The Quantum Drift

In this episode, Robert Loft and Haley Hanson dive into AI2’s Tülu 3, the newest open-source tool shaking up AI post-training. Forget needing a tech giant to fine-tune your AI—this democratizes the process entirely.

We discuss:

• Post-training’s importance: Why raw models like Llama aren’t ready for prime time out of the box.
• How Tülu 3 gives developers control to customize models for tasks like coding, math, or multilingual abilities.
• The end of reliance on big tech for custom AI: Why Tülu 3’s open-source approach is a game changer.

Can open tools like Tülu 3 level the playing field between scrappy innovators and AI giants? Join us as we explore the future of accessible AI innovation!

AI is the new battleground, and Amazon just doubled down with an additional $4 billion investment in Anthropic, the AI powerhouse behind Claude. In this episode, Robert Loft and Haley Hanson unpack how this move positions Amazon as a pivotal player against OpenAI and Google in the race for AI dominance.

We explore:

• Anthropic’s revolutionary Claude AI capabilities, from navigating websites to executing multi-step tasks autonomously.
• How Amazon’s AWS partnership gives its customers a head start with fine-tuned AI tools.
• The $1 trillion generative AI market forecast and what’s at stake.

Is Amazon rewriting the rules of AI with Claude, or is it simply keeping pace with the competition? Tune in to see how this strategic partnership could reshape tech's future.

Robert Loft and Haley Hanson dive into Microsoft's latest AI advancements unveiled at Ignite 2024. With Copilot Actions, the tech giant is redefining automation in Microsoft 365. Imagine automating your weekly reports, meeting agendas, and even multilingual PowerPoint translations—all with minimal input.

We explore:

• The rise of Copilot AI agents in SharePoint, Excel, and Outlook.
• How these tools aim to eliminate repetitive tasks and boost productivity.
• The broader implications of AI-powered macros for businesses and individuals.

Can Microsoft lead the way in turning AI into your daily digital assistant? Join us as we unpack their vision for the workplace of tomorrow.

Could AI workspaces redefine how we collaborate with technology? In this episode, Robert Loft and Haley Hanson dive deep into Claude Artifacts, Anthropic’s groundbreaking feature for seamless human-AI teamwork.

We explore:

• What are Artifacts? A powerful tool to separate complex projects from casual AI chats.
• Interactive features: Version control, targeted editing, real-time previews for code, diagrams, and more.
• The Claude advantage: How Artifacts outperform ChatGPT Canvas with specialized previews and built-in collaboration tools.

Is this the next leap in AI productivity, or just hype? We’ll show you how Claude Artifacts could transform workflows for coders, creators, and teams alike.

This week, Robert Loft and Haley Hanson dig into a provocative idea from tech veteran Marissa Mayer: ad-supported AI chatbots. At the Cerebral Valley AI Summit, Mayer painted a picture of a future where AI tools provide highly detailed, ad-driven answers by integrating directly with advertisers' inventory.

Join us as we debate:

• Will advertisers embrace radical data sharing to fuel AI experiences?
• The potential of AI partnerships with StubHub, Ticketmaster, and other platforms to revolutionize search and chat.
• How this vision could reshape the business model of AI providers like OpenAI and Perplexity as compute costs rise.

Is ad-supported AI the future or a privacy nightmare waiting to happen? Tune in for this thought-provoking discussion.

The rise of AI is shaking up the workforce, and Robert Loft and Haley Hanson are here to unpack the latest. In this episode, we dive into a groundbreaking study revealing a 21% drop in demand for freelance jobs vulnerable to automation, like writing and graphic design, just months after ChatGPT’s debut.

Join us as we discuss:

• Why routine tasks are vanishing and what this means for creative industries.
• The growing divide between AI-complementary skills and jobs that are fading away.
• How tools like ChatGPT and Midjourney are impacting compensation and competition in the gig economy.

Are critical thinking and creativity enough to future-proof your career? Find out as we explore the high-stakes evolution of work in the AI era.

AI agents are everywhere, but can they deliver consistently? Robert Loft and Haley Hanson unpack H2O.ai’s latest innovation: h2oGPTe, a platform combining generative and predictive AI for enterprise workflows.

We discuss:

• The power of predictive AI for consistent, reliable agent responses.
• How multimodal agents analyze charts, historical data, and trends to inform decision-making.
• Why H2O.ai’s automated testing tools may set a new standard for agent accuracy.

Are generative AIs enough, or do enterprises need the consistency predictive models offer? We’re exploring the future of AI agents—and how they’ll transform industries like finance, healthcare, and beyond.

This week, Robert Loft and Haley Hanson explore DeepSeek's newly unveiled R1-Lite-Preview, the AI model causing ripples in the reasoning world. Dubbed as a competitor to OpenAI's o1, this reasoning-first LLM offers transparency with its "chain-of-thought" approach, allowing users to follow its step-by-step logic in real-time.

We break down:

• What makes R1-Lite-Preview unique and how it stacks up against industry heavyweights.
• The potential of transparent reasoning in building trust with AI.
• DeepSeek’s vision to open-source its cutting-edge models.

Is this the model to watch as the AI reasoning race heats up? Or just another splash in the competitive LLM ocean? Tune in to find out!

In this episode, Robert Loft and Haley Hanson dive into Rabbit’s groundbreaking "Teach Mode" for the R1, a feature that lets users train the AI to automate tasks across any website. From tweeting to grocery shopping, the R1’s teachable AI promises convenience, but does it deliver?

We discuss:

• What is Teach Mode, and how it works to mimic your actions.
• The vision for a future "agent store" where users sell custom automations.
• Rabbit’s lofty claim of creating the ultimate multi-platform navigator.

Can Rabbit's R1 redefine automation, or is this just another step in a long journey toward true AGI? Let’s break it down.

Ever wondered what’s happening behind the scenes of AI like ChatGPT? Join Robert Loft and Haley Hanson as we explore the magic and mechanics of Large Language Models (LLMs). Inspired by a mind-blowing animated walkthrough, we’ll break down how these models:

• Tokenize words and numbers: The first step in understanding human language.
• Predict outputs: How AI decides what comes next in a sentence.
• Learn patterns: Training LLMs to handle everything from coding to poetry.

We’ll also talk about why some see these models as “digital sorcery” and debate if AI’s inner workings are truly comprehensible. If you’re curious about the tech but don’t want to drown in jargon, this episode is your cheat sheet.