The long road to this paper began at UCSF, survived the disruptions of the pandemic, a lab move to MIT, and many more obstacles along the way. We are deeply proud of this work. 2/
https://t.co/scS4CkWXDU
- Hallucination on provided documents/context is a non-issue once customers test the product, assuming each query is limited in scope to a small number of documents. There is no reasonable support for "summarize these 100 proprietary documents" in today's LLM world.
3 months after soft launching https://t.co/oxjFP7vw9o we've learned a lot about early customer interest in LLMs. Here are some takeaways:
- Customers are willing to pay for "traditional" SaaS features like user management, SSO, and automatic document syncing.
- Consumers of documents are sometimes more excited about LLM tech than the document owners. Ex. sales teams would love access to a chatbot of their product roadmaps; the PMs are more cautious.
- Users are surprisingly tolerant of slow chat response times.
Today we're announcing our $14M Series A led by
@davidu and @kimberlywtan at @a16z
We're hiring across the board, come join the team!
🙏 @Kyle_L_Wiggers@TechCrunch for telling our story
https://t.co/frxiHKhMJ0
Sentient ingests documents from Google Drive, Notion, Confluence, etc. and has a web interface and Slack app.
It's built using GPT-3 but will switch over to ChatGPT once that is API available.
Here's the demo: https://t.co/qPFTFzxFAe
I built a chatbot for company documents.
Ask questions about HR policies, product strategy, market positioning, and get immediate answers.
Give it a whirl: https://t.co/oxjFP7w3YW
My thesis is that every company will soon have an AI “librarian” for its employees to interface with.
As a PM I constantly had to answers questions that were already well-documented - I know my colleagues in HR, Sales, etc felt the same. Hoping Sentient can help!
🚨Biggest @langchain drop yet🚨
⚡️Dynamic, zero-shot composition of multiple chains
Easily plug in different subchains (google, REPL, wikipedia, DBs) just by telling the router LLM (in English!) when and how to use them
`pip install langchain==0.0.8`
https://t.co/vxBzH0v5F9
TL;DR: SAMs break current user interface dichotomy of GUIs for people / APIs for third parties, shifting control of how a user accesses a service from the company to the user.