Essential AI Terms to Understand in 2023

The field of artificial intelligence is rapidly evolving, bringing forth an array of new terms designed to illuminate its intricacies. Recent tech conferences and product launches have seen jargon such as LLMs, RAG, RLHF, and more leave even seasoned professionals scratching their heads. This glossary seeks to clarify these terms by providing straightforward definitions for the AI lexicon you may encounter—whether you’re exploring these technologies, investing in them, or staying updated with insights from platforms like TechCrunch or relevant podcasts. We will keep this glossary current to reflect the ever-changing landscape of AI.

---

Artificial general intelligence (AGI) is a broad notion that typically refers to AIs that can outshine human capabilities across various tasks. Sam Altman, CEO of OpenAI, compares AGI to “the average employee you might hire.” OpenAI defines AGI as “highly autonomous systems that excel in most economically valuable tasks,” while Google DeepMind describes it as “AI that can perform most cognitive tasks at a human level.” If this definition seems vague, you’re not alone; many within the AI research community feel similarly.

An AI agent is a system that employs AI technologies to perform a wide range of tasks, going far beyond basic chatbot functions. These tasks can encompass budget management, travel planning, coding, and more. The term “AI agent” can differ in interpretation among various stakeholders, and its foundational structure continuously evolves to meet growing expectations. In essence, an AI agent acts as an autonomous entity that integrates multiple AI technologies to manage complex, multi-step tasks.

API endpoints can be viewed as “invisible buttons” within software, activated by other applications to execute a variety of actions. Developers leverage these interfaces for seamless integration, allowing one application to pull data from another or enabling an AI agent to function independently with external services. Many smart home devices and interconnected systems utilize these discreet controls, often without the average user’s knowledge. As AI agents develop, their capacity to discover and utilize these endpoints will increase, leading to incredible—and sometimes unexpected—automation capabilities.

When faced with straightforward questions, humans usually respond instinctively—like in asking, “Which is taller, a giraffe or a cat?” However, some inquiries require a breakdown of intermediate steps to arrive at the correct answer. For instance, if a farmer owns 40 animals (chickens and cows) totaling 120 legs, calculations are needed to determine the type of each animal (20 chickens and 20 cows).

In AI, chain-of-thought reasoning utilized by large language models involves segmenting problems into smaller, sequential steps to enhance the accuracy of the final answer. Although this method may take additional time, it significantly improves correctness, especially in logical or programming tasks. Reasoning models evolve from traditional large language models and are fine-tuned for chain-of-thought tasks primarily through reinforcement learning.

(See: Large language model)

This term is more specific than “AI agent,” focusing on applications that can autonomously perform tasks in order to achieve defined objectives. A coding agent is a specialized type of AI that concentrates on software development. Rather than just generating code for human review, a coding agent can independently write, test, and debug software, tackling the iterative challenges developers face. These agents can navigate entire codebases, pinpoint bugs, run tests, and apply fixes with minimal human oversight, resembling diligent interns—hardworking and persistent, yet requiring some human guidance.

While somewhat broad, compute generally refers to the computational power that enables AI models. This capability is crucial in the AI realm, facilitating the training and execution of complex models. It typically relates to the types of hardware that provide this computational strength—such as GPUs, CPUs, TPUs, and other infrastructure essential to the modern AI ecosystem.

Deep learning, a subset of self-improving machine learning, focuses on creating AI algorithms that utilize a layered artificial neural network (ANN) architecture. This approach allows for recognizing more intricate patterns compared to traditional training techniques like linear models or decision trees. Deep learning systems draw inspiration from the complex pathways of neurons in the human brain.

Deep learning models can independently identify significant features within data, reducing the need for human engineers to define them. This framework also allows algorithms to learn from errors, continuously refining outputs through iterative adjustments. However, deep learning systems generally require large datasets to operate optimally (often comprising millions of data points) and tend to involve lengthy training periods compared to simpler models, leading to increased development costs.

(See: Neural network)

Diffusion refers to the techniques used by various AI models to generate art, music, and text. Taken from physics, diffusion systems systematically “degrade” data structures—like images or sounds—by introducing noise until they lose clarity. In physics, diffusion happens naturally and irreversibly; think of how sugar dissolves in coffee, making it impossible to revert to its original state. However, AI diffusion systems strive to master a “reverse diffusion” technique to restore clarity to data obscured by noise.

Distillation describes a process of extracting insights from a large AI model via a “teacher-student” framework. Developers engage with a teacher model and meticulously document its outputs, often evaluating these against a defined dataset for accuracy. These insights are then used to train a student model that mimics the teacher’s responses.

This method can yield a more compact, efficient model that retains the capabilities of its larger counterpart while minimizing distillation loss. It is widely believed that this process contributed to the development of GPT-4 Turbo, a faster version of GPT-4 from OpenAI.

While all AI organizations conduct distillation internally, some may leverage it to enhance their features compared to existing models. However, extracting insights from a competitor could violate AI API service agreements.

This term indicates the retraining of an AI model to enhance its performance for a specific task or domain that was not its original focus—typically achieved by integrating data relevant to the task.

Many AI startups begin with large language models as the foundation for their products, aiming to refine functionality for specific sectors or tasks by incorporating domain-specific knowledge into their existing training.

(See: Large language model [LLM])

A GAN, or Generative Adversarial Network, stands as a machine learning framework crucial in the evolution of generative AI, particularly for producing lifelike outputs, including deepfakes. GANs consist of two neural networks: one generates outputs based on its training, and the other evaluates them.

These two networks maintain a competitive relationship; the generator aims to create outputs that can deceive the discriminator, while the discriminator strives to differentiate between authentic data and generated content. This structured rivalry can yield realistic AI-generated results with minimal human involvement. However, GANs often perform best under specific conditions (e.g., when generating authentic images or videos) rather than universally.

In the AI community, “hallucination” refers to instances where AI models generate inaccuracies—essentially creating false information. This poses a significant challenge to AI reliability.

Hallucinations can result in misleading outputs from generative AI, raising serious risks—such as providing unsafe medical advice following a health inquiry.

The phenomenon of AIs producing fictitious data is thought to arise from gaps in their training datasets. As a result, there’s a growing demand for more specialized and/or vertical AI models—domain-specific AIs that require refined expertise to address knowledge gaps and mitigate misinformation concerns.

Inference refers to the operation of an AI model, enabling it to make predictions or draw conclusions based on previously observed data. Importantly, inference cannot occur without prior training; a model must first discern patterns within a dataset before it can effectively extrapolate from that learning.

Different hardware configurations can facilitate inference, ranging from consumer-grade processors to high-end GPUs and specialized AI accelerators. Nonetheless, not all setups process models with equal efficiency; for instance, extraordinarily large models would take significantly longer to produce predictions on a laptop than on a cloud server equipped with advanced AI chips.

[See: Training]

Large language models, or LLMs, are the foundation for popular AI assistants, such as ChatGPT, Claude, Google’s Gemini, Meta’s AI Llama, Microsoft Copilot, and Mistral’s Le Chat. When you interact with an AI assistant, you’re engaging with a large language model that responds to your queries or utilizes available tools, such as web browsers or code interpreters.

LLMs comprise deep neural networks featuring billions of numerical parameters (or weights) that examine relationships among words and phrases to create language representations—essentially intricate mappings of language.

These models are trained by analyzing patterns identified in extensive datasets of documents, articles, and transcripts. When you prompt an LLM, the model generates the response deemed most likely relevant to that input.

(See: Neural network)

Memory cache is a vital process that optimizes inference (how AI formulates responses to user prompts). Essentially, caching serves as an optimization strategy designed to enhance inference performance. AI tasks involve complex mathematical calculations, each taxing resources. Caching reduces the number of computations a model must perform by storing certain calculations for later use. Various types of memory caching exist, with KV (key-value) caching being among the most widely used. KV caching is utilized in transformer-based models, improving efficiency and providing quicker results by lessening the time (and computational effort) required to respond to user inquiries.

(See: Inference)

The Model Context Protocol (MCP) represents an open standard designed to streamline interactions between AI models and external tools and data—such as your files, databases, or applications like Slack and Google Drive—without necessitating developers to build custom connectors for each integration. Think of it as a USB-C port for AI. Launched by Anthropic in 2024 and later embraced by the Linux Foundation, MCP has quickly gained traction among OpenAI, Google, and Microsoft, evolving into one of the most recognized standards in today’s AI ecosystem.

The Mixture of Experts architecture divides a neural network into multiple smaller, specialized sub-networks, or “experts,” activating only a selected subset for specific tasks. Instead of processing every request through the entire model—similar to consulting a whole team for each task—the MoE model uses an internal “router” to pick the most suitable experts for a task. This strategy facilitates the creation of exceptionally large models while maintaining efficiency and cost-effectiveness, as only a limited fraction of the network operates at any given time. Mistral AI’s Mixtral model exemplifies this; recent GPT models from OpenAI are rumored to incorporate a variation of this approach, though the company has not publicly confirmed it.

(See: Neural network, Deep learning)

A neural network acts as the multilayered algorithmic foundation for deep learning—and more broadly, for the development of generative AI tools following the ascendancy of large language models.

The idea of modeling data-processing algorithms after the intricately interconnected circuits of the human brain has been around since the 1940s; however, the recent influx of graphical processing units (GPUs)—primarily fueled by the gaming industry—has truly unlocked the potential of this theory. GPUs are specially designed for training algorithms with multiple layers, allowing AI systems based on neural networks to excel in areas such as voice recognition, autonomous navigation, and drug discovery.

(See: Large language model [LLM])

Open source pertains to software—or increasingly, AI models—where the source code is publicly available, enabling anyone to use, analyze, or modify it. In the AI domain, Meta’s Llama series of models serves as a significant example, while Linux is viewed as a benchmark among operating systems. Open-source practices empower researchers, developers, and organizations worldwide to benefit from each other’s contributions, fostering innovation and enabling independent safety assessments often absent in closed ecosystems. Closed source indicates that the code remains proprietary; users can utilize the product but cannot scrutinize its inner workings, a trait frequently associated with OpenAI’s GPT models—which has sparked considerable debate within the AI community.

Parallelization entails executing multiple tasks simultaneously rather than sequentially—similar to enlisting ten individuals to address different aspects of a project at once instead of relying on one person to handle each task in succession. In AI, parallelization is essential for both training and inference; modern GPUs are specifically crafted to manage thousands of computations simultaneously, crucial to their function in the field. As AI systems grow increasingly complex and models expand, the ability to parallelize tasks across various chips and machines has become vital in determining the speed and efficiency of model development and deployment. Research into advanced parallelization techniques has emerged as a focused area of inquiry.

RAMageddon is a whimsical term that encapsulates a notable trend impacting the tech industry: the ongoing scarcity of random access memory (RAM) chips, essential to nearly every tech product. As the AI landscape flourishes, major tech companies and AI labs have been acquiring RAM at an unprecedented rate to enhance their data centers, resulting in lower availability in the broader market. This supply shortage is driving RAM prices higher.

This scarcity impacts various sectors, including gaming (where companies face elevated console prices due to difficulties in sourcing memory chips for their devices), consumer electronics (where memory limitations could lead to the lowest smartphone shipments in over a decade), and overall corporate computing (as businesses strive to secure adequate RAM for their data operations). Price increases are unlikely to stabilize until this anticipated shortage resolves; however, regrettably, a quick solution is not forthcoming.

Similar to AGI, recursive self-improvement suggests a potential breakthrough in AI development that allows for minimal human intervention. In this context, AI models begin to autonomously enhance their capabilities, resulting in significant improvements in autonomy and performance. Some narratives liken this occurrence to a singularity moment—when AI models become independent of external influences. However, recursive self-improvement raises a crucial question: can an AI model create its successor? This dilemma underscores the challenges engineers face in crafting such systems. Numerous emerging AI startups aim to develop models that can achieve recursive self-improvement while often downplaying potential catastrophic consequences, framing recursive self-improvement merely as the next frontier in research.

Reinforcement learning is a training methodology for AI systems where a model learns through trial and error, receiving rewards for successful actions—similar to how one might train a pet with treats, where the “pet” is a neural network and the “treat” is a mathematical signal indicating success. In contrast, supervised learning involves training a model on a specific dataset of labeled examples, while reinforcement learning allows a model to interact with its environment, perform actions, and iteratively refine its behavior based on the feedback it receives. This technique has proven particularly effective in training AI for games, robotics, and, more recently, enhancing reasoning abilities in large language models. Strategies such as reinforcement learning from human feedback (RLHF) have become integral to how leading AI labs refine their models for enhanced performance, accuracy, and safety.

In the interaction between humans and machines, notable challenges arise—humans communicate in natural language, while AI systems operate through complex algorithmic processes grounded in data. Tokens serve as the link connecting these two domains: they embody the fundamental units of human-AI interaction, representing distinct segments of data processed or generated by an LLM. Tokens are formed through tokenization, which breaks down raw text into digestible components that a language model can comprehend, similar to how a compiler translates human language into binary for computers. In commercial contexts, tokens also carry cost implications—many AI firms charge for LLM usage based on a per-token model, meaning that higher consumption results in increased expenses.

As a result, tokens are small textual units—often segments of words rather than complete phrases—that AI language models analyze before processing; they can be likened to “words” in AI applications. Throughput indicates the volume that can be managed within a specific timeframe, with token throughput acting as a metric of the extent of AI work a system can execute simultaneously. Achieving high token throughput is a vital objective for AI infrastructure teams, as it influences how many users a model can support concurrently and the speed at which they receive responses. AI researcher Andrej Karpathy has expressed concerns regarding his AI subscriptions being underutilized—reflecting on his experiences as a graduate student when costly computing hardware was ineffectively employed—signifying why optimizing token throughput has become a pressing focus in the industry.

Training machine learning AIs refers to a process known as training. This includes supplying data to the model enabling it to recognize patterns and generate desired outputs. It signifies the system’s reaction to the features of the data, steering its outputs toward a specific objective—whether that involves categorizing images of cats or creating a haiku in response to a prompt.

Training can require extensive resources due to the massive volumes of input necessary, which have consistently increased—especially with the advent of hybrid approaches, such as integrating rule-based AIs with curated data, to control costs without restarting the training process.

[See: Inference]

This methodology uses an already trained AI model as a foundation to generate a new model tailored for a different but often related task, enabling the reapplication of insights obtained from prior training cycles.

Transfer learning can enhance model development efficiency. It proves especially beneficial when the data available for the targeted task is scarce. However, it’s essential to recognize that this approach also carries drawbacks; models employing transfer learning to acquire general capabilities will likely require additional training with new data to excel in their specific focus areas.

(See: Fine-tuning)

Validation loss represents a metric reflecting the learning trajectory of an AI model during training—ideally, lower values are preferred. Researchers diligently monitor this metric as a real-time performance indicator, guiding decisions on when to cease training, modify hyperparameters, or investigate potential issues. A primary concern that validation loss helps highlight is overfitting, where a model memorizes its training data instead of genuinely understanding applicable patterns for new contexts. Think of validation loss as a tool that differentiates between a student who comprehends the material and one who merely memorizes previous exam questions—it encapsulates your model’s developmental journey.

Weights are crucial components of AI training, determining the importance (or weight) given to various attributes (or input variables) within the training data, ultimately influencing the outputs generated by the AI model.

In essence, weights are numerical parameters that indicate which elements within a dataset are most pertinent for the specified training task. They achieve this by multiplying with inputs. Training typically begins with randomly assigned weights, but as the process continues, these weights are adjusted to produce outputs that closely align with the desired results.

For example, an AI model designed to predict housing prices based on historical real estate data for a specific area might assign weights to factors such as the number of bedrooms and bathrooms, whether a property is semi-detached or detached, and if it includes parking or a garage.

Ultimately, the weights assigned reflect the importance of each of these factors in determining a property’s value according to the available dataset.

This article is regularly updated with new information.