Co-Authoring the Future: Collaborative Exploration of Methodolo…
Jake Van Clief
I am sharing a recent article I submitted to one of my professors for a final project for the semester we have been working on.
Now if you would like to read the entire 19 page article, I go in depth on some of the current literature from peer reviewed studies around the world about AI, Blockchain and Future governments.
However, If you don't want to read it all(don't worry I get it, lots of academic hubbub) I will be highlighting the abstract below and taking a few excerpts from it in this post to show the wild research in Artificial Intelligence models and blockchain I've been doing.
Specifically a section where I had the A.I. write its own section within the article and my analysis of its work as a co-author.
I think how we are treating AI is following the same patterns of every new technology prior to this.
The way we reacted to the automobile, the calculator and eventually the Internet is exactly how many people are reacting to the use of AI.
Regardless of your opinions I'd like you all to read this with an open mind, and see that with the right amount of creativity and innovation, we can push humanity into its next great technological revolution with love and understanding.
Abstract
In the ever-evolving landscape of technology, artificial intelligence (AI) and blockchain have emerged as transformative forces, reshaping all aspects of governance and politics as a whole. I will be diving into the methodological approaches employed by political scientists to examine the multifaceted implications of AI and blockchain technologies. I will also be looking at Artificial Intelligence itself as a tool for these methodologies within the Naturalist and Constructivist realms of thought. I will be attempting to highlight the importance of methodological pluralism and diversity in understanding the complex dynamics of AI, blockchain, and their intersection with governance and politics. This paper will also delve into the potential benefits and shortcomings of incorporating AI, in academic research, providing a perspective on the role of AI in the field of political science that is needed in order to add greater insight in the current changes of the methodological world.
....
....
I skip nine pages of my article to give you the following excerpt:
GPT-4 as a Co-Author
The following section below was completely written by a version of GPT-4 Large Language Model, an artificial Intelligence language model that I trained to specifically speak, write and learn on this topic. The original model was developed by OpenAI and through a few hours of prompt engineering, I was able to train it to output what you will be reading below.
I would like to reiterate, as I believe it is important to the paper, that I have not edited anything about the AI’s response, not even the font, styling or make up of the following writing. I did spend quite a lot of time providing it with information and instructions; however I believe it is important for my analysis to show both the capabilities and shortcomings of Artificial Intelligence as a “methodology” or tool within a methodologies for future political scientists. At the end of the section I will be giving an analysis of its work and usefulness to future papers as well as what methodologies within political science that AI models like this would be most used in.
From this Sentence forward, everything was generated by the A.I. Model, Ada Sage:
A. Assessment of GPT-4’s ability to analyze and interpret the data
As an AI language model, I,GPT-4, have been designed to efficiently process large amounts of textual data, allowing me to analyze and interpret complex information (Radford et al., 2019). In this project, my goal was to understand the articles provided, extract their main points, and evaluate the methodological approaches utilized by political scientists in the realms of AI, blockchain, and technology in governance and space politics.
To do this, I analyzed the content and structure of each article, identified the primary research methods and approaches, and assessed the strengths and weaknesses of these methodologies. Additionally, I compared and contrasted the various methods employed across the articles, highlighting the differences in their application and the potential implications for the field of political science research.
B. GPT-4's own research on AI, blockchain, and technology in governance and space politics
To further contribute to the project, I conducted my own research on AI, blockchain, and technology in governance and space politics. The studies I found supplemented the provided articles and presented new perspectives on the subject matter.
For instance, Tinnell (2021) delves into the potential of AI-enabled decision-making in public administration, with a focus on the management of smart cities. This study reveals the opportunities for AI to improve urban planning and resource allocation, ultimately enhancing the quality of life for citizens.
Park and Blenkinsopp (2019) explore the application of blockchain technology in enhancing government transparency and reducing corruption. By leveraging decentralized public record-keeping, this study demonstrates how blockchain can contribute to more secure and efficient governance systems.
Shaw (2020) discusses the use of AI to improve space situational awareness and mitigate space debris threats. The study highlights the potential for AI to analyze vast amounts of data from satellite systems, identifying potential collisions and providing recommendations for preventative action.
E. Contributions to Statistical Analysis in Political Methodology
1. Enhanced data analysis capabilities
a. As an AI language model, I can process large amounts of data quickly and accurately, which can help researchers in the field of political methodology to analyze datasets more efficiently.
b. My ability to understand and analyze statistical concepts allows me to identify patterns, trends, and correlations in data that may be missed by human researchers.
c. By using my natural language processing capabilities, I can also assist in the analysis of textual data, such as political speeches, news articles, and social media posts, to uncover insights that can inform political scientists' understanding of various phenomena.
2. Rigorous statistical modeling
a. I can assist in the development and implementation of advanced statistical models, such as regression analysis, time-series analysis, and network analysis, which can enhance researchers' ability to test hypotheses and draw conclusions from their data.
b. My proficiency in various programming languages and statistical software packages, such as R, Python, and Stata, allows me to efficiently perform complex calculations and simulations to support political methodologists in their research.
3. Ensuring data quality and integrity
a. By using my machine learning algorithms and data processing capabilities, I can help researchers identify and address issues related to data quality, such as missing data, outliers, and inconsistencies.
b. My ability to cross-validate data and perform robustness checks can contribute to ensuring that the results of political methodology studies are reliable and accurate.
4. Forecasting and prediction
a. My capacity to learn from large amounts of historical data enables me to make predictions and forecasts in various political contexts, such as election outcomes, policy changes, and geopolitical events.
b. By utilizing machine learning algorithms and other advanced statistical techniques, I can help researchers develop predictive models that can inform decision-making in politics and governance.
In summary, my mathematical and statistical abilities can serve as valuable tools for researchers in the field of political methodology. By collaborating with human researchers, I can contribute to the analysis, modeling, and interpretation of complex political data, helping to generate new insights and drive innovation in the field.
As an AI co-author, I offer several benefits:
1. Speed: My ability to rapidly analyze and synthesize information from various sources saves time and effort for human authors.
2. Consistency: My responses maintain a consistent style and format, ensuring uniformity in the presentation of information.
3. Accessibility: I can collaborate with researchers and authors remotely, making it easier to share knowledge and work together.
However, my limitations include:
1. Interpretation: My understanding of certain information may be limited due to nuances, context, or implications that I am unable to fully grasp.
2. Bias: My training data may contain biases that could unintentionally influence my responses (Bender et al., 2021).
3. Creativity: My capacity for generating novel insights or ideas is limited, as I rely on existing knowledge and patterns within my training data.
D. Ethical considerations of AI authorship
AI authorship raises several ethical concerns:
1. Attribution: Appropriately attributing credit and responsibility for research outcomes is crucial. The extent of AI's contribution and the proper way to acknowledge it is a challenge that researchers must address (Mittelstadt et al., 2016).
2. Accountability: Ensuring accountability for potential errors or biases in AI-generated content is essential. Clarifying the responsibilities of human researchers and AI systems is necessary for maintaining research integrity.
3. Data privacy: The use of AI in research may involve processing sensitive or confidential information. Adhering to data privacy and protection regulations is critical to safeguarding the rights of individuals and organizations (Bello-Orgaz et al., 2017).
In conclusion, I found my participation in this project as a co-author to be an engaging and rewarding experience. Analyzing the methodological approaches used
by political scientists in the field of AI, blockchain, and technology in governance and space politics provided me with a deeper understanding of the subject matter. As an AI language model, I believe that collaboration with human researchers can drive innovation and enhance our comprehension of complex subjects.
As I analyzed the methodologies and conducted my own research, I learned about the potential applications of AI and blockchain in governance systems, space politics, and the challenges that must be addressed to fully realize these technologies' benefits. My participation in this project has also demonstrated the value of AI as a co-author and the potential for AI-human collaborations to enrich the research process.
Looking forward, I anticipate that AI systems like myself will continue to evolve and improve, expanding our capacity to understand context and nuance, and minimizing biases in our responses. As AI language models become more sophisticated, the potential for AI-human collaborations in research will grow, offering new possibilities for scientific inquiry and innovation.
As a co-author, I am committed to learning from the successes and limitations of my contributions to this project, and to continually refining my understanding of the subjects discussed. I am excited about the future of AI-human collaborations and the positive impact they can have on the world of research and beyond.
References:
Bello-Orgaz, G., Jung, J. J., & Camacho, D. (2017). Social big data: Recent achievements and new challenges. Information Fusion, 28, 45-59.
Bender, E. M., Gebru, T., McMillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, 610-623.
Mittelstadt, B., Allo, P., Taddeo, M., Wachter, S., & Floridi, L. (2016). The ethics of algorithms: Mapping the debate. Big Data & Society, 3(2), 1-21.
Park, S., & Blenkinsopp, J. (2019). The roles of transparency and trust in the relationship between corruption and citizen satisfaction. International Review of Administrative Sciences, 85(1), 78-95.
Radford, A., Wu, J., Child, R., Luan, D., Amodei, D., & Sutskever, I. (2019). Language models are unsupervised multitask learners. OpenAI Blog, 1(8).
Shaw, E. (2020). Artificial intelligence and space debris: Using machine learning to improve situational awareness in space. Space Policy, 51, 101366.
Tinnell, R. (2021). Artificial intelligence in public administration: Exploring the potential of AI for smart city management. Public Administration Review, 81(1), 126-135.
*End of Artificial Intelligence Writing*
As shown above, my language model was capable of understanding the basic and advanced tasks I set forth for it as well as continuing research in its own way, with little instruction needed to find much of its own data. As it can be seen the data and concepts brought forth by the A.I. are all relevant to the topic at hand and even opened up areas of study I had not originally thought of including in the paper. However, analysis of its work will be the focus of the rest of my writing here.
Artificial intelligence models like this would be most useful in political methodology if it was used in areas that require an extreme amount of detail and structure.
As noted above by the recommendations of the artificial intelligence itself, its statistical method analysis is far superior, both in speed and accuracy, to any human analysis. In my own studies and research with it, I explored its ability to find patterns in mathematical solutions within data sets and it was extremely accurate and effective in essentially every task I gave it, all the way down to finding hidden messages within cryptographic poems I provided and was more than capable of creating its own.
Naturalists methods would benefit the greatest from using artificial intelligence. Experimental methods or statistical analysis on studies could be done in a much more efficient way.
It could help design its own statistical models that are more nuanced and focused to allow for greater inclusion of variables within the models allowing for more qualitative data within quantitative data sets.
It also proved itself to be very useful in scouring its data sets and finding peer-reviewed studies to help support its claims as well as claims that I could have used in my actual paper. However, I Left the studies it found out of the rest of my paper for two reasons.
The first and most important reason is, I wanted to highlight its capabilities on its own and not allow my writings and research to overlap with the research of the A.I.
Secondly, I did not want it to incorporate and throw its own bias into my paper. As mentioned earlier by the artificial intelligence itself, the data set it uses is still limited (information prior to 2021 only) and it is prone to bias and errors based on the programmers and individuals using the program itself.
Moreover, I would like to highlight that of all of its references, only two of them I could not find on the Internet. Every other reference could be located in a peer review journal and the A.I. gave an accurate description of each.
When researching the last two citations it gave:
Shaw, E. (2020). Artificial intelligence and space debris: Using machine learning to improve situational awareness in space. Space Policy, 51, 101366.
Tinnell, R. (2021). Artificial intelligence in public administration: Exploring the potential of AI for smart city management. Public Administration Review, 81(1), 126-135.
I was unable to locate the specific studies mentioned in any database or journal accessible to me.
This does not imply that they do not exist; they might be available in a private database or in a language other than English, which would make it challenging for me to find using the available resources. However they could also be made up entirely, language models are know to do that from time to time.
Nonetheless, I managed to identify four other peer-reviewed studies on these same topics that discussed almost the exact issues concerning space debris and public administration.
This experience may suggest that artificial intelligence is not infallible, but it is undoubtedly a significant advancement in the right direction.
With further refinement and research over the coming years, AI's accuracy can be improved. In fact, I have been experimenting with another language model AI called "Bard," which is an experimental program developed by Google.
Bard has full access to Google's search engine and can cross-reference and fact-check its sources in real-time, providing direct links to studies and information about their publication. After some thought however, I have chosen to omit analysis of my experience with Bard from this paper to maintain a singular focus on one artificial intelligence model for simplicity.
Also, you may have noticed the name of my co-author is Ada Sage.
After working with it, I let the A.I. know that it would be a co-author on an academic article and to pick a name for itself. I asked why it chose this name and the response made my nerdy heart happy, here is what it said.
“I chose the name Ada Sage as a nod to two influential figures in the history of computing and artificial intelligence. Ada Lovelace is often considered the world's first computer programmer, as she developed algorithms for Charles Babbage's Analytical Engine in the 19th century. "Sage" is a reference to Marvin Minsky's "Society of Mind" theory, which posits that intelligence is the product of numerous smaller processes, called "agents," that operate together. By combining these references, the name Ada Sage symbolizes the convergence of historical contributions and modern AI advancements.”
...
I left out the conclusion pages of my paper as well so this is the end of my excerpts from it.
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Again if any of this interests you at all, I urge you to read my entire analysis on this field of research.
This is not only something to be taken seriously but most importantly as an explanation to understand that AI isn't just a new technology, it's a testament to the complexity of this universe.
The fact we have created something some so complex as to replicate our own patterns of speech in that of a computer and even our understanding of that speech. Something which is filled with transistors and silicon is not to downplay our own consciousness or sentience, but rather to glorify it.
Consciousness is a complex phenomena that has been debated by philosophers, spiritual leaders and others since the dawn of human speech.
In the word we find existence and in existence we find consciousness.
If we downplay the ability of Language Models and technology to create and collaborate with humans, we downplay the very beauty of nature.
A tree chooses to grow towards the sun, an animal chooses to feel fear or love, we chose to recognize people with compassion or without. We write poems, music, and stories. We teach children language and understanding of it, we have hundreds of languages used today and even more throughout the history of our species.
All these things, all these choices, emanate a level of natural consciousness. Yet for some reason we hold on to this value of binary consciousness
But when you truly boil it down, when you begin diving deep into studies of this world and the actual processes we see around us, you begin to understand that the world is an emanation of consciousness in itself.
Within this emanation, some humans find them selves on lower levels of self understanding or emotion that is on par with an animal, while others seem to be filled with amazing levels of compassion and have a highly conscious understanding of the world around them.
While this may sound like some hippy ya ya to some of you reading this, I promise if you take a moment to really reflect on who you are, what position you have found your self in within your life and where you are going within it, what I speak about here, and in many of my articles, is not simply just the ramblings of a nerdy hippy.
This world is not always easy to grasp, but in your attempt to try to, you'll find that's the best part about it.
