An epic AI Debate—and why everyone should be at least a little bit worried about AI going into 2023. Noam Chomsky, Gary Marcus, Dileep George, Yejin Choi, and Francesca Rossi all expressed their concerns about the current approach to artificial intelligence. Each of them focused on different aspects, but all shared a common worry about whether AI could truly replicate the complexities of the human mind.

An epic AI Debate—and why everyone should be at least a little bit worried about AI going into 2023. Noam Chomsky, Gary Marcus, Dileep George, Yejin Choi, and Francesca Rossi all expressed their concerns about the current approach to artificial intelligence. Each of them focused on different aspects, but all shared a common worry about whether AI could truly replicate the complexities of the human mind.

Chomsky, a renowned linguist and cognitive scientist, was concerned about understanding what makes the human mind what it is. He questioned whether the current approach to AI could ever provide insights into this fundamental aspect of humanity.

Marcus, an AI researcher, focused on four key aspects of thought that he believed any intelligent machine should possess: reasoning, abstraction, compositionality, and factuality. He worried that contemporary approaches to AI were not adequately addressing these aspects.

George, a DeepMind researcher and AI startup cofounder, raised an interesting analogy with dirigibles like the Hindenburg. He expressed concern that scaling alone would not be enough to achieve general intelligence, just as dirigibles were eventually surpassed by airplanes.

Choi, a MacArthur-winning AI professor, raised important questions about the "dark matter of AI", specifically commonsense reasoning. She also discussed the ethical implications of AI and the need for value pluralism and ethical reasoning.

Rossi, an IBM Fellow and President of AAAI, focused on the ethical behavior of AI systems. She argued that current approaches may not be sufficient in ensuring AI systems behave ethically and emphasized the importance of involving humans in the loop.

While these concerns may seem disparate, they all point to the need for a more comprehensive and holistic approach to AI development. It is clear that there are still many challenges to overcome in order to achieve true artificial intelligence.

In addition to these concerns, there is another aspect that needs to be considered: the rise of aggregators. Aggregation Theory describes how platforms, or aggregators, dominate industries in a systematic and predictable way. These aggregators have three key characteristics: a direct relationship with users, zero marginal costs for serving users, and demand-driven multi-sided networks with decreasing acquisition costs.

Aggregators thrive on the abundance of digital goods, which allows them to provide value through discovery and curation. As more suppliers join the platform, customer acquisition costs decrease, leading to winner-take-all effects. Netflix is a prime example of a Level 1 Aggregator, while Google and social networks fall under Level 3 Aggregators.

The rise of super-aggregators like Facebook and Google, which operate multi-sided markets with zero marginal costs, raises concerns about their control over user relationships. This ownership of the user relationship gives aggregators significant power and influence, which may require regulation to ensure fair competition and protect users.

Regulating aggregators is a complex task, as they offer superior services that users willingly choose. Traditional regulatory measures, such as breaking up companies or limiting their markets, may not be effective in the digital age. Instead, a more nuanced approach that balances competition and user protection may be necessary.

In conclusion, the AI debate highlights the need for a deeper understanding of the human mind and the challenges of achieving true artificial intelligence. The concerns raised by Chomsky, Marcus, George, Choi, and Rossi all point to the complexities and limitations of current AI approaches. Additionally, the rise of aggregators in various industries raises questions about user relationships and the need for regulatory measures to ensure fair competition and user protection.

To address these issues, here are three actionable pieces of advice:

1. Foster interdisciplinary collaboration: AI development requires insights from various fields, including linguistics, cognitive science, and ethics. By bringing together experts from different disciplines, we can gain a more comprehensive understanding of the human mind and develop AI systems that better replicate its complexities.

2. Encourage research on reasoning and abstraction: Reasoning, abstraction, compositionality, and factuality are crucial aspects of thought that AI systems need to possess. Investing in research and development in these areas can help bridge the gap between current AI capabilities and human-level intelligence.

3. Implement responsible AI practices: Ethical considerations must be at the forefront of AI development. Involving humans in the loop and ensuring the ethical behavior of AI systems are essential steps. This can be achieved through stringent regulations, transparency, and accountability measures.

By addressing these concerns and taking actionable steps, we can navigate the future of AI with a more comprehensive understanding and ensure that its development aligns with our values and aspirations.