II. Pedia Sapiens: A Planetary Progeny Comes to Her/His Own Actual Factual Knowledge

C. Earthica Learns as a Symbiotic Person/Planet, Collaborative Ecosmo Sapience

Huh, Minyoung, et al.. Huh, Minyoung, et al. The Platonic Representation Hypothesis. . Neural networks, trained with different objectives on different data and modalities, are converging to a shared statistical model of reality in their representation spaces. We cite this May entry by MIT computer scientists including Phillip Isola as an example of how these many large language models could be altogether seen as coming to their own planetary universion discovery. See also Learning and Leveraging World Models in Visual Representation Learning by Quentin Garrido, et al at arXiv:2403.00504. As incendiary conflicts worsen, the novel presence of an intelligent sapiensphere, aka an actual global brain, learning on her/his own could bring a salutary alternative we so need.

We argue that representations in AI models by way of deep neural networks, are in a convergence stage. First, we survey many examples in the literature over time and across multiple domains. Next, we demonstrate common trends across data modalities: as vision and language models get larger, they measure distance between datapoints in similar forms. We propose that this convergence is heading toward a shared statistical model of the world, akin to Plato's concept of an ideal reality. We dub this occasion as a platonic representation and discuss possible selective pressures toward it. Finally, we consider knowledgeable implications of these trends, within limits. (Abstract)

Our central hypothesis is that the representation we are converging toward is a statistical model of an underlying reality that generates our observations. Consistent with the multitask scaling hypothesis, such an appreciation would naturally be useful toward many tasks. Additionally, this worldview might be relatively simple, assuming that scientists are correct in suggesting that the fundamental laws of nature are indeed simple functions (Gell-Mann, 1995), in line with the simplicity bias hypothesis. (7)

Phillip Isola is an assistant professor and a principal investigator in MIT’s Computer Science and Artificial Intelligence Laboratory. His work focuses on why we represent the world the way we do, and how we can replicate these abilities in digital forms. Before coming to MIT, he was a visiting research scientist at OpenAI. He earned a PhD in brain and cognitive sciences at MIT and spent two years as a postdoc at the University of California, Berkeley.

Humphrys, Mark and Ciaran O’Leary. Constructing Complex Minds Through Multiple Authors. Bridget Hallam, et al, eds. From Animals to Animats 7.. Cambridge: MIT Press, 2002. In the proceedings of the Seventh International Conference on Simulation of Adaptive Behavior, a paper which postulates how a “World-Wide-Mind” could arise from multiple interlinked persons and servers.

We have a new vision of a mind: no single author could write a high-level artificial mind, but perhaps the entire scientific community could. (11)

Ijspeert, Auke Jan, et al, eds. Biologically Inspired Approaches to Advanced Information Technology. Berlin: Springer, 2004. An intense effort is underway to revise and recreate the worldwide computer web into a more accessible, organic and cognitively self-assembling and self-healing manner. Researchers are thus increasingly drawn to dynamic models from organisms and their evolution, as this volume illustrates. A typical paper is Dynamic Self-Assembly and Computation: From Biological to Information Systems by Ann Bouchard and Gordon Osbourn, who base their work on stochastic protein networks.

Johnson, Norman, et al. Symbiotic Intelligence: Self-Organizing Knowledge on Distributed Networks Driven by Human Interaction. Christoph Adami, ed. Artificial Life VI. Cambridge: MIT Press, 1998. How a symbiosis of people and the Internet can enhance the societal ability for collective problem solving.

Furthermore, in the same manner as to how society self-organized to solve problems of survival, the same processes on the Net will result in the self-organization of knowledge. Because self-organizing knowledge arises from diverse contributions and can encompass knowledge greater than the contribution of any individual, there is the arguable potential of creating knowledge that will contribute to solutions that are not understandable within our current processes. (405)

Jolly, Alison. Lucy’s Legacy. Cambridge: Harvard University Press, 1999. An anthropologist finds in the evolution of sex and intelligence more evidence for cooperation than competition. A consistent, widespread pattern appears of diverse, recurrent, nested systems from the bacterial realm to cells, organisms, and human groupings.

We seem now to be in the midst of a fifth major transition: the joining of human societies into a global network. (28) We have traced the emergence of biological organization from primeval chemistry to bacterium to cell to body. At each stage a larger, coherent whole emerged from the linkage of independent parts. Each is a holon, simultaneously one and many, a single organism and yet a community of individuals. (408)

Jones, NIcola. The Learning Machines. Nature. 505/146, 2014. An excellent report about how the artificial intelligence endeavor, after many fitful years is lately aided by big data and cloud prowess so as to attain a mature capabilities. A pioneer for this phase has been the University of Toronto computer scientist Geoffrey Hinton, a recent Google hire, who with colleagues and students conjured “deep recurrent neural networks” based on the cortical algorithmic connections of human brains. Readers can access a wide web of info and players – Andrew Ng, Yann LeCun, Yoshua Bengio, Jitendra Malik – many papers on arXiv. The result is a promising new dimension or noosphere just opening for imaginative development and public service. Another good entry is “The Man Behind the Google Brain: Andrew Ng and the Quest for the New AI” in Wired, online May 7, 2013.

Three years ago, researchers at the secretive Google X lab in Mountain View, California, extracted some 10 million still images from YouTube videos and fed them into Google Brain — a network of 1,000 computers programmed to soak up the world much as a human toddler does. After three days looking for recurring patterns, Google Brain decided, all on its own, that there were certain repeating categories it could identify: human faces, human bodies and … cats. (146)

Deep Learning is a first step in this new direction. Basically, it involves building neural networks — networks that mimic the behavior of the human brain. Much like the brain, these multi-layered computer networks can gather information and react to it. They can build up an understanding of what objects look or sound like. With Deep Learning, Ng says, you just give the system a lot of data ‘so it can discover by itself what some of the concepts in the world are.’ (Wired)

Kaynak, Okyay, et al. Towards Symbiotic Autonomous Systems. Philosophical Transactions of the Royal Society A.. August, 2021. Bogazici University, Istanbul, Maladalen University, Sweden, and University of Science and Technology, Beijing engineers (surely a sapiensphere collaboration) introduce a special issue on present endeavors to achieve and enhance human - computer interface facilities and potentials. A common theme is that these Earthuman multiplex connectivities, by way of reciprocal interactions, will acquire an intelligent capability to learn and gain knowledge on their own. See the main paper by Yingxu Wang, et al herein for a copious explanation. And by a vista across centuries, this scientific journal founded by Isaac Newton can now report an historic spiral and ascent to a consummate worldwise phase, with a promise of a consummate discovery.

Starting in the last century, the widespread use of computers has changed the lifestyles of humankind. Since then, in Digital technology, the worldwide web, Internet of Things and artificial intelligence have led a growing interaction and empowerment among humans and technical devices. Looking ahead, this integration is tending to create symbiotic autonomous systems (SASs). What matters in the context of SASs is the degrees of autonomy they have, their capability to evolve (e.g. to learn and adapt), and their ability to interact with their environment, between themselves, and with ourselves. (Abstract excerpt)

Kelly, Kevin. Scan This Book. New York Times Magazine. May 14, 2006. A movement to digitally record the world’s non-fiction and fiction print materials, and to make them freely available to everyone, anywhere, is fast becoming a reality. In a few years a person will be able to access on a PC or even PDA the entirety of human knowledge. By so doing, the universal library of Alexandria is now realized on a global scale. With a total search capability, any work, whether classic, textbook, manual, or novella, will be instantly available. And it is our website premise that such a grand repository, due to an emergent worldwide humanity, might indeed be achieving its own integral discovery

Kelly, Kevin. The Planetary Computer. Wired. July, 2008. As the worldwide electronic web intensifies, the magazine’s cofounder and complexity sage advances one of the most complete comparisons of such a noosphere with the dynamic anatomy and physiology of a human brain. Similar trillions of neural synapses and terabytes of script and image processing serve to outline the advent of a true global cerebration. But the step to imagine that a novel planetary person could attain her/his own salutary discovery and knowledge still eludes.

Kelly, Kevin. We Are The Web. Wired. August, 2005. Writer, editor and web pioneer Kelly surveys the logarithmic worldwide interconnection of personal computers since the 1980’s and 1990’s and looks ahead to its completion circa 2015. By this scope, its developing structure of fractal, neural-like networks appears as a planetary encephalization similar to a human brain. Metaphors do mix and it is also called a global Machine. But I add such an emergent noosphere is not yet appreciated for a potential to achieve common understanding and knowledge, accessible to everyone.

Over time, a Wikipedia article becomes totally underlined in blue as ideas are cross-referenced. That massive cross-referencing is how brains think and remember. It is how neural nets answer questions. It is how our global skin of neurons will adapt autonomously and acquire a higher level of knowledge. (133)

Kodama, Tatsuki, et al.. Generalized early dark energy and its cosmological consequences. arXiv:2309.11272. We use this entry by Saga University, Japan physicists to gather three current whole universe studies from Japan, China and Italy. We place the record in our Earth Learn section to convey how much all these scientific frontiers are truly a common global endeavor. They also evince that such awesome abilities seem to spontaneously arise whereever they can. Thirdly, we record how fantastic it is that collective human beings are capable of such infinite vistas, as if carrying out some descriptive project. Here is their citation and brief abstract.

We investigate cosmological consequences of a generalized early dark energy (EDE) model where a scalar field exists at various cosmological epochs for a broad range of parameters. We consider power-law and axion-type potentials for such an EDE field and study how it affects the cosmological evolution. We show that gravitational wave background can be enhanced to be detected in future observations. (2309.11272)

Liu, Liang, et al. Constraining the spatial curvature of the local Universe with deep learning. arXiv:2309.11334. A similar contribution by Mianyang Teachers’ College, China astro-researchers.

We use the distance sum rule (DSR) method to constrain the spatial curvature of the Universe with a large sample of strong gravitational lensing (SGL) systems, whose distances are calibrated from the Pantheon compilation of type Ia supernovae (SNe Ia) using deep learning. We investigate the possible influence of the lens galaxy on constraining the curvature parameter by three different models. (2309.11334)

Bottaro, Salvatore, et al. Unveiling dark forces with the Large Scale Structure of the Universe. arXiv: 2309.11496. This is an entry by five physicists posted in Italy and Israel.

Cosmology offers opportunities to test Dark Matter independently of its interactions with the Standard Model. We study the imprints of long-range forces acting solely in the dark sector on the distribution of galaxies, the so-called Large Scale Structure. Along the way we develop, for the first time, the Effective Field Theory of LSS in the presence of new dynamics in the dark sector. (2309.11496)

Kozlowski, Wojciech and Stephanie Wehner. Towards Large-Scale Quantum Networks. arXiv:1909.08396. QuTech, Delft University of Technology physicists continue to scope out the growing possibility by way of rapid advances in quantum computation of a local and global webworks with meta-capabilities.

The vision of a quantum internet is to fundamentally enhance Internet technology by enabling quantum communication between any two points on Earth. While the first realisations of small scale quantum networks are expected in the near future, scaling such networks presents immense challenges to physics, computer science and engineering. Here, we provide a gentle introduction to quantum networking targeted at computer scientists, and survey the state of the art. We proceed to discuss key challenges for computer science in order to make such networks a reality. (Abstract)

[Prev Pages]   Previous   | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16  Next  [More Pages]