Category: AI
-
Quantum computing will be the rocket booster to allow AI to scale to the unimaginable. 🚀
We have all been the situation, searching for a friend in a crowd in a sea of people. A classical computer (and you) would scan the crowd person by person … a real-life game of “Where’s Waldo” that could take forever. Quantum computing, however, is like having infinite versions of yourself checking every person at…
-
Is it time for RAG to RIP? Googles improved RIG (Retrieval Interleaved Generation) has some potential.
Generative AI is evolving fast, and if you’re leveraging Retrieval-Augmented Generation (RAG) like so many other organizations, it might be worth exploring the next potential evolution: Retrieval Interleaved Generation (RIG). Besides one letter, whats the difference? RAG retrieves information from external sources before generating a response. It collects relevant data upfront and then generates an…
-
Simple misunderstandings in communication are feeding AI hallucinations.
Isn’t it ironic that even among humans, our words can lead to misunderstandings? So, when we communicate with AI, it’s no surprise that similar challenges arise—especially when language barriers and ambiguities come into play. Consider the word “java.” To a programmer, it refers to a programming language. To a coffee enthusiast, it’s a cup of…
-
Is Enterprise GenAI Good Enough?
You’ve subscribed to the enterprise version of an LLM, so you have all the security and controls you need, right? Not exactly. An LLM gateway or acts as a custom-built bridge between your business and the LLM provider. It safeguards and monitors data flow, allowing you to filter and customize outputs, optimize performance, and enforce…
-
Numbers Game of GenAI
Trying to make sense of the numbers alongside GenAI models making your head hurt? 405B, 7B, 70B … lets try and clear things up. TDLR; those numbers are parameters. 7 Billion, 70 Billion parameters or adjustable levers to configure the model. The magic of LLMs lies in the math! Parameters are the Building Blocks of…
-
What does it mean to train in GenAI?
Training vs. Fine-tuning: The Educational Journey Training: This is the AI equivalent of a full education from kindergarten through PhD. We feed massive amounts of data to create a model from scratch. It’s time-consuming, computationally EXPENSIVE, and requires specialized expertise. True training is mostly done by organizations like OpenAI, Anthropic, and Meta. Fine-tuning: Think of…
-
Benchmarking Gen AI Models
Are You Choosing the Right AI Model? Discover the Benchmarks that Matter! Large Language Models (LLMs) have seen a rapid evolution in a short period of time. As these models continue to grow in complexity and capability, evaluating their performance through standardized benchmarks becomes essential. Benchmarks serve as standardized tests designed to evaluate and compare…
-
The Power and Peril of LLM Skeleton Keys: When AI Spills Its Secrets
Remember as a kid, when knowing the secret passcode could get you into the coolest clubhouse in the neighborhood? That thrill of exclusivity, the power of hidden knowledge – it was intoxicating. Fast forward to today, and we’ve stumbled upon something in the world of AI that evokes that same sense of discovery and excitement,…
-
In the battle of traditional vs. generative AI, who wins? Plot twist: it’s not a competition at all.
Artificial Intelligence (AI) has become a buzzword in recent years, with breakthroughs in generative AI capturing headlines and imaginations. However, it’s important to understand that AI encompasses a broad spectrum of technologies, including both traditional and generative approaches. These two branches of AI, while distinct, are not mutually exclusive and often work best when combined.…
-
Black Hat Prompt Engineering
Prompt engineering … a necessary business skill set that can be leveraged with bad intentions. What is Prompt Engineering?Prompt engineering is the art and science of crafting inputs, known as prompts, to effectively communicate with generative AI models. These models respond to textual inputs by generating human-like text based on the prompt’s instructions. In simple…