Author Archives: Ben Lorica

For those of us in AI, one of the most pressing issues is how the U.S.-led export controls on China’s chip industry will change when the Trump administration takes office in January 2025. The Biden administration has just announced sweeping new export restrictions targeting China’s semiconductor sector. These new measures limit the export of 24Continue reading “Circuit Breaker: America’s Semiconductor Defense Strategy”

As someone who works with Large Language Models (LLMs) daily, I’ve recently found myself captivated by the incredible advances in foundation models for visual data. While everyone’s talking about “multimodal” as the next big thing in AI, I want to share my excitement about systems that are already revolutionizing how we handle visual information. TheseContinue reading “Understanding Visual Language Models”

By Ben Lorica and Dean Wampler. In the rapidly evolving landscape of Generative AI, one of the most critical, yet often underestimated challenges is data processing and preparation. While models have become more sophisticated, the data pipelines feeding them have not kept pace, especially in leveraging heterogeneous computational resources like CPUs and GPUs. Traditional dataContinue reading “Paradigm Shifts in Data Infrastructure: Adapting to the Demands of Generative AI”

Teaching Machines to Value Companies Like a Wall Street Legend In the rapidly evolving landscape of artificial intelligence, a fascinating new frontier is emerging: AI assistants that aim to capture and replicate the expertise of world-class professionals. One notable example is the Damodaran Bot (DBOT), developed by researchers Vasant Dhar and Joao Sedoc at NewContinue reading “Digital Mentors: Building AI Systems That Think Like Experts”

Guardrails are safeguarding mechanisms designed to monitor, filter, and manage the inputs and outputs of Large Language Models (LLMs) to prevent undesirable or harmful outcomes. These protective boundaries ensure the safe, ethical, and controlled operation of LLMs, particularly when integrated into customer-facing applications. In practice, implementing guardrails involves creating programmable, rule-based systems or more advancedContinue reading “Guardrails Need to Be Integrated With AI Alignment Platforms”

Technical teams building AI applications with large language models (LLMs) face significant challenges in managing and scaling their projects. One major issue is the lack of rigor and structure in prompt engineering. Developers often embed prompts directly into code as simple strings or JSON objects, which becomes unmanageable as the number of prompts grows intoContinue reading “Seven Features That Make BAML Ideal for AI Developers”

The growth of generative AI is driving demand for AI Engineers across industries as organizations race to leverage this transformative technology. While this role isn’t entirely new, its scope and responsibilities are evolving rapidly with advancements like Large Language Models (LLMs) and other generative AI tools. To understand what this position means today from theContinue reading “What is an AI Engineer?”

In recent months, Chain-of-Thought (CoT) prompting has emerged as a popular technique for enhancing the capabilities of frontier models like Large Language Models (LLMs) and Large Multimodal Models (LMMs) in AI applications. CoT prompting encourages these models to generate intermediate reasoning steps before arriving at a final answer, effectively making the reasoning process explicit. InsteadContinue reading “When Chain-of-Thought Prompting Falls Short: Insights for AI Teams”