As we all know vector databases and AI is going to take over the business landscape.
Here is a question to the community. Do any of you have an idea of how we can implement this to take advantage of the coming gold rush.
I don’t think the AI chat part will be a viable solution. But maybe we can leverage our systems to upload the rows and documents to the vector databases where clients will be able to connect their chatgpt accounts to it once the weviate gpt app releases.
For those like me who had never heard of vector databases:
We all know how traditional databases work (more or less)—they store strings, numbers, and other types of scalar data in rows and columns. On the other hand, a vector database operates on vectors, so the way it’s optimized and queried is quite different.
In traditional databases, we are usually querying for rows in the database where the value usually exactly matches our query. In vector databases, we apply a similarity metric to find a vector that is the most similar to our query.
A vector database uses a combination of different algorithms that all participate in Approximate Nearest Neighbor (ANN) search. These algorithms optimize the search through hashing, quantization, or graph-based search.
Yeah sorry, so vector databases are going to be used to give LLM’s(ChatGPT) context about your company.
When you speak to ChatGPT now it has no context of what is going on in your company/life. With the new vector databases you will be able to upload files and information to it and ask ChatGPT questions about your own data.
Lets say you have a law firm and have a case file of 400 pages. You can upload it to the database and ask chatgpt questions about it or to summarize key parts.
It’s an interesting question, I assume you are talking about things like Pinecone?
It’s gonna be big, sure, but at the moment I don’t think we have a good way to integrate it.
Glide does have an OpenAI action to ask a question about a table, where they vectorize your table and retrieve the most relevant 5 rows to answer your question, that’s the closest thing we have.
If only we can do something like: Upload a file > Store vectors in a database > Be able to tell OpenAI to tokenize our question, and use knowledge in those vectors to answer the question, all through the API, then that would be great.
I imagine a world where Glide does this natively, we can enter our api keys and boom. Its set up.
Just Press CMD + K and ask your companies personal AI questions about your data.
I would kiss some feet if they could make this happen
On the one hand I find AI integrations in Glide exciting, on the other I hope we can all tread this path with caution.
Somehow I hope Glide will act responsibly and not blindly integrate things left and right.
Sorry for being a party pooper, but a big concern with the current AI race is if the downsides of everyone having nuclear and bio weapons in their pocket (AI) undermines its benefits (the same AI).
Of course there should always be caution with such powerful technology. OpenAI, or Midjourney, though how powerful they are, have had moments for us to think how we should approach this new era.
For me, I think the right word is “move at the right pace”. Don’t slow it down too much, or worse, give it a break, just keep it moving, but put the right guardrails to keep it in check.