Arvind Jain (Arvind Jain), CEO of $7.2B Glean, aka the knowledge layer for enterprise AI that recently crossed $300M ARR.. says: "Open source models are gonna dominate AI inferencing." "You'll see in the next two years we'll shift to where it's almost no open source to where it's gonna... be almost all open source." The 2 biggest needs for enterprise AI:show more

@glean 1.) Context: "All these agents that you want to automate the work that humans do, they need that context, that data, that information that humans use to do the same work."

@glean 2.) Proving ROI: "People keep complaining they can't measure the return on investment. "Where is the business value coming from?" "We can pick the right model for the right task, which is cheaper for them, but still gets the work done."

@glean "With Glean, they get that context in one shot, so we actually make most of your AI workloads much faster and very cheaper."

@sourceryy @jainarvind @glean I’ve been bullish open source but math changes a bit with spark 1.1. Less reason to use a chinese model

@jainarvind @glean we consistently discount how market participants will respond to competition. frontier labs are not going to give up the low-cost market forever. they own the full-stack and use that to be price competitive. saw the first instance of this with meta/grok recently

@jainarvind @glean Makes sense once inference cost matters more than training novelty. Open weights you can run on your own hardware start winning on margin alone, especially for companies doing high volume low complexity calls where a frontier model is overkill anyway.

@jainarvind @glean Let's see how this tailwind for open weight models turns out 👀

@jainarvind @glean the founders i talk to care about two things: can they keep their data secure and can their teams actually adopt it. everything else is downstream

@jainarvind @glean Really good = not really good

@jainarvind @glean That started happening at the edge 6 months ago. But on linked in people can't get enough of claude.

@jainarvind @glean Right. Apps like Glean betting on models losing with value migrating to the layer that decides which model runs and what context it sees because that’s where switching costs and proprietary data actually accumulate

@jainarvind @glean i didn't know abed from community was a unicorn tech ceo now.

@jainarvind @glean Great take.