs for the rest of my life. The climate crisis is a looming threat, and the earth is becoming greener due to trapped carbon for millions of years. We're living in an intriguing era, by any timescale of astronomy, geology, biology, and human history.
for me, the most worth goal to pursue is to understand my place in this universe and attempt to hack my reward function (see lebowski theorem). I see AI as a project to test how our minds truly function. There's nothing magical about the brain; we just need a few groundbreaking paradigms to develop systems similar to our minds. This is what I live for.
Thank You for sharing your comprehensive thoughts @Muhammad Rizkillah, I need some more time to digest but this is very valuable consideration for my future articles.
One thing I can say for now is that, I agree with you that freedom on its own doesn't make sense as a goal. Hence why my proposal is to put our innate drive to achieve something (urgency) on center stage as well. It's a bit abstract but I want to explore more on this while also reflecting on what you've shared.
And I fully agree that LLM is not a silver bullet, it is but one spark of innovation in the past 10 years. But we have to agree that historically human boost productivity by making tools, and those tools and/or approaches of using the tools are exactly what I want to explore (from either technology, social or philosophical perspective).
As someone who grew up in poverty, and is currently working to help the masses get jobs, inequality is at the forefront of my mind right now.
I think AI, in the long term, has the tremendous power to lift humanity's burden by taking on work that we humans don't like to do.
However, in the short-term, this may replace hundreds of millions of human workers who depend on the job to make a living. Losing their jobs can mean life and death to them and their families.
Yes, "technology distribution" is exactly my concern, which is also confirmed by the OECD research shown in the article that says that more advanced firms outperform all other firms in terms of productivity because they can adopt emerging technology, quickly.
I do not exactly have answer right now on how disruptive the situation will be or how it can be addressed, but that's exactly one topic of my exploration in this newsletter.
Throughout history, humans have demonstrated a natural inclination to delegate tasks and responsibilities. This practice has enabled us to enhance efficiency and productivity, and it remains an integral part of human nature. As technology advances, it is now possible to delegate those tasks to a bare chunk of metal instead of a living human being.
Although it is going to be harsh to say that it is good for the society because surely it will remove many jobs, but at least with machine we could just pressure a bare metal instead of a living human being with a beautiful soul. From this point of time, humans will be required to befriend technology as part of their life with the reason they could do everything much faster than us. But remember, they could only do well in a specifically given order. Jobs that requires many manual labor and muscle would not be replaced easily, such as building a home, installing electronics appliance, etc. Also jobs that required human affections such as teachers, nurse, and doctors would never ever be replaced.
While machines and AI can enhance productivity and efficiency, they cannot replicate the essence of human creativity. Human beings possess a unique capacity to create something from nothing. The journey of creation, the process of going from zero to one, encompasses the depth of human experiences and aspirations. As we delegate routine tasks to machines, we gain the freedom of time.
However, after all these things we discuss earlier, the bigger question still remains.
If we finally had our free time, what will we do with it? Is freedom itself could become the bare necessities required for us to be happy? Or the journey to achieve itself is already enough? The quest for meaning to resolve one of life biggest question will still be intact. But then again, any attempts to make the world a better place or to make people's life easier even in just the slightest manner should always be appreciated and hold into the highest regards.
Thank You for sharing the thought-provoking questions @Audi Pratama, I have some interesting follow-ups for you.
1. I'm very much aligned with your statement "..they could only do well in a specifically given order..". There's already a paper that's discuss about this: https://arxiv.org/pdf/2212.09251.pdf - you might find the section "Evaluating Sandbagging" interesting
2. "...they cannot replicate the essence of human creativity..." very much aligned as well, I'm thinking content that is fully generated by human (validated) will increase in value over time.
3. "If we finally had our free time, what will we do with it?" - Yes, this is the interesting part. My proposal was to pair the "innate drive to achieve something" (urgency) as goal along with freedom.
If you watch star trek then there's hypothetical scenario in the future where money is not required anymore because there's a technology that could generate any (most?) materials at will and they have the capability to harness enough energy to run the society. Therefore people choose to work following their passion instead with self-actualization as probably the highest motivation.
TL;DR: The distinctions between goals/values, needs, and resources are essential, and freedom itself isn't value-agnostic. Economic value definitions could fundamentally change with the automation of human capital.
I think it's crucial to differentiate between inherent non-negotiable needs (the essentials required for survival) and purpose (arbitrary values or goals one wishes to achieve). Additionally, resources (imo) can be defined as necessary and/or sufficient means to approach these goals or values. While these concepts (needs, purpose, resources) are distinct, they're not mutually exclusive. In Indonesia, where I presume both of us reside, poverty, malnutrition, water scarcity, and limited internet access are pressing issues that deserve our attention. Furthermore, global financial inefficiencies and impending ecological collapse warrant our concern.
In my view, "time" alone isn't a resource; attention (not used for earning a living) might be a more fitting concept. Finances or money serve as **convergence instrumental goals** for almost any **terminal goal**. As instrumental goals, expertise, health, and access may can be gained through money. We can narrow down generic resources to money and attention. Assuming that we require 8 hours of free time and 10 million Rupiah per month (basic living cost in Indonesia) to pursue any arbitrary goal in life, up to this level, i think it's essential to determine our life's purpose rigorously.
Research suggests a strong correlation between life satisfaction and income per capita. However, when our needs are met, we might blindly follow our innate biological needs without restraint keep running in hedonic treadmill cause habituation and compare to other (https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1010316). Addressing the problem of axiology in life is of utmost importance.
Regarding your concept of freedom, i guess it's difficult to bootstrap the notion of freedom and treat it as a value without a clear understanding of goals/values. it is like, Establishing an action space as a reward function along with a few hidden specific values/goals (in reinforcement learning). i don't think treating freedom as a "proxy goal for a better life" make problems any concreate.
It's also worth mentioning that the inevitable ecological collapse means that using LLMs to boost productivity isn't a standalone solution. like, single gpt-3.5-turbo response alone can cost 1000x than single google search. Even with access to the most advanced GPTs, the value of one's work or enterprise cannot exceed the market's upper limit unless market manipulation occurs. Moreover, the book **Bullshit Jobs** addresses the existence of jobs that add no value to the economy.
In my opinion, we live in fascinating times. All I need is decent internet access and basic living costs for the rest of my life. The climate crisis is a looming threat, and the earth is becoming greener due to trapped carbon for millions of years. We're living in an intriguing era, by any timescale of astronomy, geology, biology, and human history.
for me, the most worth goal to pursue is to understand my place in this universe and attempt to hack my reward function (see lebowski theorem). I see AI as a project to test how our minds truly function. There's nothing magical about the brain; we just need a few groundbreaking paradigms to develop systems similar to our minds. This is what I live for.
Thank You for sharing your comprehensive thoughts @Muhammad Rizkillah, I need some more time to digest but this is very valuable consideration for my future articles.
One thing I can say for now is that, I agree with you that freedom on its own doesn't make sense as a goal. Hence why my proposal is to put our innate drive to achieve something (urgency) on center stage as well. It's a bit abstract but I want to explore more on this while also reflecting on what you've shared.
And I fully agree that LLM is not a silver bullet, it is but one spark of innovation in the past 10 years. But we have to agree that historically human boost productivity by making tools, and those tools and/or approaches of using the tools are exactly what I want to explore (from either technology, social or philosophical perspective).
Thank you for the insightful article Gio.
As someone who grew up in poverty, and is currently working to help the masses get jobs, inequality is at the forefront of my mind right now.
I think AI, in the long term, has the tremendous power to lift humanity's burden by taking on work that we humans don't like to do.
However, in the short-term, this may replace hundreds of millions of human workers who depend on the job to make a living. Losing their jobs can mean life and death to them and their families.
What do you think about this?
Yes, "technology distribution" is exactly my concern, which is also confirmed by the OECD research shown in the article that says that more advanced firms outperform all other firms in terms of productivity because they can adopt emerging technology, quickly.
I do not exactly have answer right now on how disruptive the situation will be or how it can be addressed, but that's exactly one topic of my exploration in this newsletter.