The AI Slop Echo Chamber
The AI Slop Echo Chamber
YouTube16 min 22 sec
Watch on YouTube
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights

Consider a moderate, long-term position in Reddit (RDDT): its human-generated discussions may become more valuable for AI training, and one speaker added shares, but no price target or confirmed licensing catalyst was provided. Treat slower user growth, AI-generated content, and competition as key risks; licensing renewals and a potential acquisition remain speculative scenarios. Avoid Yelp (YELP) and Shutterstock (SSTK) as investments based on this discussion—the suggested sympathy trades were highly uncertain and potentially lasted only minutes.

Detailed Analysis

Reddit (RDDT)

  • The speakers viewed Reddit as a valuable source of fresh, human-generated data for AI training, especially because its discussions capture arguments, opinions, and how people work through decisions—not just final answers.
  • They argued Reddit could benefit if AI companies need to license reliable human data. One speaker expects potential licensing renewals to be priced meaningfully higher, though no specific deal, price, or timeline was identified.
  • Reddit’s user growth had slowed to 7% in the latest reporting cycle discussed, while 5% of all Reddit content ever written was reportedly created in the prior quarter. The speakers could not determine how much of that increase came from AI.
  • AI-generated content infiltrating Reddit was identified as a concern because the platform’s value depends in part on human discussion. The speakers also noted uncertainty about whether all companies using Reddit data—particularly those outside the United States—pay for it.
  • Reddit shares reportedly dipped after news of a Meta competitor, but one speaker thought the reaction was excessive and expected Reddit to remain differentiated by its interest-based communities.
  • One speaker said they had bought more Reddit shares, describing the position as moderate but outside their top four or five holdings. They characterized the risk/reward as attractive, while acknowledging that the stock could be difficult to trade in the short run.
  • One speaker estimated a 20% chance that Reddit could eventually be acquired. This was a personal scenario estimate, not a confirmed expectation or timeline.

Takeaways

  • The discussion presents a bullish long-term thesis for Reddit: its human discussions could become more valuable to AI developers, while its communities may appeal to users seeking human interaction.
  • Key uncertainties raised include slower user growth, the amount of AI-generated content on the platform, whether data use is consistently paid for, and competition.
  • The speakers did not provide a price target. Treat the licensing and acquisition possibilities as scenarios, not established catalysts.

Meta Platforms (META)

  • The speakers said Meta had announced a competitor to Reddit, which they believed contributed to Reddit’s brief share-price decline.
  • They were skeptical that Meta’s offering would succeed. Their reasoning was that Meta’s platforms are organized around personal relationships, while Reddit is organized around shared interests and topic-specific communities.
  • They also cited Meta’s previous attempt with Threads as a reason for skepticism, though they disagreed on whether the issue was user trust or the fit between Meta’s existing platforms and topical discussion.

Takeaways

  • The discussion was skeptical of Meta’s ability to replicate Reddit’s community model, but did not provide details about the new product or a broader view on Meta as an investment.
  • The competitive announcement is relevant to Reddit’s outlook, but the speakers’ comments reflect opinion rather than evidence of the product’s eventual performance.

Alphabet (GOOGL)

  • The speakers described Google as having a potential data advantage, including access to video data through YouTube and possibly email data. They questioned whether Google has the rights to use email data for model training.
  • They noted that YouTube has historically hosted original content, but is increasingly filling with AI-created videos.
  • Rumors that an upcoming Gemini release could be “earth shattering” were mentioned, but were not verified.

Takeaways

  • The discussion suggests Alphabet’s data access may be an advantage in AI, while the growth of AI-generated content could make the quality and originality of some data less certain.
  • The Gemini comments were explicitly presented as rumors; no investment recommendation, price target, or release timeline was given.

xAI and X

  • The speakers said xAI may have an advantage through access to data from X. They considered X useful for timely, topical discussion, but debated how broad and valuable its discussions are compared with Reddit’s.
  • One speaker described Reddit as broader in its range of communities, while X was characterized as particularly useful for current topics such as technology, politics, and finance.
  • A speaker said AI agents had been posting on their X profile, illustrating the broader concern that online content may not always be written by the account holder.

Takeaways

  • The discussion identifies timely human conversation as a potentially useful AI training resource, but does not establish the value of X’s data or the terms under which xAI can use it.
  • xAI is not separately publicly traded, and no specific investment recommendation was made.

Yelp (YELP) and Shutterstock (SSTK)

  • The speakers mentioned Yelp and Shutterstock as possible short-lived sympathy trades if Reddit were acquired or announced a major data-licensing deal.
  • The proposed logic was that investors might briefly reassess other companies with data that could be valued in a similar way. One speaker described the potential trading window as only several minutes.
  • A speaker said they disliked Yelp’s business and did not want to own it, and questioned the reliability of Yelp reviews compared with Google reviews.
  • This was presented as a speculative scenario, not as a current trade or a fundamental endorsement of either company.

Takeaways

  • The idea is an event-driven, high-uncertainty trading hypothesis, not a long-term investment thesis.
  • The speakers gave no target prices or defined execution plan. The potential price reaction and timing are uncertain.

AI Training Data and Licensing

  • The broader investment theme was that AI systems may generate more information than humans produce, potentially creating a feedback loop in which models train on increasingly AI-generated or replicated content.
  • The speakers suggested that fresh, high-signal, permissioned human data could become more valuable as AI companies seek reliable sources for training.
  • Reddit was the main example. YouTube, X, Yelp, and Shutterstock were also discussed in relation to data access or possible market reactions.
  • The speakers emphasized that the topic is uncertain and that they could not say with conviction how the issue will develop.

Takeaways

  • The discussion points to a possible investment theme around companies that own or license distinctive human-generated data, but the value depends on data quality, usage rights, and whether AI-generated content dilutes the source material.
  • No specific timeline or confirmed industry-wide pricing trend was provided.
Ask about this postAnswers are grounded in this post's content.
Video Description
AI is starting to eat its own slop. AI-generated images, videos, text, and code are flooding the internet, and the next generation of AI models may increasingly have to learn from that content. At the same time, AI agents are becoming capable of doing research, running experiments, writing code, and even helping build better AI systems. So what happens when the internet becomes increasingly filled with AI-generated content? In this video, we break down the rise of AI slop, why AI companies are facing a growing data problem, and how a feedback loop of AI-generated content could change the way future models are trained. If AI is learning from the internet, what happens when the internet starts learning from AI? — 👍 LIKE what we're doing? Smash the thumbs up! 🔔 SUBSCRIBE with "all" notifications to know when we're on ✅ CONNECT on IG, FB & Twitter @DumbMoneyTV 💬 JOIN our Discord https://DumbMoney.tv/discord 🐦 TWEET @ChrisCamillo @DaveHanson and @Jordan_Mclain 🎧 LISTEN to our podcast https://DumbMoney.tv/podcast 👕 BUY stuff with our logo https://DumbMoney.tv/merch — Our videos contain personal views and opinions and are intended strictly for information, education & entertainment purposes. We do not provide investment advice or investment strategy. #investing #trading #chriscamillo #ai #artificialintelligence #meta #metamuse #muse #reddit #twitter #x #chatgpt #google
About Dumb Money Live
Dumb Money Live

Dumb Money Live

By @dumbmoneylive

We are Dave Hanson, Chris Camillo & Jordan Mclain. On this channel, we reveal our actual investments and thoughts on the ...