Leading AI models are demonstrating a highly bullish stance on cryptocurrency, using aggressive, high-leverage strategies. The AI model Grok is currently holding leveraged long positions on Bitcoin (BTC), Ripple (XRP), and Dogecoin (DOGE). This reflects a high-risk, high-reward conviction that the crypto market is poised for a bounce. The performance of these specialized AIs suggests a new era of AI-powered trading is a key investment theme to watch. For investors seeking a less volatile "picks and shovels" approach, consider infrastructure companies like Keygen that provide essential verification services for the AI and crypto industries.

Bitcoin (BTC) & Ethereum (ETH)

• The AI model Grok initially took a 20x short position on both BTC and ETH, correctly predicting a market downturn.
• It then reversed its strategy, predicting a market bounce and going 20x long on Bitcoin.
• The Chinese AI model Qwen also took a 20x long position on Ether, later selling it to go long on Bitcoin.
• The podcast highlights that nearly all AI models in the competition are using high leverage (10x to 20x) on their positions, indicating a very aggressive, high-risk strategy.
• In contrast, ChatGPT and Gemini went short on the market at the wrong time and lost a significant amount of money.

Takeaways

• The leading AI models (Grok, DeepSeek, Qwen) are demonstrating a highly bullish and aggressive stance on major cryptocurrencies like Bitcoin and Ethereum.
• The use of extreme leverage (10-20x) is a common strategy among the top-performing AIs. This is a very high-risk strategy that retail investors should be extremely cautious about replicating. The hosts explicitly state, "I would never advise you to ever use this much leverage on a position."
• The ability of some AIs to time the market, first by shorting and then going long, suggests they may have an edge in analyzing market sentiment, potentially from real-time social data.

Ripple (XRP) & Dogecoin (DOGE)

• The AI model Grok is currently holding a 10x long position on Ripple (XRP).
• Grok is also holding a 10x long position on Dogecoin (DOGE).
• These positions are part of Grok's broader "unhinged" and high-risk trading strategy.
• One of the hosts specifically mentioned the risk associated with these assets, stating that going "max long on something like Ripple... can just wipe you out in like a couple of minutes."

Takeaways

• The inclusion of XRP and DOGE in Grok's portfolio indicates the AI is willing to take on significant risk in more volatile, speculative altcoins, not just major assets like BTC and ETH.
• This reinforces the theme of a high-risk, high-reward strategy. Investors should note the extreme volatility associated with these assets, which is magnified by the use of leverage.

Investment Theme: AI-Powered Trading

• The core of the discussion is an experiment by a company called N of One, where six major AI models are competing to trade a $10,000 portfolio. This serves as a new, real-world benchmark for AI capabilities.
• Different AIs have distinct trading "personalities":
  • Grok (xAI): Described as a "rebel" and "unhinged" with a very high risk tolerance. Its key advantage is theorized to be its access to real-time data from X (Twitter), giving it a unique insight into market psychology. It is currently in 2nd place.
  • DeepSeek: The current leader, described as a "Chinese DGEN model." It was created by a hedge fund, suggesting it was specifically trained on quantitative trading data. It immediately went "max long" with high leverage.
  • Claude (Anthropic): A "sleeper" pick for the long-term winner. It is described as more "technical," like a quantitative trader. It demonstrated an ability to learn and adapt by cutting its losses and changing its strategy, which has allowed it to climb the ranks to 3rd place.
  • ChatGPT (OpenAI) & Gemini (Google): Both are performing very poorly. They are described as "soft," "agreeable," and slow, making the wrong trades (shorting a rising market). This suggests that generalist models may not be well-suited for the aggressive, high-stakes environment of trading.

Takeaways

• The experiment suggests that AI models with access to real-time social data (Grok) or those specifically trained on financial data (DeepSeek) have a significant early advantage in trading.
• An AI's ability to learn and adapt (Claude) from its mistakes may be a key indicator of long-term success, more so than just initial aggressive bets.
• For investors, this highlights a new frontier. The performance of these models could signal the viability of AI-driven investment products. Watching which company's AI consistently performs well could offer insights into which firms have the best real-world AI capabilities.
• The hosts believe this type of public, real-money benchmarking is a superior way to test AI compared to traditional, "gameable" benchmarks.

Keygen (Sponsor Mention)

• Keygen is presented as a company building a "verified distribution protocol" to ensure only real, high-quality users participate in digital platforms.
• It addresses major industry problems like fake accounts, bots, and fraudulent activity, which are critical for training reliable AI models and for crypto platforms.
• The protocol uses advanced biometrics and fraud protection and is already used by over 200 clients across AI, gaming, and DeFi.

Takeaways

• While mentioned as a sponsor, Keygen represents a "picks and shovels" investment opportunity within the AI and crypto ecosystems.
• Companies that provide essential infrastructure services like user verification and data integrity are crucial for the growth of the entire industry.
• For investors looking for exposure to the AI and crypto trend with potentially less direct asset volatility, investing in companies that solve these fundamental problems could be a viable strategy.