I feel like every example shown in the course is currently an easy to use feature on ChatGPT…
- uploading documents for Q&A
- Context History / Memory (Projects)
- Chaining together several prompts to get to a final output
- Prompt templates (Custom Instructions)
- Search the web
- etc…
Was this course simply released before all of these features were launched by OpenAI and now they are obsolete?
or is LangChain’s main value proposition that you can choose your LLM and are not forced to only use ChatGTP??
please let me know if I’m missing something, I don’t want to seem ungrateful for such a cool open source project, I’m just trying to understand if i’m missing a key piece of why LangChain is valuable
thank you 
LangChain can help you when making your own project that can do something that ChatGPT does not do, like RAG (retrieval augmented generation, essentially a natural language interface to search your docs) or an agentic system. If you don’t like the ChatGPT WebUI, you can make your own (for example, ChatGPT WebUI does not have a feature to search over your old conversations.)
The main added value of LangChain is indeed that they provide adapters so you can plug and play with various LLMs and vector databases.
2 Likes
Ok understood, I really appreciate your reply
Thank you for your reply @Paul_Baclace. I have not yet waded into LLM applications territory and feel this was a well stated reply.
@Paul_Baclace
according to my understanding gpt4 and gpt 4o does use RAG technique and also stores conversion in server database.
The main difference although it is true langchain provides natural language processing inputs and outputs but one needs to have good understanding about customisation with langchain where as chatgpt built in llm using gpt fine tuned models using supervised and reinforced learning human feedback (rlhf)
You probably can do more customisation with langchain surely than chatgpt as one is open source framework to create llms and latter is built in llm with open ai.