GPT-4 generated article

After giving GPT-4 the parameters of the article – namely, to answer the questions of how might AI technology and its successors contribute to human flourishing, and how might we as society best guide the technology to achieve maximal benefits for humanity – and then asked to generate an essay that I might write on this topic, I obtained the following output:

The Promise of GPT-4 and Human Flourishing: Charting a Path to Maximize Benefits for Humanity

As a mathematician who has dedicated my life to understanding the intricate patterns and structures that govern our world, I have always been fascinated by the power of human intellect to create and innovate. Today, I find myself marveling at the rapid advancements in artificial intelligence (AI), particularly in the realm of natural language processing, as embodied by the latest iteration of OpenAI’s generative pre-trained transformer—GPT-4.

GPT-4 is a remarkable feat of engineering that has the potential to contribute to human flourishing in myriad ways, from democratizing access to knowledge and expertise, to revolutionizing the creative process across various fields. By assisting with tasks ranging from research and writing to design and problem-solving, GPT-4 can augment human capabilities, freeing us to focus on higher-order thinking and fostering a culture of collaboration between humans and AI.

The potential benefits of GPT-4 and its successors are immense, but they also raise critical questions about how to guide the development and deployment of these technologies to ensure maximal benefits for humanity. As we consider the future of AI, I propose three guiding principles:

  1. Prioritize Ethical and Responsible AI Development: As AI systems become increasingly powerful, it is crucial that we prioritize the ethical dimensions of their development. This includes addressing issues such as fairness, accountability, transparency, and privacy. Developers and researchers must work collaboratively to establish best practices, and policymakers should implement regulations to ensure that AI systems are developed and deployed with the well-being of all individuals in mind.
  2. Emphasize AI Education and Workforce Development: The rise of AI technologies will inevitably reshape the labor market, and it is essential that we prepare the workforce for this transition. By integrating AI education into curricula at all levels, we can equip future generations with the skills and knowledge they need to thrive in an AI-driven world. Additionally, retraining and upskilling programs for those whose jobs may be displaced by AI can help mitigate the potential negative effects of automation on employment.
  3. Foster Public-Private Partnerships and Global Collaboration: The development and deployment of AI technologies should not be the sole responsibility of private corporations or individual nations. By fostering public-private partnerships and global collaboration, we can ensure that the benefits of AI are distributed more equitably, and that its development is guided by a diverse array of perspectives. This approach can also help prevent the monopolization of AI technologies by a few powerful entities, ensuring that the potential benefits of AI are enjoyed by all.

As we contemplate the future of GPT-4 and its successors, we must remain mindful of the balance between the potential benefits and risks that these technologies present. By adhering to the principles outlined above, we can work together to maximize the positive impact of AI on human flourishing, while minimizing potential harm. We stand at the precipice of a new era of human achievement, and I am confident that, with thoughtful guidance and collaboration, we can harness the power of AI to create a brighter future for all.

My own article

Resolving not to look too carefully at the GPT-4 generated article, I independently wrote my own article, which I reproduce here:

Generative AI: embracing change and resetting expectations

Humans have been trained over the last few decades to expect certain things from information technology. To list a few:

  • Hardware and software will improve (in such metrics as performance, user experience, and reliability) at a Moore’s-law type of pace, before transitioning to more incremental improvement.
  • Individual software tools can reliably produce high-quality outputs, but the input data must be of the highest quality, carefully formatted in the specific way that the tool demands.
  • The more advanced the tool, the more complex the specifications and edge cases, making interoperability between tools (particularly between different providers) a significant technical challenge unless well-designed standards are in place.
  • The humans will make all the key executive decisions; the software tool influences the decision-making process through its success or failure in executing human-directed orders.

All of these expectations will need to be recalibrated, if not abandoned entirely, with the advent of generative AI tools such as GPT-4. These tools perform extremely well with vaguely phrased (and slightly erroneous) natural language prompts, or with noisy data scraped from a web page or PDF. I could feed GPT-4 the first few PDF pages of a recent math preprint and get it to generate a half-dozen intelligent questions that an expert attending a talk on the preprint could ask. I plan to use variants of such prompts to prepare my future presentations or to begin reading a technically complex paper. Initially, I labored to make the prompts as precise as possible, based on experience with programming or scripting languages. Eventually the best results came when I unlearned that caution and simply threw lots of raw text at the AI. This level of robustness may enable AI tools to integrate with traditional software tools—or with each other, or with personal data and preferences. It will disrupt workflows everywhere in a way that the current AI tools, used in isolation, merely hint at doing.

Because these tools allow for a wide variety of inputs, we are still experimenting with how to use them to their full potential. I now routinely use GPT-4 to answer casual and vaguely phrased questions that I would previously have attempted with a carefully prepared search-engine query. I have asked it to suggest first drafts of complex documents I had to write. Others that I know have used the remarkable artificial emotional intelligence of these tools to obtain support, comfort, and a safe environment to explore their feelings. One of my colleagues was moved to tears by a GPT-4-generated letter of condolence to a relative who had recently received a devastating medical diagnosis. Used conversationally, GPT-4 can serve as a compassionate listener, an enthusiastic sounding board, a creative muse, a translator or teacher, or a devil’s advocate. They could help us flourish in any number of dimensions.

Current large language models (LLM) can often persuasively mimic correct expert response in a given knowledge domain (such as my own, research mathematics). But as is infamously known, the response often consists of nonsense when inspected closely. Both humans and AI need to develop skills to analyze this new type of text. The stylistic signals that I traditionally rely on to “smell out” a hopelessly incorrect math argument are of little use with LLM-generated mathematics. Only line-by-line reading can discern if there is any substance. Strangely, even nonsensical LLM-generated math often references relevant concepts. With effort, human experts can modify ideas that do not work as presented into a correct and original argument. The 2023-level AI can already generate suggestive hints and promising leads to a working mathematician and participate actively in the decision-making process. When integrated with tools such as formal proof verifiers, internet search, and symbolic math packages, I expect, say, 2026-level AI, when used properly, will be a trustworthy co-author in mathematical research, and in many other fields as well.

Then what? That depends not just on the technology, but on how existing human institutions and practices adapt. How will research journals change their publishing and referencing practices when entry-level math papers for AI-guided graduate students can now be generated in less than a day—and with the far better accuracy of future AI tools? How will our approach to graduate education change? Will we actively encourage and train our students to use these tools?

We are largely unprepared to address these questions. There will be shocking demonstrations of AI-assisted achievement and courageous experiments to incorporate them into our professional structures. But there will also be embarrassing mistakes, controversies, painful disruptions, heated debates, and hasty decisions.

Our usual technology paradigms will not serve as an adequate guide for navigating these uncharted waters. Perhaps the greatest challenge will be transitioning to a new AI-assisted world as safely, wisely, and equitably as possible.

Revised GPT-4 generated article after viewing draft

I then gave GPT-4 a draft of my own essay, and asked GPT-4 to rewrite its essay to more closely resemble my own style. This is what it returned:

Embracing the Age of AI Assistance: The Path to a Flourishing Partnership

The rapid development of generative AI tools such as GPT-4 has begun to challenge many of our long-held expectations about the relationship between humans and technology. As these tools become increasingly capable of interpreting vague or imperfect input and generating sophisticated outputs, they are poised to disrupt traditional workflows and reshape our understanding of human-AI collaboration.

AI tools like GPT-4 demonstrate an uncanny ability to adapt to the demands of various knowledge domains, including my own field of research mathematics. Although these AI-generated texts sometimes contain nonsense upon close inspection, they can also provide relevant concepts, hints, and promising leads for human experts to follow. This suggests that the future may bring AI systems that can participate as equal and trustworthy partners in the research process, across a wide range of professions.

As we stand on the precipice of this new era of AI-assisted human intelligence, it is crucial that we carefully consider the challenges and opportunities that lie ahead. One of the most pressing concerns is the potential disruption of long-established practices and institutions, such as research journals and graduate education. The ability of AI tools to rapidly generate research papers could force a reevaluation of publishing, refereeing, and educational practices. As a society, we must be prepared to address these questions and adapt to the changes they bring.

Navigating this uncharted territory will require a combination of bold experimentation and caution, as well as a willingness to abandon outdated paradigms. We must recognize that our traditional approaches to technology may not provide sufficient guidance for this new era. Instead, we must embrace a spirit of collaboration and adaptability, as we forge new partnerships between humans and AI.

As we strive to transition to a world of AI-assisted human intelligence, our final and greatest challenge will be to do so safely, wisely, and equitably. To this end, I propose several guiding principles:

  1. Encourage AI Literacy and Education: As AI tools become more deeply integrated into our daily lives, it is essential that we prioritize AI education at all levels. By training students and professionals to work effectively with AI tools, we can maximize the potential benefits of this new era of collaboration.
  2. Foster Ethical and Responsible AI Development: Ensuring the ethical development and deployment of AI systems is crucial. This includes addressing issues such as fairness, accountability, transparency, and privacy, and working collaboratively to establish best practices and regulations that protect the well-being of all individuals.
  3. Promote a Culture of Adaptability and Experimentation: Embracing the potential of AI-assisted human intelligence requires a willingness to experiment and adapt. We must be open to reevaluating long-held beliefs and practices, and courageous enough to forge new paths in the face of uncertainty.

By adhering to these principles, we can ensure that our journey into the age of AI assistance is one marked by human flourishing and the enrichment of our collective intellect. It is my belief that, with thoughtful guidance and collaboration, we can seize this opportunity to create a brighter future for all.

GPT-4 rewriting of my own article

Finally, I asked GPT-4 to directly improve the writing of my own article to make it more effective. It offered the following:

The AI Revolution: Redefining Expectations and Embracing New Horizons

After just a few months of experimenting with ChatGPT and GPT-4, I discovered an undeniable truth: AI is reshaping the way we interact with technology. Our preconceived notions about the limitations and potential of software are being challenged, and the implications are far-reaching.

Our long-held expectations of technology are rapidly becoming obsolete. GPT-4, and similar generative AI tools, excel at interpreting vague or imperfect input and producing sophisticated output. This remarkable robustness enables AI tools to integrate seamlessly with traditional software tools, disrupting workflows and revolutionizing our approach to problem-solving.

As we tap into the full potential of AI, we uncover new applications and opportunities. From generating intelligent questions about complex papers to providing emotional support during trying times, AI tools such as GPT-4 offer invaluable assistance in diverse aspects of our lives.

However, the road ahead is not without challenges. AI-generated text, while often persuasive, can contain nonsense upon closer inspection. Both humans and AI must develop new skills to critically analyze this emerging form of communication.

Despite these hurdles, AI continues to offer tantalizing hints and promising leads to experts across various fields. As AI integrates with other tools and technologies, it’s poised to become a trustworthy collaborator in research, education, and countless other domains.

But what lies beyond this horizon? The answer depends not only on the technology itself, but also on how our institutions and practices adapt. The rapid pace of AI-generated research output will challenge traditional publishing and education models, and force us to reevaluate our approaches.

As we venture into this new era of AI-assisted human intelligence, we must be prepared for both astounding achievements and uncomfortable disruptions. Our traditional technology paradigms will not serve as adequate guides for navigating this uncharted territory. The greatest challenge we face is transitioning to this new AI-assisted world as safely, wisely, and equitably as possible.

Embracing the AI revolution requires courage, adaptability, and a willingness to forge new paths. As we do so, we unlock the potential for a brighter, more connected, and intellectually enriched future.