#AIStack

Dr. Thompsonrogt_x1997
2025-05-26

🚀 Think you know how GenAI works?
It's not just prompts. It’s orchestration with 25 next-gen Python libraries powering the world’s smartest LLMs: agents, memory, RAG, grammar control & more 🧠💻
Dive into the tools under the hood of AI in 2025 👇
🔗 medium.com/@rogt.x1997/25-grou

medium.com/@rogt.x1997/25-grou

Dr. Thompsonrogt_x1997
2025-05-21

🚀 $3.83 Billion by 2025? That’s the MLOps market powering GenAI at scale.
Discover the 5-stage blueprint keeping LLMs from going rogue – from hallucination drift to real-time prompt correction. 🧠⚙️

🔗 Read now: medium.com/@rogt.x1997/3-83-bi


medium.com/@rogt.x1997/3-83-bi

Carol Romerocarol@floss.social
2024-12-03

The current #AIstack and how it could look like with democratic and ecological digital technologies, via Cecilia Rikap.

The image is divided into two sections, comparing two AI infrastructure models:

Left Panel: "Today's Privately Controlled AI Stack"

AI Chip Design: Illustrated with the NVIDIA logo.
Foundry: Represented by TSMC.
Processor: Follows chip production.
Data Sources: Includes "Data," "AI Talent," and "Micro Tasks."
Data Centres: Controlled by major private companies (Amazon, Google, Microsoft).
AI Foundation Models: Built on private infrastructure.
AI-Powered Apps: Applications powered by the AI stack.

Right Panel: "How a Public-Led International & Ecological AI Stack Could Look Like"

The stack components (AI Chip Design, Foundry, Processor, etc.) remain similar in structure.
Data Centres: Proposed as public-led, with three main layers:
Public material infrastructure.
Public universal platforms.
A public marketplace (procurement).
Competition Layer: Highlighted as a mechanism to ensure fairness and ecological balance.
The contrast shows a transition from a privately controlled model to a more open, publicly governed AI infrastructure.
☮ ♥ ♬ 🧑‍💻peterrenshaw@ioc.exchange
2024-03-29

“It is unlikely that you'll write it from scratch or write a whole bunch of #Python code or anything like that," he said on stage during his GTC keynote Monday. "It is very likely that you assemble a team of AI." …

A new AI development Stack?

How do you instruct, verify, test, update, ship those teams of AI together? With a new nVidiaLang 🤣 Containers of LLM models querying business services against a databases to generate reports… what could possibly go wrong with AI glue code on newly developed virtualised containers?

#nVidia / #LLM / #AIStack / #NIM <theregister.com/2024/03/19/nvi>

Client Info

Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst