Friso W.
344 posts
@TheSpickle
COO of @build_vertical. Ex Global Head fin. Services Daraz (Alibaba Group). Co-founder Koko Group.
1) The Vertical Gateway A decentralised web app that serves as the hub for all $VERTAI utilities. We strongly value these utilities, both to support our token economy and to incentivise holders to grow our web2 ecosystem. Here's what we're building 👇
The countdown has begun... In 13 days, we’re launching our new product Vertical Motion ⚡️ The first AI Motion Video Director - a game-changer for video creators 🎥 Join the waitlist and be among the first to get access to the platform - FCFS! Link in the comments 👇
You have no idea what’s coming The Vertical ecosystem, its utilities, & loyalty rewards will all be accessible through the Gateway Stay tuned 😏
Why Vertical Knowledge is our AI moat ⤵️ In short: Vertical Knowledge Graphs (VKG) lets AI models keep learning from new data without retraining. That makes your AI faster, cheaper, and more up-to-date than traditional methods. So what is it, and why is it different? 🔹Fine-tuning = static You train on periodic data snapshots. Updating takes GPUs, time, and a release cycle. 🔹RAG = reactive Fetches info from a vector DB but doesn’t truly understand or connect it. Retrieval, not reasoning. 🔹VKG = living Continuously links entities, facts, and media into an evolving graph. The model reasons through it and picks up new info instantly. How it will work in the Studio: 1️⃣ Ingest: Drop in your text, tables, images, audio and/or video. 2️⃣ Link: The graph connects who, what, where, when, why across sources. 3️⃣ Reason: The model follows these paths to explain, verify, and cite. 4️⃣ Learn: User feedback adjusts the graph’s edges and weights, not the model’s full parameters. Why this matters for @build_Vertical users: 🔸Faster time-to-market 🔸Lower cost 🔸Explainability 🔸Multimodal by default 🔸Domain-agnostic Real-world use cases: ▫️Support & Sales Reflect product or price changes in real time, with source citation. ▫️Compliance Laws and policies stay current, with audit trails ▫️Ops & R&D Link incidents, commits, metrics for root-cause insights. ▫️Education Connect lessons to videos, chapters, and quizzes seamlessly. So, how does this compliments fine-tuning? Fine-tuning is still great for style, tone, or task-specific behaviors. But for a knowledge refresh, VKG is far superior for adding new knowledge to your LLM system, especially compared to RAG. It shifts updates from “train the model” to “update the graph”. It lets every model inside Vertical keep learning efficiently, continuously, and infinitely. Welcome to the era of living AI ⚡️
JPMorgan is allowing institutions to use crypto as collateral. it includes 2 assets. one is bitcoin, the other is ethereum. eth institutional adoption has begun.
