Category: AI
-
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
-
Could Multi-Billion Token Prompts Make Retrieval-Augmentation Obsolete?
The ability for large language models to understand and generate relevant responses hinges on the amount of context they can process at once—the “prompt window” size. For current models like GPT-4, this is limited to around 8,000 tokens. To compensate, techniques like Retrieval-Augmented Generation (RAG) have been developed. RAG models first retrieve relevant documents or
-
What is LOrA?
When it comes to Gen AI, efficiency and adaptability are key. Low-Rank Adaptation (LOrA), is enabling more optimized AI operation. What is LOrA?LOrA is a technique that allows us to adapt pre-trained AI models to new tasks or datasets without the need for extensive retraining. Instead of updating all parameters in a model, LOrA focuses
-
Empowering the Citizen Data Scientist with AutoML
Gen AI is commanding attention, but what about automating traditional machine learning, what about AutoML? Although it has been pushed a bit to the back burner of excitement due to Gen AI, a valuable functionality is now widely available in the industry – Automated Machine Learning (AutoML). Giving rise to the citizen data scientist, it
-
There are Security Risks with Gen AI but You can Control Them
As a leader, you can’t help but feel intrigued or maybe compelled by the concept that you most likely hear many times a day, generative AI. This technology is like a magic wand that can automate mundane tasks, spark creative genius, and supercharge productivity across your organization. But you’re also a level headed person. You
-
Why you should learn prompt engineering now, even if it will not be needed in the future …
Forrester predicted, a whopping 60% of employees will be trained in prompt engineering by 2024! (Forrester) But what exactly is prompt engineering, and why is it the skill everyone will soon be scrambling to master? Prompt engineering is the craft of designing queries that steer Generative AI models like ChatGPT and Claude to deliver highly
-
The Rising Value of Vector Indexes for Generative AI
As Gen AI solutions begin to become a common component of data ecosystems, one approach that’s quickly becoming a foundational feature is Retrieval-Augmented Generation (RAG), which combines the power of large language models with external knowledge retrieval. RAG systems work by first retrieving relevant information from a knowledge base (e.g., Wikipedia, corporate documents) and then