🎧 161 - Perplexity is so back, and AI doing something useful

This week, we're diving into the tangible, system-shaking ways AI is finally proving its worth—from crushing opaque billing systems to navigating complex legal waters. Plus, we're giving a shout-out to Perplexity as it continues to impress, and breaking down a massive legal threat to OpenAI that could cost billions. Get ready for a preview of the real AI revolution.

In This Episode:

• The $162,000 Win: How one person used Anthropic's Claude to reduce a hospital bill by 83% by exposing duplicate and illegal charges, demonstrating AI's power to flatten systemic opacity.

• OpenAI Policy Clarified: We cut through the confusion regarding ChatGPT's ability to help with medical and legal issues, confirming the model can still assist with personal matters.

• Perplexity's Power Play: Exploring why Perplexity is gaining traction, especially as a "best-in-class" finance tool with useful features like live stock tickers and providing accurate, up-to-date information.

• The Copyright Bomb: Details on the lawsuit against OpenAI for allegedly pirating books to train their models, which could lead to billions in damages if found liable.

• Google's AI Domination: Breaking down the massive $1B/year deal for Apple to integrate Google's Gemini model into Siri, a huge win for Google in the AI race.

🎙️ 160 - Grokipedia

• Sponsor of todays episode - The AI Anxiety Solution: Our sponsor, Airia, offers an enterprise AI orchestration and security platform to eliminate your AI anxiety, allowing for fast, compliant AI deployment. AIRIA.com

• • The End of Consulting? Ryan discusses a provocative article suggesting AI is killing consulting, citing how machine learning can now perform complex project work for a fraction of what firms like Accenture charge.

  • Music's New Reality: We look at how Suno v5's AI-generated music is becoming virtually indistinguishable from human-made songs, plus the breaking news that OpenAI is reportedly stepping in to develop its own music-generation AI.

  • The Wikipedia War: The launch of Grokipedia is "enraging" Wikipedia editors as the Grok-powered platform automates content creation, effectively replacing human authors and editors with a "just facts" approach.

  • Content Slop on Steroids: An Instagram Reels AI agent is analyzed, demonstrating how AI can scrape, analyze, and extract every creative insight from trending content, fueling the creation of optimized, high-volume "slop."

  • Nvidia's Dominance: Nvidia's stock price soars on a flood of massive new partnerships—including Eli Lilly, Palantir, Hyundai, Samsung, and Uber (for a 100,000 robotaxi fleet)—and a $1 billion investment in Nokia.

  
  

  Mentioned in this Episode:

  @Jukanlosreve@WesRothMoney@Grummz@lsanger@mikefutia

  Grokipedia vs Wikipedia:

  https://x.com/lsanger/status/1983337876042723740

159: ChatGPT Atlas - The new search engine or another forgotten piece of tech?,

• Begins with a "vibe gut check" of 6, which Ryan attributed to the launch of ChatGPT Atlas and the general, growing AI anxiety related to job security and rapid innovation.

• 
  

• • Ryan expresses strong skepticism about ChatGPT Atlas, calling it "clunky" and essentially a blend of Google Chrome and ChatGPT, and predicts that users will inevitably return to established search engines.

    
    

  • Ryan highlights that Atlas is already being "nerfed," noting that it blocked searches for certain historical topics (like videos of Hitler) due to propaganda concerns, and is also viewed by some as primarily a tool to collect large amounts of data for AI training.

    
    

  • The general online reception to Atlas is noted as overwhelmingly negative; Ryan observes a significant lack of positive promotion on X, suggesting a rare consensus that the product is underwhelming.

    
    

  • The discussion included skepticism of Sam Altman's comments on AI risk (translated as managing expectations for a disaster) and the idea that all the new image generators and data collection are leading to powerful, highly surveilled systems.

  Mentions:

  • @AnthonyHigman

  • @newstart_2024

  • @LuizaJarovsky