VTeam AI

Happy Birthday ChatGPT !! How the AI space has evolved in the past one year

As ChatGPT commemorates its first anniversary, the landscape of AI has been markedly reshaped by a wave of new Large Language Models (LLMs) and innovative applications built around them. Amidst concerns and excitement, the AI industry has surged, with demand for AI expertise reaching unprecedented heights. The AI space, brimming with new entrants and pioneering technologies, now gazes ahead, pondering the next milestones in the Gen-AI odyssey.

On 30th November, ChatGPT will celebrate its 1st birthday. And what a year it has been to remember. A flurry of LLMs coming in; from big-tech giants to small-scale startups, everyone rushing for the Gen-AI race; Employees in different sectors fearing technology to take over, and whatnot. One thing is for sure, guys in the AI space had a great time and were in great demand as well. So what is this post about? To recap what major tech trends started in the AI space due to Gen-AI and where is the Gen-AI space heading right now? We will divide this post into the below segments

  • LLMs and lots of LLMs
  • Building Apps using LLMs
  • Autonomous AI-Agents


LLMs and lots of LLMs

The 1st big trend that the AI-Space saw was everyone trying to build an LLM which led to numerous models like GPT-4, PaLM, Llama, Falcon, etc. Even today, almost every day we hear a new LLM being trained and made open-source. We can divide all these LLMs into 3 major categories

Text generation or Q&A models: Large Language Models (LLMs) have transformed the field of Natural Language Processing (NLP) by excelling in various language-related tasks. Text generation and Q&A are some of the prominent applications of LLMs. These models can generate coherent and contextually relevant text based on the input they receive. Popular text generation LLMs include GPT-3, Flan, and BERT, each offering specific strengths in different contexts.

Multi-Model models: Multi-Model Language Models (LLMs) represent a significant advancement in the field of artificial intelligence, particularly in the domain of natural language processing. These models go beyond traditional single-task language models to simultaneously handle a variety of tasks and modalities, such as text, images, and more. They leverage multiple inputs and outputs, enabling a broader range of applications. Example: GPT-4

Domain-specific LLMs: Domain-specific LLMs are Large Language Models fine-tuned for particular domains or industries to enhance their performance in tasks relevant to that domain. They are pre-trained on a vast amount of general text data and then further fine-tuned on domain-specific data to understand and generate content specific to that field. Example: BloombergGPT, BioGPT, etc.

One problem that has been figured out in this last 1 year to use LLMs for production-level tasks is the deployment issues that might come up due to such big sizes. To cater to that, the companies are focussing on smaller models that are domain-specific in nature. 

Moving On,

Building Apps using LLMs

The next big concept that kept the AI world on its toes is how to use these LLMs for daily tasks i.e. building apps using LLMs. A few notable developments that have been used to build apps are


LangChain is an open-source framework designed for developing applications powered by large language models (LLMs). It offers several key features to facilitate the development of powerful language model-driven applications. Here are the features of LangChain explained pointwise:

  1. LLMs and Prompts Management:

    • LangChain simplifies the management of LLMs (Large Language Models) and prompts.
    • It provides tools to optimize and create a universal interface for LLMs.
    • Offers utilities for working effectively with LLMs.
  2. Chains:

    • LangChain supports the creation of chains, which are sequences of calls to LLMs or other utilities.
    • These chains enable developers to design complex workflows involving language models.
  3. Context-Aware Applications:

    • LangChain allows applications to be context-aware by connecting LLMs to sources of context, such as prompt instructions and few-shot examples.
    • Applications can ground their responses in relevant content to enhance contextual understanding.
  4. Reasoning with LLMs:

    • LangChain empowers applications to rely on language models for reasoning.
    • Developers can build applications that make informed decisions based on the input provided.
  5. Data Integration:

    • LangChain is data-aware and agentic, enabling connections with various data sources.
    • It facilitates the integration of data for creating richer, personalized user experiences


LlamaIndex, previously known as GPT Index, is a data framework designed for Large Language Models (LLMs) apps with a focus on ingesting, structuring, and accessing private or domain-specific data. Here are the features of LlamaIndex and LangChain explained pointwise:

LlamaIndex Features:

  1. Data Integration: LlamaIndex is primarily designed for integrating and managing private or domain-specific data into LLM applications.
  2. Data Framework: It serves as a data framework that enables LLMs to access and work with external data sources effectively.
  3. Tools for Data Integration: LlamaIndex provides a set of tools to facilitate the integration of private data into LLMs, enhancing the models' capabilities to work with domain-specific information.

Retrieval Augmented Generation :

The Retrieval Augmented Generation (RAG) framework is a machine learning approach that combines the capabilities of large language models (LLMs) with external knowledge or information retrieval systems. It enables LLMs to retrieve and incorporate external information to enhance their generative abilities. Here are the features of the RAG framework and LangChain explained pointwise:

RAG Framework Features:

  1. External Knowledge Integration: RAG allows LLMs to access external knowledge bases to retrieve facts, data, and information, improving the accuracy and informativeness of their responses.
  2. Fact-Based Generation: It emphasizes grounding generative processes in factual and up-to-date information from external sources, ensuring consistency and reliability in LLM outputs.
  3. Insightful User Interaction: RAG provides insights into the generative process of LLMs, giving users a better understanding of how model responses are derived.

Vector Databases

Vector databases, often referred to as vector DBs, are specialized database systems designed to efficiently store, manage, and query high-dimensional vector data. These databases are particularly useful in applications that involve vector-based data, such as machine learning, recommendation systems, image recognition, natural language processing, and more.

Now, after covering how building apps around LLMs came into the picture, let's talk about the last trend that has recently picked up i.e.

Autonomous AI-Agents

autonomous AI agents are a type of artificial intelligence system that possess the capability to independently understand objectives, create tasks to achieve those objectives, execute those tasks, adapt their priorities, and learn from their actions until the desired goals are reached. These agents exhibit a level of autonomy and decision-making that allows them to function without constant human intervention. Here are the key features and characteristics of autonomous AI agents:

  1. Objective Understanding: Autonomous AI agents can comprehend specific goals or tasks they are assigned, allowing them to work towards achieving these objectives.

  2. Task Creation: They have the ability to generate tasks and actions autonomously based on the defined objectives, effectively planning how to achieve those goals.

  3. Execution: These agents can execute the tasks and actions they create, taking concrete steps to work towards the desired outcomes.

  4. Adaptation: Autonomous AI agents are capable of adapting their priorities and actions based on changing conditions or new information, ensuring flexibility in their decision-making.

  5. Learning: These agents learn from their experiences and actions, enabling them to improve their performance over time by processing and analyzing data.

  6. Independent Decision-Making: They can make informed decisions based on their knowledge and objectives, often involving logic, statistical analysis, or machine learning algorithms.

  7. Feedback Loop: After taking action, autonomous AI agents can receive feedback from the environment. This feedback loop allows them to learn from their own experiences and adapt to new situations and environments.

A few major AI-Agents that have become popular 


AutoGPT is an advanced AI system that stands out due to its remarkable features and capabilities. Here are its salient features:

  1. Self-Prompting: AutoGPT can generate its own prompts and instructions, making it highly autonomous in task execution.

  2. Generative Power: It harnesses the capabilities of GPT-4 and GPT-3.5, large language models (LLMs), for performing tasks without the need for specific, predefined prompts.

  3. Learning Abilities: Like a smart assistant, AutoGPT learns from its own experiences, allowing it to continuously improve its performance and adapt to new situations.

  4. Diverse Applications: It is a versatile AI system that can handle a wide range of tasks, making it suitable for a variety of applications.

Baby AGI

Baby AGI is an innovative AI platform designed to train and evaluate various AI agents in a simulated environment. It replicates early cognitive development observed in infants, including perception, sensory processing, motor skills, and cognitive development. Its salient features include:

  1. Early Learning Emulation: Baby AGI aims to replicate the early learning capabilities observed in infants, making it a valuable tool for understanding cognitive development.

  2. Simulated Environment: It provides a simulated environment for training and evaluating AI agents, enabling research and experimentation.

  3. Multi-Faceted Cognitive Development: Baby AGI focuses on various aspects of cognitive development, including perception, sensory processing, and motor skills.

  4. Innovative AI Research: This platform is instrumental in advancing AI research by providing a unique focus on early-stage cognitive development.

Concluding this long post, in the ever-evolving landscape of artificial intelligence, the inception of ChatGPT marked a significant milestone. Since then, several noteworthy developments have unfolded, shaping the field of AI. Notably, other Large Language Models (LLMs) have emerged, each with its unique capabilities and applications. Google's Pathways Language Model 2 (PaLM 2) and the open-source multilingual LLM known as Bloom have gained attention in the LLM leaderboard

Furthermore, ChatGPT's success has paved the way for the rise of AI agents, pushing the boundaries of what machines can achieve. These AI agents, like ChatGPT, produce human-like text and can even tackle complex problem-solving tasks, making them valuable tools in various industries Additionally, frameworks like LangChain have entered the scene, combining ChatGPT's conversational capabilities with LangChain's language understanding expertise. This union creates a powerful force in the AI space, opening up new possibilities for AI-driven solutions across industries 

In conclusion, the legacy of ChatGPT extends far beyond its initial release. It has ignited a wave of innovation, propelling the development of various LLMs, frameworks, and the rapid rise of AI agents, each contributing to the ever-expanding potential of artificial intelligence.

Disclaimer: The views and opinions expressed in this blog post are solely those of the authors and do not reflect the official policy or position of any of the mentioned tools. This blog post is not a form of advertising and no remuneration was received for the creation and publication of this post. The intention is to share our findings and experiences using these tools and is intended purely for informational purposes.