Author Archives: Ben Lorica

The ability of a machine to reason—not merely regurgitate information but to engage in structured, logical, multi-step problem-solving—is swiftly emerging as a key trait of the most advanced large language models (LLM). We are transitioning from models that simply mimic patterns to those that can genuinely think, deconstructing complex challenges into a series of interpretableContinue reading “Beyond Imitation: How Reinforcement Learning is Reshaping AI Reasoning”

The start of a second Trump administration is prompting speculation about the future of technology policy in the United States. This analysis explores potential shifts across various sectors, from artificial intelligence and cryptocurrency to government efficiency and national security. The predictions outlined here are based on a combination of past statements made by individuals associatedContinue reading “AI and Tech under Trump 2.0”

Inside the High-Stakes World of Generative AI Red Teaming We’re at a pivotal moment for GenAI: these systems are evolving at an astonishing pace, integrating seamlessly into our daily routines, yet bringing with them a host of new security and safety challenges. That’s why I find Microsoft’s recent study—where its AI Red Team conducted extensiveContinue reading “Red Teaming AI: Why Rigorous Testing is Non-Negotiable”

In the wake of the export controls that shook the semiconductor industry, as detailed in our earlier piece, Circuit Breaker: America’s Semiconductor Defense Strategy, the U.S. government has now turned its attention to frontier AI technologies. The Commerce Department’s Bureau of Industry and Security (BIS) has published a comprehensive policy aimed at regulating advanced AIContinue reading “Navigating Washington’s New Frontier on AI Controls”

Meta today unveiled a significant shift in its content moderation strategy, discontinuing its third-party fact-checking program in the United States and introducing a “Community Notes” model for verifying and contextualizing user posts. Simultaneously, Meta will ease restrictions on certain political speech, now deemed part of mainstream discourse, while maintaining stricter enforcement for illegal content andContinue reading “The Moderation Dilemma: A Balanced Look at Fact-Checkers vs. Community Notes”

DeepSeek’s approach to AI underscores that high-performance large language models do not have to be prohibitively expensive or proprietary. By combining open-source development with resource-optimized techniques like Mixture-of-Experts architectures, FP8 mixed-precision training, and Multi-Token Prediction (MTP), DeepSeek-V3 demonstrates a robust and efficient path forward for teams of varying sizes.  Having followed coverage of DeepSeek-V3 acrossContinue reading “DeepSeek: What You Need to Know”

A noteworthy development in AI this year is the introduction of o1, an OpenAI model that stands out for its reasoning abilities and its approach to complex problem-solving. Although many Large Language Models (LLMs) are capable of producing coherent text, o1 stands out by going beyond self-supervised learning, integrating reinforcement learning, employing sophisticated search algorithms,Continue reading “Inference Scaling for Enhanced Reasoning”