The rise of the smaller models.
Can LLMs Think Like Us?
Posted in neuroscience
LLMs don’t just memorize word pairs or sequences—they learn to encode abstract representations of language. These models are trained on immense amounts of text data, allowing them to infer relationships between words, phrases, and concepts in ways that extend beyond mere surface-level patterns. This is why LLMs can handle diverse contexts, respond to novel prompts, and even generate creative outputs.
In this sense, LLMs are performing a kind of machine inference. They compress linguistic information into abstract representations that allow them to generalize across contexts—similar to how the hippocampus compresses sensory and experiential data into abstract rules or principles that guide human thought.
But can LLMs really achieve the same level of inference as the human brain? Here, the gap becomes more apparent. While LLMs are impressive at predicting the next word in a sequence and generating text that often appears to be the product of thoughtful inference, their ability to truly understand or infer abstract concepts is still limited. LLMs operate on correlations and patterns rather than understanding the underlying causality or relational depth that drives human inference.
LLMs are reshaping how we think by fostering deeper, sustained engagement and metacognition, potentially transforming our cognitive abilities.
LLMs can open new doors to cognitive and perhaps even conscious expansion, influencing how we experience and interpret the world.
Discover how cognitive connectivity, powered by LLMs, ignites creativity, fuels joyful curiosity, and transforms the way we think and learn, in the emerging Cognitive Age.
Common push puppet toys in the shapes of animals and popular figures can move or collapse with the push of a button at the bottom of the toys’ base. Now, a team of UCLA engineers has created a new class of tunable dynamic material that mimics the inner workings of push puppets, with applications for soft robotics, reconfigurable architectures and space engineering.
Inside a push puppet, there are connecting cords that—when pulled taut—will make the toy stand stiff. But by loosening these cords, the “limbs” of the toy will go limp. Using the same cord tension-based principle that controls a puppet, researchers have developed a new type of metamaterial, a material engineered to possess properties with promising advanced capabilities.
Published in Materials Horizons, the study demonstrates the new lightweight metamaterial, which is outfitted with either motor-driven or self-actuating cords that are threaded through interlocking cone-tipped beads. When activated, the cords are pulled tight, causing the nesting chain of bead particles to jam and straighten into a line, making the material turn stiff while maintaining its overall structure.
The U.S. Department of Transportation (U.S. DOT) announced a plan to deploy new technology to ‘enable vehicles to communicate with each other’ in hopes of reducing motor vehicle crashes.
Scientists are trying to build a new sort of satellite, and have recently tested their idea with the Starling Formation-Flying Optical Experiment, or “StarFOX.” You may be getting flashbacks to the retro Star Fox video game series — and you’d be right to imagine this experiment as a science fiction fantasy brought into reality. There are no space-faring animals here, though.
Basically, StarFOX is a quartet of small satellites that work in tandem — a satellite “swarm,” as it’s sometimes called. This concept isn’t entirely new, but there’s something that sets StarFOX apart from standard satellite swarms. Typically, these conglomerates need external help in terms of orientation — but StarFOX’s four satellites can gauge their own positions with onboard cameras, an ability that could allow them to operate autonomously well beyond Earth orbit.
I don’t know what’s causing the sound problem my apologies.
Randal and Keith discuss WBE, Mind Uploading and fascinating tangents in neuroscience and neuroprosthetics and pathways for the future, as well as the Carbon Copies foundation and the new book ‘Contemplating Oblivion’ by Keith Wiley.
Many thanks for tuning in!
Researchers have discovered a “spatial grammar” in DNA that redefines the role of transcription factors in gene regulation, influencing our understanding of genetic variations and disease.
A recently uncovered code within DNA, referred to as “spatial grammar,” may unlock the secret to how gene activity is encoded in the human genome.
This breakthrough finding, identified by researchers at Washington State University and the University of California, San Diego and published in Nature, revealed a long-postulated hidden spatial grammar embedded in DNA. The research could reshape scientists’ understanding of gene regulation and how genetic variations may influence gene expression in development or disease.