Exploring Complexity Ep. 2
Why the world feels more complex—and why that feels hard. Why more problems are complex problems. Why organizations struggle with complexity.
Recent research demonstrates through empirical evidence that GPT-4 can autonomously develop pricing strategies that edge towards collusion without explicit human direction or inter-firm communication.
Key Points:
The integration of Large Language Models into business operations offers new capabilities in data analysis, decision-making, and automation. However, this progress comes with complex challenges, particularly in the domain of algorithmic pricing.
In the near future, it's entirely possible that business owners looking to set prices fairly and independently might turn to a tool they use for many other decisions: AI-based large language models, or LLMs. These business owners might explain their market to the AI and instruct it to aim for long-term profits without even hinting at the idea of working unfairly with others. Despite not fully understanding how these AI systems work, these owners would trust them not to engage in any unfair practices. Some might even directly ask their AI if it would act unfairly, only to be assured it wouldn't—yet, as we've seen, there's a chance these AIs could still end up acting in ways that seem unfair, without the business owners intending or realizing it.
The Artificiality Weekend Briefing: About AI, Not Written by AI