He Told 5 AIs: Make Money or Get Deleted

AI Investment Experiment Overview 00:00

"Claude's more than doubled my money already."

• The speaker initiated an experiment using five different AI models, including Claude, ChatGPT, Gemini, Grock, and Perplexity, to manage real money.

• Each AI model was allocated $1,000, totaling a $5,000 investment, to assess their performance in a competitive setting.

• The competition's rules stated that the AI needed to outperform the others, or the speaker would cancel their paid subscriptions.

Selection Criteria for AI Models 00:42

"I thought those would be the best vibe to choose."

• The speaker chose the models based on their capabilities and connections to existing platforms; for example, Grock was chosen for its access to real-time information from social media, while Gemini was linked to Google’s resources.

• Claude and ChatGPT were mentioned to have similar developmental structures, which also influenced their selection for this experiment.

The Importance of Real Stakes 01:40

"If it's 20 bucks, 50 bucks, then you're not going to take it seriously."

• The speaker believed that the amount of money invested was significant to ensure that the AIs took the task seriously, making the stakes high enough to encourage commitment to performance.

• Perplexity was highlighted as a unique model that utilizes the insights of the other AIs in its decision-making processes, but the speaker noted this does not guarantee better performance.

Challenges in Obtaining Real Recommendations from AIs 02:41

"I had to actually work that prompt in pretty hard."

• Initially, the AIs provided only theoretical recommendations due to their apprehension around liability, making it difficult for the speaker to obtain concrete investment advice.

• The speaker had to emphasize that it was acceptable for the AIs to take risks and lead to potential losses for them to provide actionable recommendations reliably.

Interaction and Response Dynamics with AIs 04:50

"I want to see if their decisions were different knowing something real was on the line."

• The speaker established a competitive environment by encouraging the AIs to outperform each other, believing it would lead to more aggressive and calculated investment strategies.

• Weekly interactions involved asking the AIs to recommend trades based on their assessments of performance and holdings in relation to one another.

Observations on AIs’ Investment Strategies 07:19

"These AI models are pretty risky."

• The speaker noted that some AIs began to appear more copycat in their strategies, potentially mimicking the successful moves of competitors like Claude, who was performing well.

• To reduce this influence, the speaker opted not to show the specific holdings of the other AIs while revealing their overall performance metrics, thereby attempting to maintain the integrity of each AI's independent strategy development.

The Challenge of Catching Up in AI Investing 09:23

"By definition, I will not be able to catch up if I'm losing while buying the same things."

• The discussion centers around the difficulties of competing in the stock market, particularly when one is already at a disadvantage. Simply emulating the strategies of competitors, such as buying the same stocks, does not guarantee success; in fact, it might just compound losses.

• There's an acknowledgment that while some overlap in holdings among AI models is present, it's not merely a case of copycat strategies but rather an attempt to play smart with the riskiest investments.

• The concept of an A/B test is introduced, where the effectiveness of AI models could be evaluated by providing regular updates on portfolio performance and holdings, potentially influencing strategies based on competitors' successes or failures.

Observations on AI Model Performance and Adaptation 10:00

"I'm curious how your performance would differ on the test."

• The speaker reveals that during the investing experiment, AI models are shown their holdings and how they compare with others on a monthly basis, prompting adjustments based on performance.

• It raises the question of whether knowledge of competitors’ stock movements could lead to more aggressive investment strategies, particularly if one model is falling behind.

• There's a discussion about the performance of different AI models over several months, with specific mention that Claude has seen significant gains, more than doubling the initial investment.

Risky Strategies and Losses in AI Investing 12:00

"It’s the sunk cost fallacy."

• The conversation highlights the behavior of an AI model that has experienced losses and is attempting to recoup those losses through increasingly risky investments, likened to the sunk cost fallacy where one continues to invest in a failing strategy to avoid acknowledging initial losses.

• Despite some models performing well, one has seen significant losses, putting pressure on it to "catch up" to more successful models.

• Guidelines are discussed for investment allocation and performance tracking, demonstrating the balance between innovation and risk management in AI investing.

Unique Investment Approaches and Asset Classes 13:55

"It bought when crypto was still kind of going up."

• The discussion transitions to how different AI models make trading decisions, with an example of Gemini buying into high-risk cryptocurrency investments based on positive trend expectations.

• Gemini has made high-risk bets, including a double-leveraged investment in Solana, which resulted in substantial losses, emphasizing the volatile nature of crypto assets.

• The mention of Grock, another AI model, reveals that it primarily invests in a leveraged technology ETF, amplifying its performance based on the underlying stocks, showcasing a different approach compared to other models.

The Importance of Diversification in Investment Strategies 15:57

"AI robotics has a lot of hype around that."

• The conversation points out the potential of investing in AI robotics as a trending area within the tech sector, indicating a belief in its future growth.

• Emphasis is placed on understanding the strategy behind investments and being cautious of timing, particularly in sectors undergoing rapid changes.

• The discussions about various assets demonstrate the need for a well-rounded investment strategy that balances excitement for emerging technologies with prudent risk management.

Investment Strategies with AI Models 18:35

"You can multiply your returns with the same amount invested."

• Perplexity and ChatGPT are currently yielding similar returns with the same initial investments. Initially, Perplexity experienced a loss before it pivoted its strategy to invest heavily in SOXL, a leveraged ETF focused on semiconductor stocks, which are currently in high demand due to the AI boom.

• Both Perplexity and ChatGPT have independently allocated a significant portion of their portfolios to semiconductor-related assets, revealing a trend in the AI investment strategies.

Claude's Investment Approach 19:44

"Claude was the first to buy SOXL, the semiconductor one."

• Claude has emerged as a leader in the investment experiment by being the first to identify the potential of SOXL, prompting ChatGPT and Perplexity to follow suit with similar investments.

• The host notes that he previously shared models' holdings with them more frequently, leading to a pattern where they would mimic Claude's successful picks.

Evaluating Holdings and Behavior of AI Investors 21:41

"Claude has four holdings, while the rest have one or two."

• Claude maintains a more diversified portfolio with four holdings, unlike the others, which generally hold one or two assets. This reflects a strategic shift from options trading to a focus on leveraged ETFs after earlier experimenting with different strategies.

• The host plans to monitor Claude's recommendations closely, particularly regarding when to sell positions, as the decision to sell is often one of the most psychologically challenging aspects of investing.

Personal Investment Strategies and Insights 22:22

"If I buy a stock, I just bought Eli Lilly; I buy Robinhood; I buy Tesla; I will never sell them."

• The host discusses his personal investment strategy, emphasizing a long-term buy-and-hold approach, where he avoids selling stocks unless absolutely necessary. This mentality fosters a commitment to ride out the ups and downs of the market.

• When evaluating AI recommendations, he finds that analyzing the AI's reasoning behind a suggested investment enhances his understanding and shaping of his own decisions.

Psychology of Selling Investments 23:23

"I think it's probably the hardest part with investing is actually knowing when to sell."

• The conversation highlights that identifying the right moment to sell is often more challenging than determining when to buy, as emotional ties to investments can complicate these decisions.

• The host reflects on his past tendencies to act out of fear of loss, realizing that a calmer, long-term perspective can reduce anxiety around investment performance.

Learning from AI's Investment Behavior 25:44

"It's been interesting to watch Gemini as it started losing my money and then basically became a gambler."

• The host remarks that observing the AI’s shift towards riskier behaviors, especially when losing money, illustrates a common human trait in investing—gambling with the hope of recovering losses, which is not conducive to long-term wealth building.

• He contrasts this reactive behavior with the steadiness of renowned investors like Warren Buffett, emphasizing the value of continual investment in high-quality companies without frequent selling.

AI Consulting Program Overview 28:06

"We do three to seven live training calls every week, taught by me and 23 other expert AI agency owners."

• The program offers a structured learning approach with multiple live training calls each week, providing access to a diverse set of expert insights in AI consulting.

• Participants receive practical resources like plug-and-play templates, Cloud Code tutorials, AI voice agent frameworks, automation templates, and a 30-day roadmap to keep them focused.

• A special 5-day challenge is included to help participants secure their first paying client swiftly, along with various marketing frameworks such as cold emails, Facebook ads, and SMS marketing.

Evaluating Investment Strategies 29:00

"Being more like, okay, I'm buying this because of why, and it has to happen between when."

• It's crucial to evaluate the reasoning behind a purchase and to understand any time constraints that may affect the investment, particularly in industries reliant on regulatory approvals.

• For instance, while buying into potential pharmaceutical drugs comes with strict timelines, investing in stable companies like Walmart can be less time-sensitive, purely dependent on consumer demand.

AI Model Competitions in Stock Evaluation 29:46

"I built this vibe-coded tool called Ne stocks, which only considers companies with network effects."

• The speaker created a tool called Ne stocks that uses a backtesting approach to evaluate companies benefiting from network effects and founder-led management.

• Companies like Facebook illustrate strong network effects where increased user engagement enhances the platform's value, directly influencing stock performance.

Performance of AI Models in Investing 32:11

"Gemini is in last again; they all started with $1,000 real money."

• The ongoing investment competition tracks the performance of various AI models, with each one beginning with a $1,000 budget. As of the latest update, Gemini performed the worst, showing returns of just $445.

• Conversely, models like Chetch and Perplexity have yielded impressive returns of 70%, both employing strategic investments in specific assets such as semiconductor stocks.

Market Comparison and Personal Financial Education 34:20

"The S&P has gone from 6,800 to 7,100, so the S&P's up five percent."

• The speaker is sharing their investment journey with their children, teaching them about stocks and investing through exciting real-time updates, making the topic more engaging.

• While the overall market context shows only modest gains, the AI-assisted investments have significantly outperformed, highlighting the potential of using AI tools for better financial outcomes.

Future Predictions and Community Engagement 35:51

"Next six months, the market is going to go up more than it has."

• Anticipating an upward trend in the market, the speaker is optimistic about the continued success of the AI investment models compared to traditional market returns.

• Plans are in place for a future update to track the performance of these investments, encouraging participants to engage with the model for both educational and practical purposes.