The world altogether is grappling with how to apply AI in the most useful ways. We'll lay the groundwork for Torah to be applied to any existing application.
By developing datasets, models, and other infrastructure components, we will provide a platform agnostic plug and play solution to integrate with all available AI applications, as the industry develops.
AI-generated video and audio shiurim on a variety of topics, variety of styles, on-demand, interactive.
Books
Children's books with illustrations; novels in the style of Umberto Eco - but teaching a lesson. eBooks, print-on-demand - choose your own ending.
Video
Youtube or Social Media - Generate for all audiences, at all lengths, and for myriad applications. Parable takes on a life of its own.
Indices
Comprehensive indexes to help users navigate the vast Torah corpus. Dictionaries/Glossaries (think: Sefer Halikutim); encyclopedic works based on enriched texts.
Articles
Blog posts, feature publications, etc. on user interests and current events - relevant and engaging content rooted in Torah.
Audience-specific summaries. Rashi and Tosfos which know the user. Rashi on Chumash systematically explained.
Beyond
Stock images, Music generation, Digitization/OCR. The sky really isn't the limit anymore: just an overflow of good content.
Beis Medrash of Bots
1
Tagging
Contextual tagging to provide deeper understanding and connections.
Automated definitions and explanations of key terms and concepts. Knowledgebasing internally as well as user-facing.
2
Mapping
Identifying textual and subject matter connections and relationships.
Comparing and contrasting: what was added in this piece? Which is primarily cited? And so on.
3
Structuring
Applying the learning methodology that a person would, to organize and 'enrich' the content in a meaningful way.
We can perform a full audit on Torah; we can think anything which hasn't been thought yet - and according to the rules of Torah
Structured AI for Torah
Aligning machine learning with Torah learning. Torah is the most structured, scientific corpus: strictly governed and exceptionally peer reviewed.
Ensuring accuracy foremost.
Structuring data and queries
Mapping / codifying 'Klalei Hashas' and so on
Audit trail
Structured insights.
1
Not displaying books; this is to write books.
Quality and quantity which the world has never seen.
2
Compatibility for Torah
Bringing Torah off of the shelf, and creating an accessible frontend - and backend…
3
As much an art as a science
Interfacing and connecting Torah and the Jewish people.
Applying Rashi's rules, the Rambam's rules, the Klalei Hashas — and the Rebbe's methodologies, at scale and where users are.
Editing Content
Editing
Editing human posts and articles before publication - not just for grammar but for content too.
Fact-checking
Revise the entire Wikipedia for accuracy. Audit news articles and produce real objectivity.
Tools for Teaching
Personalized
Curriculum tailored by level, from advanced to Hebrew school.
No one size fits all. Not even a classroom setting.
Interactive
Interactive lessons that adapt to the user's needs and learning style.
Analyze progress, identify areas which may need additional support, and adjust the lessons accordingly. One-on-one, audio-visual: the most out of a learning experience.
Plain Old Search?
Vector Search
The very advent of search is a huge contribution, and isn't yet embraced in full (and far from it). Detailed search filters and algorithms to provide relevant results for both academic and lay users.
Search becomes more than keywords: we're talking about real contextual appreciation - finding relationships and correlations on the idea level. Discovering search results, rather than retrieving.