Future of Artificial Intelligence
Re: Future of Artificial Intelligence
The process really highlights how low information density is in a podcast. I largely stopped listening to them, because of that.
One use of AI in a corporate setting, is generating meeting notes. It does ok, but there's a lot of implicit context the AI is blind to. That makes increasing information density (similar to a podcast to article) tough.
From the AI's perspective, text vs. speech isn't dramatically different. The input/output factors are well understood problems. It's a question of what's in the tokenized input, and how well the model can transform it the new desired set of tokens.
I found this book to be a reasonably accessible overview:
https://www.manning.com/books/how-gpt-w ... =serp_auto
It has me much less concerned about a true runaway AI, while recognizing everything is going to change.
One use of AI in a corporate setting, is generating meeting notes. It does ok, but there's a lot of implicit context the AI is blind to. That makes increasing information density (similar to a podcast to article) tough.
From the AI's perspective, text vs. speech isn't dramatically different. The input/output factors are well understood problems. It's a question of what's in the tokenized input, and how well the model can transform it the new desired set of tokens.
I found this book to be a reasonably accessible overview:
https://www.manning.com/books/how-gpt-w ... =serp_auto
It has me much less concerned about a true runaway AI, while recognizing everything is going to change.
Re: Future of Artificial Intelligence
Here is a briefing doc generated by NotebookLM for your two Stoa youtube videos. It can also generate a study guide, FAQs, a table of contents and a timeline in addition to the podcast highlighted earlier. Could be useful for someone who does not want to watch a few hours of video.jacob wrote: ↑Sun Nov 03, 2024 11:42 amA simple test to whether the asymmetry holds is whether there are also tools that can generate an article from three conversations, now that we've seen it the other way around. I still doubt that because to me written posts contain substantially more information as well as more complications than conversations.
Briefing Doc: Resilient Lifestyle Design and Solving the Meta-Crisis
Main Themes:
Transitioning from consumerist to post-consumerist lifestyles: This involves shifting from dependence on external systems (jobs, markets) to developing self-reliance and resilience through diverse skillsets and interconnected resource flows.
Systems thinking: This approach emphasizes understanding the interconnectedness of different aspects of life and designing solutions that address multiple goals simultaneously.
Emergent movements: Individual change, when aligned with actionable plans and diverse motivations, can lead to broader societal shifts.
Addressing the meta-crisis: This refers to tackling the complex, interconnected global challenges of our time, such as climate change, economic instability, and societal fragmentation.
Most Important Ideas/Facts:
From "A Systems Approach to Resilient Lifestyle Design w/ Jacob Lund Fisker":
Lifestyle Design: A deliberate process of shaping one's life and way of living, moving beyond societal defaults.
The Salary-Consumer Model: This dominant model emphasizes specialized jobs and consumption-driven spending, leading to fragility and waste.
Skill Development: Transitioning to a resilient lifestyle requires developing diverse skills beyond specialized employment, enabling greater self-reliance and reduced dependence on external systems.
The "Diamond Model": This model represents an interconnected system where resources (time, skills, materials, etc.) flow within a closed loop, minimizing waste and maximizing efficiency.
Plato's Cave Allegory: This allegory serves as a metaphor for moving beyond the limitations of consumerist thinking and embracing new ways of perceiving and interacting with the world.
Quotes:
"The money will sort of follow from [a resilient lifestyle] automatically."
"Anything that's like super specialized here is is not resilient."
"You can never do just one thing. Whenever you do some activity that always has some side effects, some waste."
"The attitude towards spending [in a post-consumer system] is also something that has changed significantly... Now you see it more as more as a kind of friction, as a sign of... a bad philosophy in life."
From "Resolving the Meta-Crisis With Emergent Movements and Post-Consumerist Praxis w/ Jacob Lund Fisker":
The Meta-Crisis as a Failure of Scope: People tend to focus on short-term, local concerns, making it difficult to address large-scale, long-term problems.
Individual vs. Societal Solutions: Both individual and societal approaches are necessary to address the meta-crisis, but often struggle due to limitations in scope and implementation challenges.
The Importance of Transdisciplinary Knowledge: Developing expertise in multiple, diverse fields allows individuals to connect seemingly disparate ideas and create novel solutions.
Designing for Emergent Movements: Effective solutions must consider human limitations, provide individual benefits even without widespread adoption, and offer a clear path for transition from the present to a desired future.
Ethical Considerations: Solutions should be honest, sustainable, and respect individual agency, avoiding unintended negative consequences.
Quotes:
"Resolving the meta-crisis... requires... considering [the personal, the societal, and the terrain] simultaneously."
"Human knowledge has been divided and conquered... by the system that sits on top of us."
"[We need] to make more meaning-making errors and... become jack of more traits and mastering some of them."
"[The plan] has to work for a specific individual within the current system... [and] for all individuals if everybody does it."
"[Solutions] have to respect human limits and maintain solution integrity even when it's misinterpreted."
Key Takeaways:
Building a resilient lifestyle requires a shift in mindset and skills, focusing on self-reliance, interconnectedness, and long-term thinking.
Individual solutions, when designed strategically, can contribute to emergent movements that address larger societal problems.
Solving the meta-crisis demands collaborative efforts that bridge the gap between individual and societal approaches, while respecting human limitations and ethical considerations.
Further Research:
Explore Jacob Lund Fisker's blog and forum for further insights into resilient lifestyle design and the FIRE movement.
Investigate concepts such as "systems thinking," "transdisciplinary knowledge," and "emergent movements" for a deeper understanding of the proposed solutions.
Research organizations and initiatives working on specific meta-crisis issues to explore potential avenues for involvement.
-
- Site Admin
- Posts: 17167
- Joined: Fri Jun 28, 2013 8:38 pm
- Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 77
- Contact:
Re: Future of Artificial Intelligence
The thing is I'd like to see an article. Figure that I was talking for an hour at 100WPM then that's 6000 words of raw material for each presentation, so I'm generously looking for an article that's 1500 words for a 4x reduction that convincingly argues the point I was making. If generative AI can do an adequate job of that compared to what I am able to write myself, it's time to hang up my [writing] gloves.Ego wrote: ↑Sun Nov 03, 2024 1:40 pmHere is a briefing doc generated by NotebookLM for your two Stoa youtube videos. It can also generate a study guide, FAQs, a table of contents and a timeline in addition to the podcast highlighted earlier. Could be useful for someone who does not want to watch a few hours of video.
While AI typically gets the subject and the resulting words right, the main give away of AI writing tends to be that it lacks depth (it's shallow) in terms of the skeleton of the argument and double down by filling up the space with platitudes. It's like listening to a sales rep or a politicians talking in circles w/o saying much when asked a question they don't know the answer to. In short, the actual information density is low. There's a lack of "surprising" details like a high school essay demonstrating no actual [life-]experience about the topic they've been forced to write about.
Re: Future of Artificial Intelligence
I had a lazy go at it:
A Systems Approach to Resilient Lifestyle Design: Beyond Consumerism
In an era marked by escalating environmental crises, economic instability, and societal upheavals, the need for resilient lifestyle design has never been more pressing. Jacob Lund Fisker, author of Early Retirement Extreme, presents a compelling argument for adopting a systems approach to lifestyle design that emphasizes resilience, self-reliance, and sustainability. By critically examining our entrenched consumerist habits and embracing systems thinking, individuals can not only achieve financial independence but also contribute to a more robust and adaptable society.
The Salary Consumer Paradigm: A Fragile Foundation
The traditional consumerist model, which Fisker terms the "salary consumer" paradigm, is predicated on a linear and compartmentalized approach to living. Individuals typically specialize in a narrow set of skills to perform a job that yields monetary compensation. This income is then exchanged for goods and services, often without consideration of the interconnectedness of various aspects of life or the broader environmental and societal impacts.
This model is inherently fragile due to its heavy reliance on external systems for basic needs. The COVID-19 pandemic starkly illustrated this vulnerability as supply chain disruptions led to shortages and individuals found themselves ill-equipped to meet their own needs without external support. The salary consumer is bound by a cycle of earning and spending that leaves little room for adaptability or resilience in the face of systemic shocks.
Transitioning to Advanced and Deliberate Consumption
Recognizing the limitations of the salary consumer model, some individuals progress to what Fisker describes as the "advanced salary consumer" stage. In this phase, people begin to optimize their finances through budgeting, saving, and strategic spending. They might invest in financial instruments and start to consider the long-term implications of their economic choices. However, their lifestyle remains fundamentally tied to the consumerist framework, and they continue to depend on external systems for most of their needs.
Further along this spectrum is the "deliberate consumer." Individuals at this stage start to develop a broader skill set and seek to obtain different yields beyond mere financial gain. They might engage in activities like gardening, repairing their own goods, or developing side gigs that provide alternative sources of income or value. Their approach to spending shifts from consumption to investment in their own capabilities and resources.
Post-Consumer Systems and the Power of Systems Thinking
The culmination of this progression is the adoption of a "post-consumer system," wherein individuals fully integrate systems thinking into their lifestyle design. Systems thinking involves understanding the complex interdependencies between different components of one's life and optimizing them for resilience and sustainability.
In a post-consumer system, activities are interconnected in a way that each action serves multiple purposes and contributes to the overall resilience of the individual’s lifestyle. For example, gardening not only provides food but also contributes to physical health, reduces grocery expenses, and can even serve as a social activity. This integrated approach reduces waste, lowers dependency on external systems, and enhances the capacity to adapt to changing circumstances.
Fisker introduces the concept of the "web of goals," a multidimensional network where various life activities support and reinforce each other. This web creates a resilient structure that can withstand disruptions because the failure of one component does not collapse the entire system.
Skill Development and the Learning Curve
A critical aspect of transitioning to a resilient lifestyle is the development of a broad skill base. Fisker outlines a learning curve that progresses from imitation to innovation:
1. Copying: Imitating existing models or instructions without understanding the underlying principles.
2. Comparing: Evaluating different models to understand their relative merits.
3. Compiling: Gathering various methods and strategies to create a personalized approach.
4. Computing: Applying knowledge contextually to solve problems and make decisions.
5. Coordinating: Integrating multiple skills and knowledge areas to achieve complex goals.
6. Creating: Innovating new solutions and systems.
By advancing along this curve, individuals become more capable of addressing their needs without relying solely on external services. This not only leads to financial savings but also enhances personal autonomy and adaptability.
Philosophical Perspectives: Escaping Plato's Cave
Fisker employs Plato's Allegory of the Cave as a metaphor for the consumerist paradigm. In the allegory, prisoners are confined to a cave, perceiving reality only through shadows projected on a wall. Similarly, consumers are often confined by societal norms and marketing influences, perceiving needs and desires shaped by external forces rather than intrinsic values.
Escaping the "cave" requires a fundamental shift in perception and understanding. However, those who attempt to leave may face resistance or incomprehension from others still within the consumerist mindset. This highlights the psychological and social challenges inherent in adopting a radically different lifestyle.
The Wheaton Levels: Navigating the Transition
To facilitate the transition between different levels of consumerism and self-reliance, Fisker introduces the concept of "Wheaton Levels," a framework that delineates stages of understanding and practice. Each level represents a significant shift in mindset and capability, and progression requires both learning and unlearning.
Key principles of the Wheaton Levels include:
Incremental Progress: One cannot skip levels; each stage builds upon the previous one.
Relative Perspective: Individuals tend to see those one level ahead as inspiring and those several levels ahead as extreme or unrealistic.
Community Dynamics: People often struggle to understand or accept those at significantly different levels, leading to social friction.
Understanding this framework can help individuals navigate their own transitions and manage expectations regarding how others may perceive their choices.
Practical Steps Towards Resilient Lifestyle Design
Fisker offers several practical strategies for individuals seeking to adopt a systems approach:
1. Buy Nothing Year: Committing to a year of purchasing only essential items forces a reevaluation of needs versus wants and encourages finding creative solutions to problems.
2. Minimalism: Reducing possessions to the essentials simplifies life and reduces dependency on consumer goods.
3. Zero Waste: Focusing on reducing waste leads to more sustainable consumption patterns and a greater appreciation for resource cycles.
4. Maker Culture: Engaging in DIY projects and learning to repair and create items fosters self-reliance and skill development.
5. Financial Independence Retire Early (FIRE): Prioritizing savings and investments to achieve financial independence shifts focus from consumption to long-term security.
6. Permaculture: Adopting agricultural practices that work with natural systems promotes sustainability and provides direct benefits such as homegrown food.
Implementing these strategies requires adopting systems thinking to ensure that each action contributes to multiple aspects of resilience and well-being.
Individual and Community Approaches to Change
While individual actions are crucial, Fisker acknowledges the limitations of solely personal efforts. Sustainable change often requires community involvement to address systemic issues and create supportive networks.
There is a tension between individualistic approaches, which emphasize personal autonomy and self-reliance, and community-based approaches, which focus on collective action and shared resources. Fisker suggests that integrating both perspectives is essential for meaningful and lasting change.
Communities can amplify the impact of individual efforts by sharing knowledge, resources, and support. Conversely, individuals can bring innovative ideas and practices to communities, fostering a culture of resilience and adaptability.
Conclusion: Embracing a Resilient Future
Jacob Lund Fisker's systems approach to resilient lifestyle design offers a transformative pathway for individuals seeking to navigate the complexities of the modern world. By moving beyond the limitations of the consumerist paradigm and embracing systems thinking, individuals can develop lifestyles that are not only financially sustainable but also environmentally responsible and personally fulfilling.
This approach requires a willingness to challenge societal norms, invest in skill development, and think critically about the interconnectedness of various aspects of life. While the journey may present challenges, the rewards include greater autonomy, resilience, and the potential to contribute.
Re: Future of Artificial Intelligence
Decided to erase Jacob. I'm the expert now! No idea if the references are real. They sure look good though:
A Systems Approach to Resilient Lifestyle Design
In an era marked by economic volatility, environmental degradation, and social upheaval, the quest for a resilient lifestyle has never been more pressing. Traditional consumerist models, characterized by high levels of consumption and waste, are increasingly proving unsustainable. This article explores a systems approach to lifestyle design that emphasizes resilience, self-reliance, and sustainability, drawing on interdisciplinary insights from systems thinking, ecology, economics, and philosophy.
The Limitations of the Salary Consumer Model
The prevailing lifestyle in developed nations is often centered around the "salary consumer" model. In this paradigm, individuals specialize in a particular job to maximize income, which they then use to purchase goods and services to meet their needs and desires. This model is underpinned by the principles of comparative advantage and specialization from orthodox economic theory, which suggest that individuals and societies benefit when everyone focuses on what they do best.
However, this approach has significant limitations. It creates a high degree of interdependence and reliance on complex supply chains that are vulnerable to disruption. The COVID-19 pandemic, for instance, exposed the fragility of these systems, as lockdowns and economic shutdowns led to shortages and unemployment. Furthermore, this model encourages a cycle of consumption that generates excessive waste, contributing to environmental problems such as pollution and climate change.
Systems Thinking and Lifestyle Design
To address these challenges, a systems approach to lifestyle design advocates for viewing one's life as an interconnected web of activities and resources. Systems thinking involves understanding how different components of a system interact and influence one another, rather than viewing them in isolation. By applying this perspective to personal lifestyle choices, individuals can create more resilient and sustainable ways of living.
This approach aligns with the principles of permaculture and ecological design, which emphasize the importance of creating systems that are regenerative and self-sustaining. By considering the inputs and outputs of various activities, individuals can design lifestyles that minimize waste and make efficient use of resources.
Moving Beyond Specialization
One of the key shifts required in this systems approach is moving away from extreme specialization towards developing a diverse set of skills. While specialization can lead to efficiency in a narrow domain, it also creates vulnerability when circumstances change. Generalists, who possess a broad range of competencies, are better equipped to adapt to new situations and solve complex problems.
This concept is supported by research in cognitive science and education, which suggests that diverse experiences and knowledge bases enhance creativity and problem-solving abilities (Kell & Lubinski, 2013). By cultivating skills in areas such as gardening, cooking, repairing, and basic construction, individuals increase their self-reliance and reduce dependence on external systems.
The Role of Minimalism and Voluntary Simplicity
Minimalism and voluntary simplicity are philosophies that advocate for living with less and focusing on what truly adds value to one's life. These approaches can be integral to designing a resilient lifestyle. By reducing material possessions and unnecessary consumption, individuals can lower their environmental footprint and decrease financial pressures.
Studies have shown that materialistic values are associated with lower well-being and increased environmental harm (Kasser, 2002). Embracing minimalism can lead to greater life satisfaction and more sustainable living practices.
Implementing a Resilient Lifestyle
Implementing a resilient lifestyle involves several practical steps:
1. Assessing and Reducing Consumption: Conducting a thorough evaluation of one's consumption patterns to identify areas where reductions are possible. This might involve tracking expenses, analyzing waste production, and considering the necessity of various purchases.
2. Developing Diverse Skills: Learning new skills that enhance self-reliance, such as cooking from scratch, repairing household items, or growing food. Online resources, community workshops, and local skill-sharing groups can facilitate this learning.
3. Building Community Connections: Fostering relationships with neighbors and local communities to share resources, knowledge, and support. Community resilience is enhanced when individuals collaborate and pool their skills and resources (Gilchrist, 2009).
4. Adopting Sustainable Practices: Integrating sustainable practices into daily life, such as composting, using renewable energy sources, and choosing environmentally friendly products.
5. Financial Independence and Security: Working towards financial independence by saving aggressively, reducing debt, and investing wisely. The concept of financial independence emphasizes living below one's means to gain greater freedom and security (Dominguez & Robin, 1992).
Philosophical Underpinnings: Plato's Allegory of the Cave
The transition from a conventional consumerist lifestyle to a resilient, systems-oriented one can be likened to Plato's Allegory of the Cave. In the allegory, prisoners are chained in a cave, perceiving reality only through shadows projected on a wall. When one prisoner escapes and experiences the outside world, he realizes the limitations of his former perceptions. Upon returning to the cave to enlighten others, he faces resistance and disbelief.
This metaphor illustrates the challenge of breaking free from ingrained societal norms and perceptions. Adopting a resilient lifestyle often requires questioning deeply held beliefs about success, happiness, and the "good life." It involves recognizing the illusory nature of consumerism as the sole path to fulfillment and being open to alternative ways of living.
The Wheaton Levels: Stages of Lifestyle Transformation
Paul Wheaton, a permaculture advocate, conceptualized the process of transitioning to a more sustainable lifestyle in terms of levels. Each "Wheaton Level" represents a stage of understanding and practice in sustainability and self-reliance. Progressing through these levels involves incremental learning and adaptation, with each stage building upon the previous one.
This model acknowledges that drastic changes are often unsustainable and overwhelming. Instead, individuals can focus on manageable steps, gradually incorporating more sustainable practices into their lives. This approach aligns with behavior change theories that emphasize the importance of incremental progress and self-efficacy (Prochaska & DiClemente, 1983).
Overcoming Barriers to Change
Transitioning to a resilient lifestyle can be impeded by various barriers, including:
Financial Constraints: The initial costs of adopting sustainable technologies or practices can be prohibitive. However, many sustainable choices, such as reducing consumption or learning to repair items, can save money in the long run.
Social Pressures: Deviating from societal norms can lead to social isolation or criticism. Building a supportive community of like-minded individuals can mitigate these challenges.
Knowledge Gaps: Lack of knowledge or skills can hinder the adoption of new practices. Accessing educational resources and engaging in community learning opportunities can address this barrier.
The Importance of Community and Social Networks
While individual actions are crucial, the role of community cannot be overstated. Social networks facilitate the sharing of resources, information, and support, amplifying the impact of individual efforts. Community gardens, tool-sharing libraries, and local cooperatives are examples of how collective action can enhance resilience.
Research has demonstrated that communities with strong social ties recover more effectively from disasters and are better equipped to handle economic and environmental stresses (Aldrich & Meyer, 2015). Building social capital is therefore a key component of resilient lifestyle design.
The Emergence of a Post-Consumer Society
The cumulative effect of individuals adopting resilient lifestyles can contribute to the emergence of a post-consumer society. In this envisioned future, value is placed on sustainability, community, and well-being over material consumption. Economic systems are reoriented to support regenerative practices, and technological innovation is harnessed to solve environmental challenges.
This transition aligns with concepts from ecological economics, which advocate for a steady-state economy that operates within the ecological limits of the planet (Daly, 1991). It also resonates with the degrowth movement, which calls for a reduction in consumption and production to achieve sustainability and social equity (Kallis et al., 2018).
Conclusion
A systems approach to resilient lifestyle design offers a holistic framework for addressing the interconnected challenges of our time. By moving beyond the limitations of the salary consumer model and embracing practices that enhance self-reliance and sustainability, individuals can create lifestyles that are not only more resilient but also more fulfilling.
This transformation requires a shift in mindset, recognizing that well-being is not solely derived from consumption but from meaningful activities, relationships, and a harmonious relationship with the environment. Through individual actions and community engagement, it is possible to contribute to a broader societal shift towards sustainability and resilience.
References
Aldrich, D. P., & Meyer, M. A. (2015). Social Capital and Community Resilience. American Behavioral Scientist, 59(2), 254-269.
Daly, H. E. (1991). Steady-State Economics (2nd ed.). Island Press.
Dominguez, J., & Robin, V. (1992). Your Money or Your Life: Transforming Your Relationship with Money and Achieving Financial Independence. Viking Penguin.
Gilchrist, A. (2009). The Well-Connected Community: A Networking Approach to Community Development. Policy Press.
Kallis, G., Kerschner, C., & Martinez-Alier, J. (2018). The Economics of Degrowth. Ecological Economics, 154, 175-179.
Kasser, T. (2002). The High Price of Materialism. MIT Press.
Kell, H. J., & Lubinski, D. (2013). Spatial Ability: A Neglected Talent in Educational and Occupational Settings. Roeper Review, 35(4), 219-230.
Prochaska, J. O., & DiClemente, C. C. (1983). Stages and Processes of Self-Change of Smoking: Toward an Integrative Model of Change. Journal of Consulting and Clinical Psychology, 51(3), 390-395.
Re: Future of Artificial Intelligence
A 2024 Roundtable: Marx, Darwin, and Sterne on the Meta-Crisis
A dimly lit room, filled with the hum of futuristic technology. Three figures, each a giant of their respective fields, sit around a holographic table. Karl Marx, Charles Darwin, and Laurence Sterne, though centuries removed from their physical bodies, are present in consciousness.
Marx: "This meta-crisis, as you call it, is merely the latest iteration of class struggle on a global scale. The bourgeoisie, with their insatiable greed and reckless exploitation of the planet, have sown the seeds of their own destruction. The proletariat, the true engine of history, must rise up and seize the means of production."
Darwin: "The meta-crisis is a natural consequence of the relentless struggle for survival. Species adapt or perish. Human society, too, must evolve to meet the challenges of a changing environment. Those who are best adapted will thrive, while those who cling to outdated ideologies will inevitably decline."
Sterne: "Ah, a meta-crisis! A most curious predicament. A crisis of crises, a tangle of contradictions. We are creatures of both reason and passion, of science and sentiment. Perhaps, in this moment of great upheaval, we should embrace the absurd, the irrational, the comic. A good laugh, a well-timed joke, might just be the antidote to despair."
Marx: "Sentimentality will not save us, Sterne. We need concrete action, a revolutionary transformation of society."
Darwin: "And yet, Marx, it is our capacity for empathy and cooperation that has allowed us to survive thus far. We must nurture these qualities, even as we strive for progress."
Sterne: "Perhaps, gentlemen, the solution lies in a blend of your two approaches. A revolution of the heart, a Darwinian evolution of consciousness. We must strive for a society that is both just and compassionate, a world where reason and emotion are in harmony."
The holographic table flickers, and the three figures fade. The room is silent, the only sound the distant hum of technology. Yet, their words linger, a testament to the enduring power of ideas.
-
- Site Admin
- Posts: 17167
- Joined: Fri Jun 28, 2013 8:38 pm
- Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 77
- Contact:
Re: Future of Artificial Intelligence
I came across a conversation in the chat of one of the games I play in which someone was describing how he was using chatGPT to write books to sell on amazon. The "business" strategy was the old Tim Ferriss playbook on steroids. Find some niche subject, pretend to be the expert, and let the algorithmic Matthew effect do the rest as initial sales beget exponentially more sales (he didn't mention helping it along with paid reviews). The entire process of publishing a book went as follows:
The dude claimed he was making $18,000/month. It was only a question of throwing enough A/B tested mud on the wall to see what would stick. In terms of ethics he saw no need to disclose that the book was entirely written by AI. Caveat Emptor!
In conclusion: AIs can write books well enough to fool some people most of the time or most of the people some of time. It would also stand to reason that book platforms have little incentive to run zeroGPT-type checks on the stuff they're publishing as long as it keeps selling. I'm kinda wondering (hoping) that this book enshittification eventually leads to a reformation of the publishing/writing industry. Failing that, I guess I'll start avoiding any book that was published after 2022.
- Find popular niche subject with little competition.(1)
- Have AI write a 100 pages.
- Do some light editing if needed.
- Upload the document to amazon's kindle services which will instantly turn it into a book.
- A/B-test the cover design.(2)
- Profit
The dude claimed he was making $18,000/month. It was only a question of throwing enough A/B tested mud on the wall to see what would stick. In terms of ethics he saw no need to disclose that the book was entirely written by AI. Caveat Emptor!
In conclusion: AIs can write books well enough to fool some people most of the time or most of the people some of time. It would also stand to reason that book platforms have little incentive to run zeroGPT-type checks on the stuff they're publishing as long as it keeps selling. I'm kinda wondering (hoping) that this book enshittification eventually leads to a reformation of the publishing/writing industry. Failing that, I guess I'll start avoiding any book that was published after 2022.
Re: Future of Artificial Intelligence
It's not fooling people if they just buy it, and then realise it's just some ai generated bs.
It's dishonnest, but it's not worst than self proclained expert that write books, which has been a problem for a very long time. They just lose less time writing useless crap.
I don't think we will lose more time reading bs than before ai, but the sea of bs books might make it harder for good book to be discovered, because it will make the expected gain of reading a random book so low, that people will only read recomanded books, even those that usually take the risk to read tons of random books.
It's dishonnest, but it's not worst than self proclained expert that write books, which has been a problem for a very long time. They just lose less time writing useless crap.
I don't think we will lose more time reading bs than before ai, but the sea of bs books might make it harder for good book to be discovered, because it will make the expected gain of reading a random book so low, that people will only read recomanded books, even those that usually take the risk to read tons of random books.
-
- Site Admin
- Posts: 17167
- Joined: Fri Jun 28, 2013 8:38 pm
- Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 77
- Contact:
Re: Future of Artificial Intelligence
Many people will not realize this though. At no point in that chat conversation did I see any concern that the AI generated output might not be factual. No mention of fact-checking. Indeed, I gathered that the "author's" belief in AI was strong and that the profit was mainly made on the laziness of people to ask an AI themselves.
Some domains are more susceptible to people being fooled than others. Whenever the output is abstract or vague, AI does just as fine as a human. Examples would include small talk, academic bloviating, or inspirational overviews. The output is so vague or subject to interpretation that hallucinations don't matter. When the output is specific or detailed, the AI error rate goes up by a lot. Examples would include the infamous "glue on pizza"-example or historical details.
Re: Future of Artificial Intelligence
It will lead to a decay in reading (already in progress due to other forces). Secondly, due to those who still read, it will lead to an increase in the importance of publishing within the context of an already developed online personal brand (so you can persuade your readership you actually wrote this). This is also already under way.
Re: Future of Artificial Intelligence
No, they still publish it. I think it shows up on the listing somewhere but I’m not certain. Interestingly enough, you do not have to disclose if you have “AI assisted” content.
From their terms:
Artificial intelligence (AI) content (text, images, or translations)
We require you to inform us of AI-generated content (text, images, or translations) when you publish a new book or make edits to and republish an existing book through KDP. AI-generated images include cover and interior images and artwork. You are not required to disclose AI-assisted content. We distinguish between AI-generated and AI-assisted content as follows:
AI-generated: We define AI-generated content as text, images, or translations created by an AI-based tool. If you used an AI-based tool to create the actual content (whether text, images, or translations), it is considered "AI-generated," even if you applied substantial edits afterwards.
AI-assisted: If you created the content yourself, and used AI-based tools to edit, refine, error-check, or otherwise improve that content (whether text or images), then it is considered "AI-assisted" and not “AI-generated.” Similarly, if you used an AI-based tool to brainstorm and generate ideas, but ultimately created the text or images yourself, this is also considered "AI-assisted" and not “AI-generated.” It is not necessary to inform us of the use of such tools or processes.
You are responsible for verifying that all AI-generated and/or AI-assisted content adheres to all content guidelines, including by complying with all applicable intellectual property rights.
Re: Future of Artificial Intelligence
Autocatalytic transformers: -> data -> training -> testing ->
A dataset is exploited during training to be tested for satisfactory out of distribution activity to be fed back into the underlying dataset. Both training and testing have their own scaling trends driving the cost of tokens down for a similar amount of performance. Distillation between larger and smaller models also seems to work well. Training for longer runs into diminishing returns based upon the complexity of the dataset, and testing runs into diminishing returns based upon the complexity of the prompt/problem. Turning down the temperature converges closer to the original dataset, and turning the temperature up diverges from the original dataset. Some balanced approach may be reached depending on the domain to satisfy reliability vs creativity constraints. Domains like gaming, math, and coding are relatively easy to test at high fidelity so may be suitable for providing highly complex problems with automated verification. Allowing for more self-play with domain-specific scaling that can bootstrap from less data. A playground that is understood well-enough might be able to employ specialized agents to construct new problems for the general agent to solve, further catalyzing the loop. The prior generation of models partially bootstrapping the next gen.
Developments are too fast to keep up with, but the trail seems to be leading somewhere interesting for better or worse. Perhaps even sustained 5-7%+ economic growth interesting? That would be a pretty big deviation from trend. Capital, talent, resources, energy, policy, what will give?
A dataset is exploited during training to be tested for satisfactory out of distribution activity to be fed back into the underlying dataset. Both training and testing have their own scaling trends driving the cost of tokens down for a similar amount of performance. Distillation between larger and smaller models also seems to work well. Training for longer runs into diminishing returns based upon the complexity of the dataset, and testing runs into diminishing returns based upon the complexity of the prompt/problem. Turning down the temperature converges closer to the original dataset, and turning the temperature up diverges from the original dataset. Some balanced approach may be reached depending on the domain to satisfy reliability vs creativity constraints. Domains like gaming, math, and coding are relatively easy to test at high fidelity so may be suitable for providing highly complex problems with automated verification. Allowing for more self-play with domain-specific scaling that can bootstrap from less data. A playground that is understood well-enough might be able to employ specialized agents to construct new problems for the general agent to solve, further catalyzing the loop. The prior generation of models partially bootstrapping the next gen.
Developments are too fast to keep up with, but the trail seems to be leading somewhere interesting for better or worse. Perhaps even sustained 5-7%+ economic growth interesting? That would be a pretty big deviation from trend. Capital, talent, resources, energy, policy, what will give?
Re: Future of Artificial Intelligence
Analogy to think about might be the body of an organism. Each cell being a connected chunk of space (e.g. neighborhood) where chemicals/proteins are decision-making agents that possibly move around (human, app, machine, robot, vehicle, building, infrastructure, etc.), and the body itself being a civilization. A biological organism grows cell by cell where each is roughly equivalent give or take some mutations and rearrangements. Cells must work with the information around them to coordinate towards the moving attractor state of the body. In doing so they differentiate themselves accordingly while correcting the process continuously given local neuronal/electrical/chemical information. How does each cell know when to start/stop growing? Perhaps they use something like a fuzzy blockchain to verify global states.
As more virtual and real agents are added to civilization, the body grows and adapts to new environmental constraints while cells differentiate further. More levels of organization emerge and complexity increases. The "brain" develops from the instincts of each cell to coordinate interaction at the scale of body vs environment (based upon something like nested markov blankets, probability of in/out). Brains are great at top-down control, but cells retain bottom-up control to avoid fatal mistakes like touching fire and so forth. Perhaps as civilization develops a more information potent web, bottom-up coordination will require more readily provable models of the whole communicating continuously to avoid complexity reducing interactions with the environment. Exploit math and physics of local information to more securely explore global-environmental states.
A really tricky question, it seems, is where to draw the public/private boundary as the world becomes more transparent to itself? What boundaries will be drawn and how?
The boundary between body/environment is illusory in both cases from some perspectives, yet useful. Or at least that is what seems to be the case from my local vantage point where mass deception seems unlikely but always possible! The world seems coherent enough for now.
As more virtual and real agents are added to civilization, the body grows and adapts to new environmental constraints while cells differentiate further. More levels of organization emerge and complexity increases. The "brain" develops from the instincts of each cell to coordinate interaction at the scale of body vs environment (based upon something like nested markov blankets, probability of in/out). Brains are great at top-down control, but cells retain bottom-up control to avoid fatal mistakes like touching fire and so forth. Perhaps as civilization develops a more information potent web, bottom-up coordination will require more readily provable models of the whole communicating continuously to avoid complexity reducing interactions with the environment. Exploit math and physics of local information to more securely explore global-environmental states.
A really tricky question, it seems, is where to draw the public/private boundary as the world becomes more transparent to itself? What boundaries will be drawn and how?
The boundary between body/environment is illusory in both cases from some perspectives, yet useful. Or at least that is what seems to be the case from my local vantage point where mass deception seems unlikely but always possible! The world seems coherent enough for now.
-
- Site Admin
- Posts: 17167
- Joined: Fri Jun 28, 2013 8:38 pm
- Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 77
- Contact:
Re: Future of Artificial Intelligence
Wow, the Jedis are feeling that one! Popular US AI-related stocks are down 15-20% today. A Chinese company called DeepSeek came out with an open-source model that claims to be comparable to chatGPT but uses less sophisticated hardware and less energy to compute ostensible similar performance. And did I mention it's open-source?
Sounds like a brains over brawn or software over hardware situation which is quite the reversal compared to computing over the past couple of decades.
Insights?
Sounds like a brains over brawn or software over hardware situation which is quite the reversal compared to computing over the past couple of decades.
Insights?
Re: Future of Artificial Intelligence
I think the question remains as to whether or not Deepseek is lying. There are a few AI talking heads who think they actually are using roughly the same computing power as US companies but don't/can't say so. Perhaps in one instance because of restrictions/sanctions that would put Deepseek and NVIDIA in hot water respectively.
Another common response I've seen (including from Microsoft CEO Satya Nadella) is along the lines of "investors are silly! This is just another classic case of Jevons paradox. Deepseek be more efficient, but we'll still use the same amounts of energy in the end! "
Another common response I've seen (including from Microsoft CEO Satya Nadella) is along the lines of "investors are silly! This is just another classic case of Jevons paradox. Deepseek be more efficient, but we'll still use the same amounts of energy in the end! "
-
- Site Admin
- Posts: 17167
- Joined: Fri Jun 28, 2013 8:38 pm
- Location: USA, Zone 5b, Koppen Dfa, Elev. 620ft, Walkscore 77
- Contact:
Re: Future of Artificial Intelligence
In particular, how/why would such a competitive advantage be released as open source?
My only guess would be if they already have something better and want to mess with the US markets. If it's true, it does suggest that the US investment strategy in hardware is a material strategic blunder. If it's not then that kind of head fake only works once. I guess we'll see.
Re: Future of Artificial Intelligence
If it's true, not only was the large capex uneccessary, but their competition's product is free! That upends the whole business model of US AI companies to date. I'm not sure what that spells for the future of these companies. (ETA: This isn't exactly true. The chatbot is free but the APIs aren't)
Charles Hugh Smith has a good explainer article on the recent events and possible implications for anyone curious. He ends suggesting that the popping of the AI bubble will lead to the popping of the everything bubble, due to the heavy investment in AI by the Mag 7.
https://open.substack.com/pub/charleshu ... nt?r=35saq
Charles Hugh Smith has a good explainer article on the recent events and possible implications for anyone curious. He ends suggesting that the popping of the AI bubble will lead to the popping of the everything bubble, due to the heavy investment in AI by the Mag 7.
https://open.substack.com/pub/charleshu ... nt?r=35saq
I was also skeptical of the possibility--suggested to me by Adam Taggart, host of Thoughtful Money--that DeepSeek could conceivably prove to be the Black Swan that pops the AI bubble. This now strikes me as less of a possibility and more of an inevitability. There is simply no longer enough advantage generated by super-energy-consuming, costly chips in terms of generating a product that is worth paying for when equivalent tools are already available for free that can run offline on free-standing devices--which means there can't be any back-door stealthy "calling home" by the software.
And given that the Mag 7 stocks, most of which are heavily invested in AI, are almost 40% of the total market capitalization of the U.S. stock market, and the global concentration of equity wealth in the U.S. market, then the AI bubble popping will pop the entire global stock market bubble--and with that reduced to rubble, the entire global Everything Bubble in risk-on assets.
Last edited by theanimal on Mon Jan 27, 2025 5:13 pm, edited 1 time in total.