LLMs - The Stack (Page 3)

Synthetic focus groups and RAG in the contact centre: Bayer, Verizon, WPP on their AI deployments

"You can create a consumer, a brand strategist, a brand marketer, client, encoded with actual ground truth data, then critique the content that's been generated by the system with agents playing off against each other."

“Scant evidence” – Google’s AI chemistry claims were very misleading

AI discovered 2.2m new "materials" said DeepMind. Chemistry professors investigated, found hallucinations, repetition, known crystals.

DataStax

DataStax snaps up Langchain-democratiser Langflow

"You have to chain user input, system prompts, and DB data to feed the LLM and then lots of processing to deliver that magic AI agent experience to the user"

LLMs

Databricks releases most powerful open LLM yet

Inference “up to 2x faster than LLaMA2-70B” for new model, trained on 12 trillion tokens.

LLMs

Musk open-sources 314-billion parameter LLM "Grok" (including weights), but...

Apache 2.0 licence and weights, but no training code/reproducible datasets.

Europe’s AI Act demands extensive "logs" - targets biometrics, bias, black boxes

Emotion recognition banned in workplaces, classrooms.

Opinion

The key to making Large Language Models work is becoming clearer.

"GenAI with too low a temperature lacks creative spark... Too high a temperature and it will strongly hallucinate" -- Neo4j’s Jim Webber discusses new ways of delivering GenAI value.

Researchers go back to the 80s to jailbreak LLMs :-)

Generative AI models can be fooled using ASCII art, while "rainbow teaming" pushes LLM semantic defenses to limit...

LLMs

PDFs, RAG, and LlamaParse: Generative AI's "Swiss Army Knife" adds a welcome new toolkit

"Building production-grade RAG remains a complex and subtle problem... unlike traditional software, every decision in the data stack directly affects the accuracy of the full LLM-powered system."

OpenAI’s “Sora” lets you create video from text: Be awed, be worried.

Company says it is releasing examples "to give the public a sense of what AI capabilities are on the horizon" as one expert, NVIDIA's Dr Jim Fan emphasised that "if you think OpenAI Sora is a creative toy like DALLE... think again"

Google’s massively multimodal Gemini 1.5 Pro preview shows a giant warming to its task

"Coming soon, we plan to introduce pricing tiers that start at the standard 128,000 context window and scale up to 1 million tokens"

Red RAG to a Bull? Generative AI, data security, and toolchain maturity. An ecosystem evolves...

"Here's where people end up in RAG hell, with a bunch of unfamiliar tools and in many cases immature tools...”

See all