ALL INPredictions
Last updated Nov 29, 2025
Prediction
Sacks
← Back to hostEpisode
Sacks @ 00:38:15Right
aitech
Future AI models used within Glue (and similar enterprise chat tools) will become capable of automatically inferring which internal data source or repository to query, without users having to explicitly specify it, making AI assistance more seamless than the May 2024 state.
in the future it's going to figure it out on its own. So it's going to become more and more seamless, but.View on YouTube
Explanation

By late 2025, several enterprise chat/assistant products are doing essentially what Sacks predicted: models automatically route to the relevant internal data sources without users having to name a specific repository each time.

  1. Glue itself has moved in this direction. Glue AI now:

    • Sits in every conversation and "leverages context from your workspace" (threads, attachments) and web data when you simply @‑mention it and ask a question, rather than you pointing it to a particular thread or file each time.
    • Is described as using context from "threads, attachments, and connected apps (with permission)" to answer questions, implying it can pull from whatever connected sources are relevant rather than requiring the user to specify which one on each query. (docs.glue.ai)

  2. Glue added MCP-based integrations that let the model pick the right external tool. In 2025 Glue shipped MCP support and a curated directory of MCP servers for apps like Linear, Asana, Notion, Sentry, Stripe, Vercel, etc. In use, you can just ask Glue AI in a thread to do things like "create a Linear issue" or "look up data from an internal system," and the system invokes the appropriate MCP server/tool behind the scenes rather than you manually choosing an API endpoint. (docs.glue.ai)

  3. Anthropic’s Claude Enterprise Search is a clear realization of the prediction in a “similar enterprise chat tool.” Claude’s 2025 Microsoft 365 connector and Enterprise Search feature are explicitly designed so that one question to Claude can search across a company’s connected data sources–SharePoint, OneDrive, Outlook, Teams, and other apps–in a single shot. Anthropic’s own example: ask about your company’s remote‑work policy and Claude automatically pulls from HR documents in SharePoint, email discussions in Outlook, and team guidelines from various sources into one answer. Users no longer have to say "now search SharePoint" vs. "now search Outlook"—the model infers which repositories to query based on the question. (anthropic.com)

  4. Broader ecosystem trend confirms the direction. Microsoft 365 Copilot’s Researcher agent similarly leans on Microsoft Graph to draw from meetings, emails, chats, and SharePoint files, with users mainly expressing what they want rather than formulating per‑repository queries—evidence that mainstream enterprise assistants are converging on automatic source selection. (techcommunity.microsoft.com)

Given these shipped capabilities in Glue and closely related enterprise assistants like Claude for Enterprise and Microsoft 365 Copilot, the world has clearly moved toward exactly what Sacks described: AI models in enterprise chat that mostly “figure out on their own” which internal data sources to hit, making assistance more seamless than it was in May 2024. Therefore, the prediction is best classified as right.