Is there an AI coding bubble? Plus Meta’s new SlopTok product and more | E2184

Key Concepts

• TAM (Total Addressable Market): The total market demand for a product or service.
• Work Slop: Creating work product using AI that is perceived as low-quality or untrustworthy by coworkers.
• Think Slop: Sloppy thinking resulting from over-reliance on AI and neglecting fundamental reasoning processes.
• First Principled Thinking: A method of reasoning that breaks down complex problems into basic elements and reassembles them from the ground up.
• Jevon's Paradox: When technological progress increases the efficiency of resource use, but the rate of consumption of that resource rises due to increasing demand.
• Partisan Perceptual Bias (Partisan Motivated Reasoning): The tendency to interpret information in a way that aligns with one's political affiliation.
• AI Slop: Low-quality, AI-generated content.
• Interstitial Ad: An ad that appears between content, often during a loading period.
• GMV (Gross Merchandise Volume): The total sales dollar volume for merchandise sold through a marketplace over a certain period.

1. Market Dynamics and Founder Strategy

• Large TAM Attracts Competition: Founders should be drawn to large TAMs despite competition, as the market may be underestimated and innovation is always possible.
• Examples:
  • Uber entering the ride-sharing market despite Sidecar and Lyft.
  • XAI entering the LLM race after ChatGPT, Claude, and Gemini.
• Innovation Opportunities: Even in crowded markets, innovation can create room for new players. Apple investing heavily in LLMs and designing iPhones around AI is an example.
• "Bronze" Can Still Be Huge: Even if a company doesn't win a market, being a significant player can still result in a large outcome (e.g., Lyft).

2. Cybersecurity for Startups

• Neon App Case Study: An app that paid users to upload phone calls for AI training had a major security flaw, exposing user data.
• Common Sense Approach: Recording phone calls is inherently risky.
• Cybersecurity Investment: Startups should prioritize cybersecurity, given the increasing number of vulnerabilities.
• Think Slop: Over-reliance on AI can lead to sloppy thinking and neglecting basic reasoning.
• Strategies to Combat Think Slop:
  • Taking notes by hand.
  • Rewriting notes for better retention.
  • Reviewing bookmarks and notes regularly.
  • Taking breaks to digest information.

3. Tik Tok Valuation Controversy

• Valuation Discrepancy: The proposed $14 billion valuation for Tik Tok's US operations is questioned, given its $16 billion revenue in 2023 and projected growth.
• Possible Explanations:
  • The $14 billion may be the amount being invested, not the total valuation.
  • The number could represent the value of the stake being acquired by the new investing group.
• Revised Valuation: A 45% stake being acquired by a new investing group (including Oracle and Michael Dell) suggests a valuation of around $31 billion.
• Comparison to Comps: Even at $31 billion, the valuation seems low compared to price-to-sales multiples of other social media and e-commerce companies (Meta, Snap, Pinterest, Shopify, Amazon).
• Keith Raboy's Perspective: The company would have been worth zero under federal law, so this is a heroic outcome.
• Administration Communication: The administration's communication style can lead to confusion and the need to verify details.

4. AI in Radiology: Reality vs. Hype

• Radiologist Demand: Demand for radiologists is at an all-time high, contradicting predictions of AI replacing them.
• Performance in the Wild: Medical imaging AI models often don't perform as well in real-world settings as they do in benchmarks.
• Data Bias: Models trained on data from one hospital may not generalize well to other hospitals.
• Single-Issue Focus: AI models are often trained for single issues, while radiologists can identify multiple problems in a scan.
• Harvard Medical School Study: AI can interfere with a radiologist's performance and accuracy.
• Human-AI Collaboration: Combining human radiologists with AI can lead to worse outcomes due to reduced human effort.
• Iteration Problem: Data going into the AI could be variable whereas the data is cleaner going in during testing.
• Cautious Optimism: Despite current limitations, startups are working on improving medical imaging AI, and progress is expected.
• Examples of Startups:
  • RAD AI: Raised $151 million, 3,229% three-year growth.
  • AI doc: Raised over $400 million.
• Jevon's Paradox: Making healthcare cheaper could lead to increased consumption of services like X-rays.

5. Government AI Pricing and Motivations

• Pricing War: XAI priced its government AI product at 42 cents, undercutting Google's 47 cents.
• Other Offerings: OpenAI and Anthropic are offering their AI products to the government for $1.
• Potential Monetization: Companies may be willing to offer low-cost AI to the government to jumpstart interest in the technologies and eventually monetize through premium features, on-premise versions, and other services.
• Cynical Take: Private sector companies may be offering low-cost services to curry favor with the White House.
• Partisan Perceptual Bias: Interpretations of business relationships between private companies and the government are often influenced by political affiliation.

6. AI-Powered Coding Assistance

• Factory's Droids: Factory, backed by Sequoia, is developing AI agents called Droids to assist developers in coding.
• Investment in AI Software Development: Significant investment is flowing into AI software development startups.
• Market Size: The market for coding assistance startups is estimated to be $10 trillion.
• Developer Growth: The number of developers is expected to increase significantly in the coming decade, potentially reaching 250 million.
• Founder Strategy: Founders should be drawn to large TAMs despite competition, as the market may be underestimated and innovation is always possible.

7. Meta's "Vibes" and the Future of AI-Generated Content

• Meta's Vibes: Meta's AI team launched "Vibes," a feed of AI-generated videos.
• Negative Reception: "Vibes" received widespread criticism and was labeled "AI slop."
• Potential Audience: The product may be targeted towards older Facebook users who are more receptive to AI-generated content.
• Shrimp Jesus Phenomena: Older Facebook users have shown a fondness for AI-generated memes, such as images of Jesus that resemble shrimp.
• Meme Lord.com: A startup raised $3 million for memelord.com, a platform for creating memes.
• Memes as a Marketing Tool: Memes can be used to evoke emotions and serve as a form of branding.
• Slop to Art: Content that is initially considered "slop" can evolve into a recognized art form over time.

8. Open AI and Advertising

• Advertising in Chat GPT: Ads are likely coming to Chat GPT.
• Open AI Revenue Growth: Open AI expects significant revenue from new products, including free user monetization.
• Open AI Hiring: Open AI is looking for an ads chief.
• Interstitial Ads: Interstitial ads during the AI's processing time could be a viable monetization strategy.
• Monetization Variability: Monetization potential varies based on user location, with North America being more lucrative than other regions.
• Amazon's Advertising Business: Amazon's advertising services revenue was $15.7 billion in Q2 of this year, up 23%.

9. Where to Wheel Pitch Deck Analysis

• Where to Wheel: A marketplace connecting off-roaders with private landowners.
• Pitch Deck Strengths:
  • Efficient and well-structured.
  • Clear explanation of the problem and solution.
  • Strong go-to-market strategy (focus on the Northeast Corridor).
  • Credible team with relevant experience.
• Areas for Improvement:
  • Include more detailed personas of target customers (women, elite customers).
  • Explore premium service offerings (e.g., "Uber Black" for off-roading).
  • Incorporate sponsorship partnerships with manufacturers and gear companies.
• Key Metrics:
  • Demonstrate consistent month-over-month growth in GMV (10-20%).

10. Conclusion

The episode covers a wide range of topics, from cybersecurity vulnerabilities in startups to the potential and limitations of AI in various industries. It emphasizes the importance of critical thinking, data-driven decision-making, and understanding market dynamics. The discussion highlights the need for founders to be adaptable, innovative, and focused on building sustainable businesses, even in highly competitive environments. The analysis of the Where to Wheel pitch deck provides actionable insights for entrepreneurs on how to effectively communicate their value proposition and attract investors.