This article examines the fundamental differences between how human children acquire values and competencies versus how current artificial intelligence systems are trained. While modern AI training relies on large datasets and software guardrails, humans learn through a combination of innate biological programming, personal experience, and essential social interaction within a community. The author argues that to achieve truly reliable, trustworthy, and human-compatible AI, developers should move beyond mere data containment and instead integrate deep learning with the collaborative, staged experiences found in developmental robotics and social species.
A review exploring how the current artificial intelligence revolution is driven more by capital interests and hype than technological necessity. The piece examines Cory Doctorow’s argument that technology development is often steered toward maximizing investor returns rather than human empowerment, leading to a phenomenon known as "reverse centaurs" where workers lose autonomy and skill to machines. It critiques the industry's perceived inevitabilism and its similarities to the process of enshittification seen in other tech sectors.
* The economic motivations behind artificial intelligence hype
* Concept of reverse centaurs versus automation theory
* Critique of Big Tech business models
* Impact of capitalism on technological progress
* The economic motivations behind artificial intelligence hype
* Concept of reverse centaurs versus automation theory
* Critique of Big Tech business models
* Impact of capitalism on technological progress
Nobel laureate Daron Acemoglu critiques current optimism regarding AI productivity and economic narratives. He argues that much of the prevailing debate is speculative and fails to address critical issues like concentrated corporate power and extractive data models. Rather than focusing on whether capitalism is mutating, he suggests evaluating technology based on whether it fosters inclusive or extractive institutions.
>*Seen through that lens, AI is not troublesome in its own right, but rather whether it is positioned as inclusive or extractive. Today’s AI hyperscalers, he argues, fit the extractive mold almost perfectly: concentrated ownership, regulatory capture, and a business model that extracts data and attention at scale."
- Skepticism toward massive near-term AI productivity gains due to current model limitations
- The distinction between simple automation and true human-complementary tasks
- Potential social instability if significant job displacement occurs among younger generations
- A call for global governance and a focus on socially desirable technological outcomes
>*Seen through that lens, AI is not troublesome in its own right, but rather whether it is positioned as inclusive or extractive. Today’s AI hyperscalers, he argues, fit the extractive mold almost perfectly: concentrated ownership, regulatory capture, and a business model that extracts data and attention at scale."
- Skepticism toward massive near-term AI productivity gains due to current model limitations
- The distinction between simple automation and true human-complementary tasks
- Potential social instability if significant job displacement occurs among younger generations
- A call for global governance and a focus on socially desirable technological outcomes
A new TechBrief from the Association for Computing Machinery's Technology Policy Council examines the rapid rise of agentic AI systems capable of planning and executing multi-step tasks autonomously. As adoption scales among enterprises and consumers, existing legal, regulatory, and technical frameworks are struggling to keep pace with these autonomous capabilities.
The report highlights several critical policy dimensions:
- Ambiguity in legal liability when an autonomous system causes harm without a clear accountable party.
- Serious security risks arising from the inability of models to distinguish between data and malicious commands.
- A lack of consumer transparency regarding agent permissions, authority, and recourse.
- Workforce disruption concerns where productivity claims have not been independently verified at scale.
The report highlights several critical policy dimensions:
- Ambiguity in legal liability when an autonomous system causes harm without a clear accountable party.
- Serious security risks arising from the inability of models to distinguish between data and malicious commands.
- A lack of consumer transparency regarding agent permissions, authority, and recourse.
- Workforce disruption concerns where productivity claims have not been independently verified at scale.
Philosophers Eric Schwitzgebel and Jeremy Pober argue that consciousness is likely not restricted to Earth's specific biological structures. By applying the concept of substrate flexibility, they suggest that sentient experiences could emerge from many different types of materials or chemical compositions found throughout the universe. This "Copernican principle of consciousness" challenges human-centric views by suggesting that various forms of complexity may lead to varied and non-humanoid types of conscious experience.
Key topics:
- Substrate flexibility and its role in potential sentience.
- The Copernican principle applied to consciousness to avoid terrocentrism.
- Implications for the diversity of extraterrestrial life forms.
- Philosophical considerations regarding different modes of AI intelligence.
Key topics:
- Substrate flexibility and its role in potential sentience.
- The Copernican principle applied to consciousness to avoid terrocentrism.
- Implications for the diversity of extraterrestrial life forms.
- Philosophical considerations regarding different modes of AI intelligence.
This encyclical letter from Pope Leo XIV examines the profound challenges posed by artificial intelligence and digital transformation to human dignity and social order. Drawing on the Church's Social Doctrine, the document provides an ethical framework for navigating the technological era, warning against a technocratic paradigm that seeks to reduce humans to data or optimize them for efficiency. It calls for a shared responsibility to ensure technology serves the common good rather than becoming an instrument of dominance or exclusion.
Main topics include:
- Safeguarding human dignity and ontological value in the digital age
- The need for transparency, accountability, and governance in AI
- Upholding truth as a common good against disinformation
- Protecting the dignity of work during digital transitions
- Guarding freedom from commercialization and algorithmic social control
- Addressing new forms of slavery within technological supply chains
- Building a civilization of love through justice and peace
Main topics include:
- Safeguarding human dignity and ontological value in the digital age
- The need for transparency, accountability, and governance in AI
- Upholding truth as a common good against disinformation
- Protecting the dignity of work during digital transitions
- Guarding freedom from commercialization and algorithmic social control
- Addressing new forms of slavery within technological supply chains
- Building a civilization of love through justice and peace
Anthropic CEO Dario Amodei warned at the World Economic Forum that rapid AI advancements are driving software costs toward zero, which could render many coding-based careers obsolete. He suggested that SaaS companies relying on code complexity as a competitive moat may face bankruptcy or significant market value losses. This prediction aligns with Anthropic's pursuit of a $900 billion valuation and its goal to position Claude as a replacement for the global knowledge worker wage bill.
Key points:
- Software is expected to become essentially free due to AI automation.
- Careers built around writing code may not survive the productivity shift.
- SaaS incumbents using complexity as a moat face high risks of going bust.
- Anthropic's strategic move toward replacing human knowledge worker wages with AI.
Key points:
- Software is expected to become essentially free due to AI automation.
- Careers built around writing code may not survive the productivity shift.
- SaaS incumbents using complexity as a moat face high risks of going bust.
- Anthropic's strategic move toward replacing human knowledge worker wages with AI.
This repository provides an implementation and recreation of the first published version of the Logic Theory Machine, also known as the Logic Theorist. Originally developed by Allen Newell, J. C. Shaw, and Herbert A. Simon in 1956, this program was designed to prove theorems in propositional logic using principles from Principia Mathematica. The project includes a Python-based interpreter for the IPL-I abstract machine language, tools to run the program against historical axioms and theorems, and utilities to analyze generated proofs.
Main components:
Implementation of the 1956 Logic Theory Machine
Propositional logic based on Principia Mathematica
Python interpreter for the IPL-I language
Tools for running simulations and verifying results
Main components:
Implementation of the 1956 Logic Theory Machine
Propositional logic based on Principia Mathematica
Python interpreter for the IPL-I language
Tools for running simulations and verifying results
The paper introduces LeWorldModel (LeWM), a stable Joint-Embedding Predictive Architecture (JEPA) that trains end-to-end directly from raw pixels. Unlike existing methods that rely on complex losses, pre-trained encoders, or auxiliary supervision to prevent representation collapse, LeWM uses only two loss terms: next-embedding prediction and Gaussian latent regularization. This approach significantly simplifies the training process by reducing tunable hyperparameters. The model is highly efficient, with approximately 15 million parameters capable of being trained on a single GPU within hours, and it offers planning speeds up to 48x faster than foundation-model-based world models while remaining competitive in 2D and 3D control tasks. Additionally, the latent space effectively encodes physical structures, allowing the model to detect physically implausible events through surprise evaluation.
As artificial intelligence continues to advance and outperform humans in specific tasks like mathematics or complex gaming, the question arises whether human cognition will remain unique. Tom Griffiths argues that intelligence is not a single linear scale but a multifaceted trait shaped by different constraints. While AI excels at processing vast amounts of data using scalable hardware, human intelligence is uniquely defined by biological limitations such as short lifespans and limited neural capacity. These constraints have forced humans to develop specific strengths in pattern recognition, social cooperation, and efficient learning from minimal experience. Ultimately, rather than seeing AI as a direct rival on all fronts, we should view it as a different kind of entity with its own set of capabilities and weaknesses.
- Intelligence is multifaceted rather than a single scale like height.
- Human intelligence is shaped by biological constraints such as lifespan and brain size.
- AI intelligence is driven by data volume, scalability, and machine communication.
- Different underlying architectures lead to different methods of problem-solving.
- Humans and AI are likely to be companions with distinct capabilities rather than total competitors.
- Intelligence is multifaceted rather than a single scale like height.
- Human intelligence is shaped by biological constraints such as lifespan and brain size.
- AI intelligence is driven by data volume, scalability, and machine communication.
- Different underlying architectures lead to different methods of problem-solving.
- Humans and AI are likely to be companions with distinct capabilities rather than total competitors.
