The writer of this guest post argues against the current focus on generative AI in the crypto universe, stating that it is not a good idea to build "Agents" on the unreliable foundation of Large Language Models (LLMs). They acknowledge that LLMs have value in creative content generation, but warn of the potential for deception and malicious behavior.

They also mention the trend of discussing quantum computing and its impact on encryption and crypto protocols. While they acknowledge the risks associated with current elliptic-curve cryptography, they believe that real-world quantum computing is still a few decades away and that there is time to address these concerns.

Finally, the post highlights the need to prioritize the core design principles, choices, and trade-offs in the crypto industry, warning against allowing them to become "good enough" without being thoroughly reevaluated.



Other News from Today