AI
Matt Ross and
AI
Observing these patterns of influence gives clues about how our models generalize from their training data. For instance, if the models responded to user prompts by splicing together sequences from the training set, then we’d expect the influential sequences for a given model response to include expressions of near-identical thoughts. Conversely,... See more
Anthropic \ Tracing Model Outputs to the Training Data
Memorization vs learning concepts and forming models of the world.
Maybe AI doesn’t raise the bar. Maybe it reveals how low we’ve let the bar drop. In a world where ghosting is normal and attentiveness is rare, a chatbot that listens is radical.
Imaginary Friends Grew Up: We Panicked
When you hear a phrase like a “7B parameter model,” 7 billion parameters is a measure of the dataset, not the model . You obviously couldn’t train such a model out of, say, a text of 50 words (well you could, but it would be highly informationally degenerate). And if you gave the training protocol 100 internets worth of data, 7B parameters may not... See more
Venkatesh Rao • A Camera, Not an Engine
While LLMs are designed to emulate human-like responses, this does not mean that this analogy extends to the underlying cognition giving rise to those responses
Testing theory of mind in large language models and humans - Nature Human Behaviour
Many experts are concerned about how the adoption of AI systems over the next decade will affect essential human traits such as empathy, social/emotional intelligence, complex thinking, ability to act independently and sense of purpose. Some have hopes for AIs’ influence on humans’ curiosity, decision-making and creativity.
Experts Predict Significant Change in Humans’ Ways of Thinking, Being and Doing as They Adopt AI - Imagining the Digital Future Center
Superhuman performance of a large language model on the reasoning tasks of a physician
arxiv.org
In a stunning rebuke of traditional GOP deregulatory instincts, the Senate voted 99–1 to kill the measure. For the first time in decades, conservative priorities placed local cultural guardianship over the absolute freedom to innovate. That vote showed just how unsteady the old coalitions have become, and how powerful the populist realignment... See more
GitHub - transformerlab/transformerlab-app: Open Source Application for Advanced LLM Engineering: interact, train, fine-tune, and evaluate large language models on your own computer.
github.com
I’ve found that it pays to make journaling as “inconvenient” as possible. After I deleted Day One, I returned to writing in a physical journal, and that dramatically improved my mental health, critical thinking, time management, and memory. Multiple studies have shown that handwriting is better for memory retention and learning compared to typing.... See more