E13: The Evolution of Everything: Search, Chatbots, and AI's Growing Pains

In this episode, the hosts discuss the evolution of search, the evolution of chatbots, and the impact of AI on human connections. They explore OpenAI's new search product and its potential implications for the digital marketing industry. They also discuss the rise of AI companions and the role they play in simulating human connections. The hosts reflect on the decline of active listening in society and the potential role of AI in filling that gap. They also touch on the fragmentation of search behavior and the future of internet usage. The conversation covers various topics related to AI, including the impact of AI on human connection, the concept of active listening, the stock market's reaction to AI investments, and the limitations of data and model collapse. The hosts discuss the challenges of information overload and the demand for active listening in a hyper-personalized world. They also explore the potential benefits of AI in providing emotional relief and support. The conversation touches on the stock market's reaction to AI investments and the debate between deflation of expectations and a temporary correction. Lastly, they discuss the limitations of data and the phenomenon of model collapse, where AI-generated content becomes less accurate and coherent over time.

Takeaways

• OpenAI's new search product could disrupt the digital marketing industry and change the way people search for information.
• AI companions are becoming popular as a way to simulate human connections, particularly among women.
• Active listening is declining in society, and AI could potentially fill the gap by providing a listening ear.
• The future of search behavior may be more fragmented, with users relying on different platforms and interfaces for information.
• The internet may evolve into a more curated and summarized experience, similar to Flipboard, where users get information without needing to click on links. In a hyper-personalized world, there is a demand for active listening and emotional support, which AI can potentially provide.
• The stock market's reaction to AI investments is mixed, with some companies experiencing a deflation of expectations while others see it as a temporary correction.
• The limitations of data and model collapse pose challenges for AI, as synthesized data can lead to a decrease in quality and accuracy over time.
• The future of AI may involve the employment of content writers and data providers to ensure originality and quality in AI-generated content.
• The impact of AI on human connection and the role of active listening are important considerations in the development and implementation of AI technologies.

Episode Transcript

Introduction

Chris: Welcome back to another episode of the Chris Rod Max show, where we discuss the latest trends in AI, interview founders building AI products and services, and interpret what it all means for humanity. I'm excited to be back with Max and Rod. Today, our theme is the evolution of everything AI-related. We'll cover three main topics:

1. The evolution of search
2. The evolution of chatbots and their impact on dating and human connections
3. The evolution of the AI market, including stock developments and data limitations

Let's dive in!

The Evolution of Search

Chris: OpenAI recently released a press release about building a better version of their product to enhance the search experience. Rod, can you explain what they're actually building?

Rod: When ChatGPT first came out almost two years ago, it was an assistant disconnected from the internet. You could ask it to summarize or help craft text, but it couldn't provide up-to-date knowledge. Over the last month, they've changed the product. Now, when you ask for information, it first crawls the internet, collects multiple responses, and aggregates them. For example, if you ask how to improve your sleep, it will gather information from multiple sources and summarize them.

Chris: Max, how would you interpret this move by OpenAI?

Max: They're coming for Google. That's what it is. Search traditionally involves indexing, searching, and ranking information to push back to the end user. The interface used to be just a search bar, but now it's becoming more of an assistant. You're talking to it, which allows for different indexing, searching, and ranking based on your conversation. Google does this to some extent, but not with a human-like interface. It'll be interesting to see how Google responds.

Chris: Rod, has anyone changed their search habits? Are you using any new tools instead of Google?

Rod: I use Perplexity from time to time. I've also noticed that Google itself now provides an AI answer for some search terms, just below the search bar, followed by classic search results. When I use ChatGPT, I assume I'm using it as a search engine of sorts, as it now primarily pulls information from the internet.

Max: I still use Google for about 80-90% of my searches. I use ChatGPT for more esoteric queries, but I often double-check the information on Google to ensure accuracy.

Chris: Max, what you're describing is interesting because it shows the fragmentation of search behavior. You're exposed to another viable option with ChatGPT, but you're not fully trusting it yet. This new beta product from OpenAI might change that, potentially shifting users completely.

The Evolution of Chatbots and Human Connections

Chris: Let's move on to our next topic: the evolution of chatbots. We've discussed chatbots and their effectiveness in customer support before. This time, there's an interesting article about AI companions. Two things caught my eye:

1. A woman who had an absent partner downloaded one of these AI companion apps and created an avatar. Through chatting with this avatar, she realized she deserved a better partner.
2. A quote stating that "active listening is becoming a dying art for humans."

What are your reactions to these points?

Rod: This trend of digital companions isn't new. We've seen it with Tamagotchis in the mid-90s and robot pets like Sony's Aibo in the late 90s. What's striking about the article is that it defies the common wisdom that AI replacing human connections would primarily appeal to men seeking digital girlfriends. Apparently, there's a strong trend of women finding more value in interacting with digital characters rather than engaging with real men.

Max: Regarding the statement that active listening is dying, I think we can see evidence of this in social media interactions, where people often attack each other's political ideologies instead of truly listening. This lack of active listening might be driving people to seek connections through AI.

Active listening is a skill that therapists and counselors spend years developing. AI could potentially fill this gap for those who can't access or afford professional therapy. However, the underlying problem of people not paying enough attention to each other remains a concern.

Chris: To close this discussion, what does active listening mean to you?

Rod: Active listening involves paying attention and accepting what the person is saying without immediately trying to provide an answer or response. It's about absorbing what the other person said and acknowledging their perspective rather than trying to insert your own version of things.

Max: Active listening is about picking up both verbal and non-verbal cues to understand the undercurrent of what someone is saying. It's not just hearing the words, but trying to understand the feelings behind them and the non-verbal signals. It's about capturing all these informational cues to get to the ground truth from the other person's perspective.

The Evolution of the AI Market

Chris: Let's zoom out and look at the bigger picture of the AI future. There are two topics I want to touch upon:

1. The stock market's reaction to some earning reports, with Alphabet losing 5% of their stock price and Nvidia's market cap decreasing by more than 15% over the last two weeks.
2. An article published in Nature about "model collapse" and the limitations of data for training AI models.

What are your thoughts on these developments?

Rod: If we look at Google Trends for ChatGPT, although it's down about 30% compared to June, interest worldwide is up by almost 60% compared to last summer. The market might be saying that AI adoption is happening, but not as fast as expected. There's a big gap between AI-focused companies and the rest of the market, based on expectations of rapid AI integration. In reality, this is happening more organically, and the market is adjusting accordingly.

Max: As Warren Buffett said, "In the short term, the stock market is a beauty contest. In the long term, it's a weighing machine." We have to see which AI companies will deliver value in the long term. The immediate value of AI isn't there yet, but the long-term trend is likely to continue.

Regarding the "model collapse" issue, it's concerning that we're already using AI to generate content that will become training data for future AI models. This could potentially lead to a situation where AI-generated content dominates, potentially causing issues with model quality and diversity of outputs.

Chris: Thank you both for your insights. It's clear that the AI landscape is rapidly evolving, with implications for search, human connections, and the broader tech market. As we continue to navigate these changes, it's crucial to consider both the potential benefits and challenges of AI integration in various aspects of our lives.