Welcome humans, here’s your 3-minute recap (without the fluff):

• NVIDIA’s Offline "Chat with RTX" chatbot

• Tell ChatGPT what to remember or forget from past conversations

• Otter's Meeting GenAI: Analyze meeting transcripts with a chatbot to answer questions

• New tools, funding updates & job postings

How to slice LLM cost in half with Mixtral in OctoAI

OctoAI recently demonstrated that Mixtral in OctoAI delivers over 50% cost reduction versus GPT-3.5-Turbo’s now even lower-priced option.

They created a text summarization demo app for this evaluation, and the table below captures the results across a range of document sizes from around ten thousand to over half a million words.

How did they achieve this?

We built OctoAI with deep ML systems expertise and experience in model compilation. That means you can pick up and run with the latest OSS models while we manage the complexities of hardware and performance.

How can you use OctoAI to improve your bottom line as your LLM app scales?

→ Read the blog and reach out to their Team to discuss how we can put this efficiency to work for you to achieve your goals in speed and cost. Or, sign up for free and try it yourself.

Want to advertise with us? Get started here.

🛠️ Tools

Shortform - If your goal is to learn more from Nonfiction books, you NEED Shortform. And at no extra cost, you get access to the Shortform AI extension that allows you to consume knowledge from the internet at hyperspeed. Get 20% off the Annual Subscription only at our link. (link)

Video to Blog - As the name shows it creates blogs from videos in seconds (link)

Leap - Helps to design, test, monitor, and deploy AI workflows to production in minutes (link)

PixVerse - Turns your ideas into visuals with video creation software (link)

📚 Resources

Nvidia is worth more than Alphabet, one day after surpassing Amazon (link)

How DBS is industrializing AI across its business (link)

Intel’s AI efforts could get a boost from Kimrey’s new sales group (link)

Tsmc has soared most since 2020 after broker upgrades on AI prospects (link)

Qantm AI CEO on AI strategy, open source models, and what’s to come (link)

US-India Business Council launches AI task force, boosting leadership in artificial intelligence (link)

P.S. You can still get The Ultimate ChatGPT Toolkit (no email opt-in).

🗞️ News

Nvidia’s offline AI

Nvidia has released a free demo of a new AI chatbot called Chat with RTX that can run locally on PCs with GeForce RTX GPUs. The chatbot can be trained on files the user selects, like documents, videos, and web pages.

This allows it to answer personalized questions about the content. It provides fast responses while keeping user data private since it doesn't rely on cloud services or internet access.

→ More details about Chat with RTX

Erase ChatGPT’s memory

OpenAI has introduced new memory controls for ChatGPT that allow users to tell it to remember or forget specific information from conversations to improve future responses.

The memory can be cleared or individual memories deleted. The memory could make ChatGPT more helpful.

→ More details about ChatGPT’s memory controls

AI meeting assistant

Otter’s new product, Meeting GenAI, includes a chatbot that can analyze transcripts from previous meetings to answer questions about decisions made, summaries, or other details.

It aims to save time by allowing users to get information from past meetings without rewatching recordings or reading notes.

→ More details about Otter’s Meeting GenAI

✅ Daily AI Challenge

🐦 Tweet Of The Day

🤖 Prompt Tutorial

Turn Likes into Leads

Give me 10 ways to use social media to generate leads and build relationships with [target customers].

Target customers = [Insert here]

That’s all.

Stay curious, leaders!

- Mr. Prompts

PS. If you missed yesterday’s issue, you can find it here.

*Our emails may contain sponsored content that helps support the team