New AI Hype or Headache?

Welcome back apprentices! 👋

Hope you enjoyed your turkey – and hopefully, it didn’t look like this 🦃.

It’s Friday, November 29th, a.k.a. Black Friday – the perfect day to grab deals and upgrade your game. And while you’re snagging discounts, we’ve got something special to help you level up even further:

Our Online Shopping Superhero

Remember the not-so-glorious days of jotting down your shopping list, dropping it somewhere between your kitchen table and the mall, then braving the crowds with no idea what you came for?

Thank goodness for Amazon… and thank goodness for Brad’s Deals, which helps you make the most of your Prime membership.

Brad leads shoppers by the hand, showing you ten lesser-known membership wonders you never knew you had—like unlimited photo storage, free release-date delivery, and early access to Lightning Deals.

Get all the deals from Brad himself and transform your shopping, streaming, and saving experience. Sign up for free with your email address!

Learn more

In today's email

• Amazon Unveils “Olympus”

• AI Grandma Trolls Scammers

• Pony.ai Hits $5.4B Valuation

• School's AI Cheating Penalty

• OpenAI’s Sora Leak

• Alibaba rewrites the AI playbook

• More AI Tools & Updates

_{Read Time: 3.5 minutes}

Quick News 

Amazon Unveils 'Olympus' AI Model: Amazon has developed 'Olympus,' a generative AI model capable of processing images, videos, and text. This advancement aims to reduce Amazon's reliance on external AI solutions and enhance its competitiveness in the generative AI space. The official announcement is anticipated at the AWS re:Invent conference.

AI Grandma Trolls Scammers, Saves the Day: Virgin Media O2 has unleashed "Daisy," an AI-powered grandma who’s here to outwit phone scammers by chatting their ears off. Think of her as your sweet old nan – if your nan could trap fraudsters in 40-minute conversations about knitting and “the good old days.” Daisy’s already handled over 1,000 scam calls, proving that sometimes the best way to fight crime is with tea and biscuits... and a lot of patience.

Pony.ai Hits $5.4B Valuation with $413M Raise: Chinese autonomous driving startup Pony.ai has secured $413 million in fresh funding, achieving a valuation of $5.4 billion amid its revived plans for a U.S. Nasdaq IPO. The company aims to challenge Tesla in the autonomous vehicle space, focusing on its Robotaxi and self-driving truck platforms. This milestone also signals renewed interest in Chinese IPOs in the U.S. market, following a lull due to geopolitical and regulatory tensions.

Court Upholds School's AI Cheating Penalty: A federal judge has ruled in favor of Hingham High School in Massachusetts, affirming the school's decision to penalize a student accused of using AI to cheat on an assignment. The student's parents had sued, arguing that the school's policies did not clearly prohibit AI use. However, the court found that the school's academic integrity guidelines were sufficient to inform students about the misuse of AI tools. 

Together with Artisan

10x Your Outbound With Our AI BDR

Imagine your calendar filling with qualified sales meetings, on autopilot. That's Ava's job. She's an AI BDR who automates your entire outbound demand generation.

Ava operates within the Artisan platform, which consolidates every tool you need for outbound:

• 300M+ High-Quality B2B Prospects

• Automated Lead Enrichment With 10+ Data Sources Included

• Full Email Deliverability Management

• Personalization Waterfall using LinkedIn, Twitter, Web Scraping & More

Book a demo and supercharge your sales team.

OpenAI Leak 
OpenAI’s Sora Leak: A Game-Changer or Just Hype?

A leaked version of OpenAI’s unreleased video generation model, Sora, has sparked controversy and mixed reactions. While showcasing promising capabilities, the leak has also highlighted tensions with creators and skepticism about its competitive edge. Here’s what we know so far:

The Leak: A protest group called "Sora PR Puppets" uploaded the model on Hugging Face, giving the public a glimpse before OpenAI swiftly shut it down.

Performance Highlights:

• Generated 1080p 10-second clips in under 10 minutes.

• Features rumored to include in-painting and image generation in future versions.

Controversy:

• Claims of unpaid testing by hundreds of artists who helped train the model.

• Strict control over generated content, limiting creative freedom and transparency.

Public Reaction: Social media users were underwhelmed, describing the output as "not better than competitors" and "dated," raising questions about Sora’s position in the rapidly evolving AI video space.

So what’s the big deal? Sora’s potential to generate high-quality video faster than before could offer efficiencies in marketing, media production, and creative industries. However, the mixed reception shows the importance of continuous improvement and addressing user expectations. Sora represents a step forward but also a reminder that staying ahead in AI requires more than just incremental upgrades – it demands user trust and collaboration.

Sora’s progress highlights both opportunities and pitfalls. And OpenAI’s handling of this leak and future updates to Sora could shape its reputation and adoption across industries. Will Sora rise to the occasion or remain a “what could have been”? Time – and feedback – will tell.

AI Globalization
Alibaba just rewrote the AI playbook – and OpenAI should probably take notes.

Just as OpenAI scrambles to mop up the mess from Sora’s leak, Alibaba confidently sails into the spotlight with QwQ-32B, its latest reasoning model. While Sora struggles with public skepticism and creative drama, QwQ flexes its 32K token muscles and challenges OpenAI head-on. It’s like watching one team plug holes in a sinking ship while the other builds a battleship.

QwQ-32B is an open-source large language model (LLM) designed to compete directly with OpenAI’s proprietary models like GPT-4 –  and here’s what you need to know:

• Benchmark Leadership: Competes closely with OpenAI's o1-preview and outperforms smaller versions like o1-mini on math and programming reasoning tasks.

• Domain-Specific Mastery: Unlike generalist models, QwQ-32B allows businesses to fine-tune it for specific industries or use cases, making it a flexible alternative to closed models.

• Open-Source Game Changer: Available on an open license, this model invites global developers to collaborate, fine-tune, and innovate.

• Built for Scale: Supports up to 32 billion parameters, enabling sophisticated problem-solving, reasoning, and natural language understanding.

This is not just another LLM – it’s a disruption in the status quo. QwQ-32B could mark the beginning of an open-source revolution in AI, challenging the dominance of proprietary systems like OpenAI and Google DeepMind. 

What could open-source models mean for your AI roadmap? Assess how QwQ-32B might reduce costs while giving you greater control over custom applications. Keep an eye on Alibaba’s advancements – they could shape a more competitive and collaborative future for AI-driven business solutions.

With Chinese firms like Alibaba leading the charge, this evolution highlights a potential shift toward global AI democratization.

AI Tools & Updates

🎬 Runway’s "Frames" Puts AI in the Director’s Chair: Runway has unveiled Frames, a groundbreaking AI-powered video-editing tool that offers professional creators the ability to edit videos frame by frame with unprecedented precision. Leveraging advanced machine learning, Frames enables automatic object tracking, scene modifications, and seamless integration with existing workflows. 

The tool has the potential to transform industries like film, advertising, and content creation by slashing manual editing time. For professionals aiming to streamline post-production, this innovation could be a game-changer, setting a new standard for efficiency and creativity.

🎵 The Swiss Army Knife of Sound: NVIDIA has introduced Fugatto, a generative AI model that can transform and create audio from text and sound prompts — think turning a trumpet into a barking dog or tweaking a voice's emotion and accent. While this groundbreaking tool has massive potential for industries like music, gaming, and advertising, there's a catch: it’s not publicly available yet. 

Fugatto remains behind NVIDIA's lab doors, leaving us all curious about when it might make its grand debut.

📚 OLMo-2: AI2's Open Multimodal Model Gets a Major Upgrade: Research institute AI2 has unveiled OLMo-2, a new family of open-source language models rivaling Meta's Llama in performance. Trained on a 5T token dataset of high-quality academic and web data, the 7B and 13B models achieve competitive or superior results while using fewer resources. 

Fully open, OLMo-2 includes source code, training data, and a dev package with recipes and evaluation tools, making it a standout in transparency and accessibility.

🔗 Anthropic’s MCP: AI’s New Passport to All Your Tools? Anthropic has introduced the Model Context Protocol (MCP), a game-changing open-source standard that allows AI systems to integrate seamlessly with external tools and data sources. With pre-built MCP servers for platforms like Google Drive, Slack, and GitHub, developers can also customize their own connectors to enable AI assistants to access data across repositories and environments effortlessly. This development is already live for Claude Enterprise users, who can test MCP servers to link AI with internal datasets and tools.

The MCP eliminates the need for creating individual connectors for every platform, addressing a key hurdle in AI systems evolving into robust, agentic tools. By streamlining integration, it lays the groundwork for AI assistants to operate as genuinely versatile systems capable of managing diverse tasks and datasets with ease.

🤖 Meet Robo-Dancer!

AI isn't just about calculations - it knows how to groove! What dance move do you think it's doing? 💃

Let's Chat About AI

This is all for today!

Replay to this email with your thoughts.

AI is more than just a buzzword. It’s a shift in how we live and work. And understanding it a bit better means you can make smarter choices about the tech you use every day.

_FEEDBACK