Introducing UrbanAI, your real-time AI Tutor with Visual Search and Retrieval
UrbanAI integrates Google search within your chatbot, with access to your friends' knowledge base, web, image and video understanding. In addition to indexing multimedia content and data types, you can also bookmark or delete conversations, highlight snippets to ask a follow up using custom instructions.
So what does a typical conversation look like?
ICML is an annual conference on machine learning. While ChatGPT is unaware of new papers accepted in 2023, a student or scholar may want to catch up on the latest AI research from Stanford:
The problem is that ChatGPT only has knowledge up to a cutoff point in 2021. You can manually provide documents or websites for it to parse, but it's laborious and unhelpful for discovery.
With UrbanAI, you can learn about papers presented at the conference, the workshop it was presented at, keypoints about the paper, and so much more:
1. Extract your site's images and text in seconds.
2. Up-to-date answers with images and citations.
3. Miss Google? Search instantly by highlighting.
4. Explain key ideas from the paper
5. But what are "heterogeneous curvatures"? Ask a follow up with a single click!
5b. You can also specify your follow up instructions
6. We offer many ways to help you master your learning journey.
-
Bookmark your favorite conversation.
-
Revisit conversations with a new perspective.
-
[Coming soon] Suggestive questions to further your curiosity
-
Coming soon: Global search and your good ol' Google portal
Features
- Up-to-date references: ingest websites, images, and PDFs (i.e. arXiv paper URL or uploads) on demand.
- Picture In-and-Out: GPT-4 answers with images and knowledge source / citations.
- Annotate & Elaborate: Google search on highlight, ask a follow up question using custom-defined template.
- Factual mode: retrieval augmented generation from you and/or your friends' data only.
- Developer marketplace to orchestrate and extend your knowledge workflows.
In our next release, you will be able to customize your LLM workflow with a selection of machine learning models, and create plugins with a marketplace that allows you to bring your productivity needs to any underlying language model.