idonthaveacoolname

Rumination on Q* and what it could potential imply.

Q-learning and STaR are, I think, what OpenAI is talking about when it references Q*.

Language models’ capacity for nuanced reasoning has been a focal point of research. Enter the Self-Taught Reasoner (STaR), a groundbreaking technique that augments language models by integrating sparse rationale examples with vast datasets. This innovative approach fosters an iterative learning process, refining models to generate coherent chains of thought for diverse problem-solving tasks.

See STaR: Self-Taught Reasoner Bootstrapping Reasoning With Reasoning for more details.

The essence of STaR lies in its ability to fine-tune models based on the correctness of generated rationale. This iterative refinement loop catapults language models to not only achieve significant performance improvements but also to rival larger, more resource-intensive models on complex tasks like CommensenseQA. Does this mean that the model has surpassed the human results? From 56% on the original trials to equal 89%, the human performance, or more?

STaR’s success embodies a pivotal shift—a leap forward in language models’ autonomous reasoning. It sets a precedent for future advancements in bridging the gap between artificial intelligence and human-like cognition, redefining the boundaries of what these models can achieve.

Beyond STaR’s iterative prowess, insights gleaned from Q-learning and Markov chains provide critical guidance for scaling language models’ performance. Studies leveraging these concepts reveal a foreseeable decline in model performance as problem complexities increase.

Q-learning is a fundamental concept in reinforcement learning, a type of machine learning. It involves an algorithm that enables an agent to make decisions in an environment to achieve a specific goal. Through trial and error, Q-learning helps the agent learn the best action to take in a given state to maximize its cumulative reward. It does this by updating a Q-table, which stores the expected future rewards for each action in every possible state. Over time, the agent refines its actions based on the values in this table, gradually optimizing its decision-making process in complex environments without prior knowledge of the environment’s dynamics.

An aside – the implications of these insights underscore the necessity of strategically balancing computational resources during both training and testing phases. This balancing act becomes imperative for ensuring sustained model performance across a spectrum of intricate problem landscapes. The parallel nature of these once linear processes is where my interests lie.

Consider a scenario where language models seamlessly engage in real-time problem-solving during emergencies, prioritizing resource allocation akin to a human decision-making process. These insights lay the groundwork for future innovations, enabling language models to navigate diverse problem spaces with enhanced adaptability and efficacy. But how future? What defines the constantly shifting reward modeling? How does it allocate rewarding?

Language models, once confined to simple word predictions and text generation, have undergone a paradigm shift. They now navigate intricate reasoning tasks, delve into problem-solving domains, and strive towards human-like cognitive capabilities.

The journey towards refining reasoning capabilities extends into the domain of mathematical problem-solving—a seemingly straightforward yet challenging realm for language models. The GSM8K dataset encapsulates this complexity, revealing the struggle even formidable transformer models face in navigating grade school math problems.

To overcome this hurdle, researchers advocate for training verifiers to scrutinize model-generated solutions. The success of these verification mechanisms showcases their potency in augmenting model performance, especially in handling diverse problem distributions. This essentially not only increases the frequency but also the total distribution of rewards in any space, a clustering of rewards. Makes sense, this mirrors real world learning.

In the pursuit of refining reasoning capabilities, the exploration of supervision techniques emerges as a pivotal aspect. A comprehensive investigation into outcome and process supervision reveals the latter’s superiority in training models for intricate problem domains. Checking each step of a process, enabling reward reinforces accuracy rates.

Process supervision, with its meticulous feedback mechanism for intermediate reasoning steps, exhibits unparalleled reliability and precision. When coupled with active learning methodologies, exemplified by the release of PRM800K, this supervision approach propels related research endeavors, promising a robust foundation for future advancements.

Consider a scenario where these models assist in personalized education, adapting to individual learning styles, or co-create narratives alongside authors, blurring the lines between artificial and human creativity. The potential for language models to revolutionize domains extends far beyond what we envision today.

And as we delve deeper into the world of language models, the future holds boundless potential. Imagine language models not just deciphering language but engaging in philosophical discussions about complex moral dilemmas or even participating in real-time collaborative problem-solving scenarios during crises. And I think that a lot of the discussion about the “Crossing of the Rubicon” in the miasma of the last week at OpenAI revolves around the fact that now capable, the ethical “wrapper” is a shadow but imperative. Their ability to actively engage in profound ethical debates remains a nascent area.

Envision language models not just decoding textual content but understanding the depth and nuances of moral quandaries. Imagine a scenario where a language model is posed with a complex moral dilemma, such as the classic “trolley problem,” where decisions involve choosing between utilitarian principles and individual rights. The model, armed with extensive knowledge of ethical theories and moral reasoning, would not only parse the scenario but engage in a dialogue, weighing the pros and cons of different ethical frameworks and articulating its stance on the matter.

For instance, such a model could explore various ethical perspectives—utilitarianism, deontology, virtue ethics, or ethical relativism—articulating arguments, counterarguments, and the implications of each stance. It could draw from historical ethical debates, ethical principles, and even contemporary ethical dilemmas to contextualize its responses.

The implications of this extend far beyond theoretical discourse. Language models proficient in ethical reasoning could aid in decision-making processes across diverse fields. They could assist in ethical assessments in various industries, offer guidance in moral reasoning to individuals facing ethical quandaries, or serve as a tool for educators to facilitate discussions on ethics and morality.

However, such advancements raise profound questions and challenges. Ethical reasoning is inherently complex and often involves subjective considerations, societal norms, cultural context, and emotional intelligence—factors that are intricate for machines to grasp fully. The ethical development of such models would necessitate a deep understanding of not just logic but empathy, context, and the ability to comprehend the subjective nature of human ethical reasoning.

Moreover, the ethical implications of deploying such models into real-world decision-making contexts warrant careful consideration. How would we ensure the models’ reasoning aligns with societal values? How do we mitigate biases or unintended consequences in their ethical assessments?

Future innovations might unveil models that not only traverse language intricacies but also navigate philosophical landscapes, challenging societal norms, and catalyzing groundbreaking innovations across diverse domains. These reflections offer a glimpse into a future where language models not only emulate human-like reasoning but also shape the realms they interact with.

The landscape of language models has traversed a remarkable journey—from simple text generation to sophisticated reasoning and problem-solving. The advent of methodologies like STaR, insights from Q-learning and Markov chains, and the exploration of supervision techniques have thrust these models into realms once deemed unattainable.

As these advancements continue, the horizon of possibilities expands, offering a glimpse into a future where language models not only comprehend language intricacies but also engage in profound philosophical discourse, challenge societal norms, and catalyze innovative breakthroughs. The journey of language models is an ongoing exploration, promising exciting possibilities and transformative impact across various domains.

I decided to rummage through my photos for the ones that I took with my Holga lens. The grit of Naples and Palermo, let alone the Pantheon on a dark and rainy night in Rome, made for some nice black and white pictures.

And some art by Caravaggio and Michelangelo.

The Canadian government was recently vilified for the appearance and recognition in Parliament of an immigrant from Ukraine who served in the Galicia Division that was formed in 1943, made up of Ukrainian volunteers to fight against the Soviet Union. I decided to look into the text of the 1985-1986 Deschênes Commission that was digitized in 2012 by Privy Council Office to educate myself on this issue. This Commission was established specifically to report on the issue of war criminals in Canada and included discussion of this specific movement. Text is provided for Part I that was designed to be published with Part II, destined to remain confidential although there are calls to have this released. The number of war criminals in Canada ranged from a “handful” to 6000 at the time of the Commission. It should be noted that outside interveners in this Commission never quoted a figure under 1000.

The Report shows the geo-political role played by Canada at the end of the War in Europe, suffering immensely due to the economic collapse of decades of hostilities and war. All Europe was now moving, in one direction or another, and Canada’s economic goals aligned with geo-political goals of managing the recovery of Europe through the acceptance of large numbers of persons displaced by war, even if many of these policies originated in Washington or London. The Report also shows the limitations of Canadian authority as it emerged from its subordinate Dominion status. The accoutrements of a mature state were not yet in place.

As a (barely) former colony of the British Crown, Canada only enacted an independent Immigration Act in 1952 and decisions made by Cabinet in this instance must be considered using this lens. Independent foreign policy was less than two decades old at the close of hostilities. Immigration was subsumed under Order in Council P.C. 1931-695, a tightly restrictive policy. The Citizenship Act of 1947 was a step away from Canada’s status as dependent but a “Canadian” immigration act was still years away. Immigration was controlled from the centre, as it was in 1919, policies put in place after the Great War.

I’ll also add that since Canada was not a signatory to the Charter or the 1945 London Agreement, it didn’t have any jurisdiction over this type of offence. Canada still did not have fully developed State authority over elements that would have been needed here. This was only to come over the next decade. As the report makes clear, there was no attempt to hide the fact that these persons were members of the German military that Canada was at war with so no basis to claim that these people hid their histories and overturn their status in Canada.

Canada began to play a key role in the emergent Cold War of the late 1940s as a “release valve” for emigrants from across Europe, including those who lived in former Nazi and fascist controlled areas across Europe. Immigrants from all over the continent wanted to immigrate out of Europe, many of them living years directly or indirectly under fascist rule, suffering oppression and genocide. The initial years of the post-war were dedicated to returning Canadian service personnel from across the globe. Immigration as we know it was non-existent and “who” could immigrate was moot. As order returned, so too did the desire to manage this migration for state goals. Demand was coming both from our allies in Europe and the US, along with Canadians who wanted their family members to join them here, not to mention economic growth. Economic growth as a key policy priority was evident in the 1947 agreement with the UN Relief and Rehabilitation Agency and the International Refugee Organization to bring people from Europe as contract workers in specific labour markets.

In June 1949 all immigration was restricted by Order in Council 2743 to those with relatives in Canada, citizens of the UK, USA, and France and agriculturalists, miners, lumberers, loggers and domestic workers. These occupations did not, of course, have Occupational Codes since no system such as the National Occupational Classification or any occupational codes were in place. This data was not to be collected for decades. But I digress. Important for us here is that the Canadian government had denied the entrance of Ukrainians currently held in the UK including the Galicia / Halychyna Division.

This Ukrainian division had surrendered to British forces in May 1945 and were held in Rimini until 1946 when they were screened and transferred to the UK. After the repeal of P.C. 2743 in 1950 by P.C. 2856 there was a change in policy to accept, among others, this group that was currently held in the UK noting that they still required screening both for the immigrant and the applicant in Canada, if applicable. The Canadian Jewish Congress objected and Ottawa paused.

The British Foreign Office stated that both Soviet and British missions had no evidence that these persons fought against Western Allies or engaged in crimes against humanity. And while the CJC continued its objection, Canada opened the doors to Ukrainian immigration from the UK. According to the Deschênes Commission, approximately 600 of these former members of the Galicia Division were in Canada in the mid 1980s. Simon Wiesenthal gave a list of 217 specific names to then Solicitor General Robert Kaplan. This list was investigated and of those, over 86% never set foot in Canada and the few that did had no specific accusations leveled against them. Investigations by both the RCMP and the Commission reached the same conclusion.

I invite you to read the report. It is an interesting story of immigration to Canada that reaches into our own day with the resignation of the Speaker of the House last month.

I have been experimenting with standard forms in Catholic art from the 13th to the 16th century. Here are some midJourney examples.

Rene Magritte, de Kooning and Cy Twombly on the top row, Modigliani, da Vinci and Philip Guston on the bottom.

Basquiat, Piet Mondrian and Jasper Johns.

Agnes Martin, another Basquiat, Lucian Freud, Fra Angelico and Leonara Carrington.

Klimt, Picasso, Egon Schiele, Francis Bacon, Cy Twombly and Banksy.

My artist friend John is sending me various artists or movements or subject and style that I am using to picture in midJourney. For me I get the chance to experiment and to learn about each of these artists since I was finding that I was repeating a lot of prompts so needed some inspiration. In this case he wanted a “seascape by Albert Bierstadt” (1830-1902), Prussian born and part of the Hudson River and the Rocky Mountain schools of American art.

As I was reading up on him and the subject matter of his paintings, I thought of how this imagery contributed to westward expansion and the colonization of America. And as an historian, of course I thought of Fredrick Jackson Turner in his influential essay, “The Significance of the Frontier in American History,” which was first presented in 1893. Turner’s thesis argued that the American frontier experience had played a fundamental role in shaping American identity, culture, and democracy. This wasn’t good news for those displaced and forcibly removed from their ancestral lands, let alone the deaths that occurred at the hands of these colonists, but these visualizations were powerful motivators and contributed to the American sense of exceptionalism.

Depiction of the Frontier as a Source of American Identity:

Turner’s thesis proposed that the existence of the American frontier had a profound impact on the development of American characteristics, including individualism, self-reliance, and a sense of democracy. Bierstadt’s paintings of the American West, with their vast, untamed landscapes and sense of limitless opportunity, visually reinforce Turner’s idea of the frontier as a defining aspect of American identity. Imagery like this pictured the West as an empty geography that promoted ideas such as Manifest Destiny, a perennial issue to us Canadians.

Bierstadt’s works often showcased the frontier as a land of grandeur and natural splendor. His paintings, such as “The Rocky Mountains, Lander’s Peak,” conveyed the idea that the frontier was a place of both challenge and beauty, where Americans could test their mettle and forge a unique national identity. Light and luminance plays a key role in his work.

Romanticizing the Frontier:

Frederick Jackson Turner argued that the American frontier was not just a geographic place but also a state of mind, a crucible where American values and democracy were forged. Bierstadt’s art embraced the Romanticism movement, which sought to evoke deep emotional connections with nature. By depicting the frontier with awe-inspiring landscapes, he contributed to the romanticization of the West, portraying it as a land of boundless promise and adventure.

Encouraging Westward Expansion:

Turner believed that the closing of the frontier, symbolized by the U.S. Census Bureau’s declaration that the American West was no longer an unsettled frontier in 1890, marked the end of an era. Bierstadt’s paintings, created during a time of westward expansion and exploration, played a role in inspiring people to head west in search of opportunities, contributing to the westward movement that Turner highlighted in his thesis.

Bierstadt’s artworks conveyed the idea that the frontier was a place where dreams could be realized, encouraging settlement and exploration, which were central themes in Turner’s thesis. His paintings portrayed the West as a vast, uncharted territory ripe for discovery (and exploitation).

Symbolizing the Closing of the Frontier:

Ironically, as Turner’s thesis emphasized the closing of the American frontier, Bierstadt’s later works, such as “The Last of the Buffalo,” depicted the impact of westward expansion on the wilderness and native populations. These paintings highlighted the environmental and cultural changes that accompanied the closing of the frontier, aligning with Turner’s concerns about the consequences of westward expansion.

Here are a few images that were created when ideating in midJourney with the keywords of Albert Bierstadt and “seascape”!

Wow! The illustrations in the new edition of Beowulf from Folio are fantastic! I have read a few translations of this epic but not one that flows this well, especially when read aloud.

Beowulf is an epic poem composed in Old English, that delves into a tapestry of timeless themes that resonate across cultures and eras. At its core, it explores the intricate interplay between heroism, fate, and the human condition. The heroic exploits of the main character, Beowulf, exemplify the archetype of the hero as a valiant warrior who confronts monstrous adversaries to protect his people. Amidst the relentless battles and grand feats, the poem contemplates the inevitability of fate, illustrating the concept of wyrd, or destiny, that shapes the lives of both mortals and supernatural beings alike.

Like other “valiant warrior” archetypes from Gilgamesh, Arjuna in the Mahabharata, to Siegfried in the Nibelungenlied, Beowulf protects his people and upholds the ideals of loyalty and duty.

Through the tension between heroic action and the transience of human existence, Beowulf meditates on the impermanence of glory and the lingering echoes of one’s deeds. Philosophically, the poem serves as a window into the cultural values of the Anglo-Saxons, revealing their emphasis on courage, loyalty, and communal bonds. It encapsulates a world where heroic actions are both celebrated and tempered by a sense of melancholy, offering profound insights into the perennial struggle of humanity to navigate the dual nature of its aspirations and limitations. As a crucial piece of early English literature, Beowulf holds historical significance by providing a glimpse into the oral tradition of storytelling, while its thematic depth invites ongoing contemplation of the complexities inherent in the human experience.

Beowulf’s historical significance is most pronounced within the context of its original audience, the Anglo-Saxons. In a time characterized by tribal conflicts and cultural upheaval, the poem served as a vehicle for transmitting cultural values, moral lessons, and ancestral narratives. It encapsulated the ethos of warrior societies, reflecting the importance of heroic deeds, loyalty to one’s lord, and the communal bonds that held disparate clans together. Moreover, Beowulf’s Christian elements, juxtaposed with pagan spiritual themes, provide insights into the complex religious landscape of the period, underscoring the process of cultural and religious synthesis that marked the era’s transition. This synthesis is reflected in the poem’s exploration of the tension between the heroic code and the Christian notions of humility and salvation. As such, Beowulf not only preserves the tales of a bygone era but also mirrors the cultural dynamics, tensions and transformations that defined its historical context.

The role of monsters and ghosts within Beowulf extends beyond mere antagonists, serving as symbolic embodiments of the unknown, the monstrous, and the existential fears that have haunted humanity throughout time. The monsters, such as Grendel and his mother, represent both external threats and internal struggles. Grendel embodies the darkness lurking at the fringes of civilization, while his mother embodies the fierce maternal instinct and the vengeful aspects of femininity.

These supernatural beings serve as reminders of the impermanence of life and the inevitability of death. The dragon that Beowulf battles in his final heroic act symbolizes both the culmination of his heroic journey (Joseph Campbell is my guide here) and the fragility of human achievements in the face of time’s passage.

The defeat of these monsters by Beowulf showcases the triumph of human courage and prowess over primal forces. Furthermore, the dragon in the later part of the poem emphasizes the inevitability of mortality and the paradox of the hero’s quest for eternal fame. The supernatural beings in Beowulf, with their connections to the uncanny and the macabre, reflect the blurred boundaries between the real and the mythical in a world where the mysteries of existence often remained uncharted. In this way, the monsters and ghosts in Beowulf not only heighten the epic’s dramatic tension but also provide a lens through which to explore the psychological and existential dimensions of the human experience, offering a contemplative bridge between the mortal and the supernatural realms.

I finally received my digital back for my analog Mamiya RB67. I am having some focus issues but I have found leaving the shutter open helps with the testing. I will install the app and maybe that will help. But in the meanwhile I have been uploading my test shots to MidJourney for visualizations. Here are the prompts from the /describe command. I love some of the results although some of the prompts aren’t easily understood!

• books on top of a computer in the style of pinhole photography, darkly comedic, #vfxfriday, quiet introspection, ninetencore, loose gestural marks, understated sophistication –ar 323:242
• a stack of books on the desk, in the style of found footage, gadgetpunk, pinhole photography, thx sound, controlled chaos, les nabis, understated sophistication –ar 323:242
• this is one of many books, in the stule of holga 120n, piles/stacks, low bitrate, nintencore, peter saville, dark and moody, ub iwerks –ar 323:242
• stack of books in a home office, in the style of pinhole camera , gadgetpunk, dark subject matter, anti-gloss, tinkercore, candid, grainy –ar 323:242

I’m not sure what “nintencore” means but it appears twice. Holga and pinhole camera make sense though. I’m not sure how the algorithm “derived” this from the photo but here are the results!

• 

• 

• 

• 

And I used the new zoom out feature on the first MidJourney image above.

Adjusting Focus: https://www.youtube.com/watch?v=v-o6CN-0dAw

In today’s media-saturated world, where digital technologies and virtual spaces shape our perception and experiences, the concept of hyperobjects and their significance has gained increasing attention. Hyperobjects, as conceptualized by Timothy Morton, are entities that are massive in scale, extending beyond our immediate sensory perception and challenging our traditional modes of understanding. This essay explores the dynamic nature of mediated experiences and their profound impact on the interpretation and value assigned to hyperobjects over time. By examining the interplay between media, technology, culture, and the evolving historical context, we can gain insight into the complex and ever-changing relationship between humans and hyperobjects.

The interpretation and value assigned to hyperobjects have been subject to significant transformations throughout history, largely driven by advancements in media and technology. From the advent of print culture and the Enlightenment era to the rise of photography, cinema, and the digital age, each technological innovation has shaped the way we perceive and engage with hyperobjects. Media have played a crucial role in mediating our experiences, allowing us to comprehend and interpret hyperobjects in distinct ways across different historical periods.

The interpretive frameworks through which hyperobjects have been understood and valued have evolved alongside media and technological advancements. Early on, hyperobjects were often perceived through religious or metaphysical lenses, representing the sublime and the ineffable. However, with the rise of print culture and scientific discoveries, hyperobjects came to be interpreted through rational and empirical frameworks. The Enlightenment’s emphasis on reason and empirical observation provided new ways of understanding and valuing hyperobjects based on their material properties and measurable characteristics.

The invention of photography in the 19th century introduced a significant shift in the interpretation and value assigned to hyperobjects. Walter Benjamin’s concept of the aura, intertwined with the reproducibility of art, offers valuable insights into this transformation. Through the lens of photography, hyperobjects could now be reproduced, detached from their original context, and disseminated widely. This democratization of access to hyperobjects challenged the traditional notions of uniqueness, authenticity, and aura, altering their interpretation and value in the eyes of viewers.

The emergence of cinema as a mass medium further transformed the interpretation and value assigned to hyperobjects. Marshall McLuhan’s media theory provides a lens to understand how cinema, as an extension of human senses, altered our temporal experience of hyperobjects. Through the cinematic medium, hyperobjects could be captured in motion, extending beyond the boundaries of time and space. This temporal dimension introduced new ways of perceiving and interpreting hyperobjects, emphasizing their dynamic and evolving nature.

The advent of digital technologies and the proliferation of the internet have brought forth a new era of mediated experiences with hyperobjects. The concept of hyperobjects has become intertwined with the digital realm, as the internet facilitates widespread access, participation, and engagement with these massive entities. The dynamic and interactive nature of digital media allows for multidimensional encounters with hyperobjects, enabling individuals to contribute to their interpretation, value, and even co-creation. Online platforms and social media provide spaces for collective discussions, collaborative engagement, and the formation of diverse interpretive communities around hyperobjects.

It is crucial to recognize that the interpretation and value assigned to hyperobjects are deeply intertwined with cultural contexts and socio-political dynamics. Different cultural perspectives and historical narratives shape the ways in which hyperobjects are understood, valued, and integrated into collective consciousness. The meaning and significance of hyperobjects can vary across cultures, reflecting the diverse ways in which societies interact with and perceive their surrounding environments.

Mediated experiences have significantly influenced the interpretation and value assigned to hyperobjects over time. From print culture to photography, cinema, and the digital age, each technological advancement has shaped our perception and engagement with these massive entities. The evolution of interpretive frameworks, the democratization of access through reproducibility, the temporal dimension of cinema, and the interactive nature of digital technologies have all contributed to the dynamic nature of our relationship with hyperobjects. Furthermore, cultural contexts play a crucial role in shaping the meaning and significance attributed to hyperobjects, emphasizing the diverse ways in which societies understand and value these entities.

Understanding the dynamic nature of mediated experiences and their impact on the interpretation and value assigned to hyperobjects provides valuable insights into our evolving relationship with the world and the ecological challenges we face. By critically examining the interplay between media, technology, culture, and historical context, we can better comprehend the complexities surrounding hyperobjects and develop more nuanced and informed approaches to engage with these vast and interconnected entities in the present and future.