People & AI - Future of AI: 2024 Prediction on Regulation, Risk, and Innovation

In the episode of the 'People & AI' podcast, titled 'Future of AI: 2024 Prediction on Regulation, Risk, and Innovation,' the co-founders of Armilla AI, Dan Adamson, Phil Dawson, and the host Karthik Ramakrishnan, discuss the major developments in the AI industry of 2023 and provide some of their biggest predictions for AI in the coming year, 2024.
March 1, 2024
5 min read

The conversation starts with a reflection on the seminal events of 2023, notably the productization of general-purpose models like OpenAI's GPT, which has led to widespread adoption and regulatory considerations of AI. The discourse then turns to how AI has begun to infringe upon creative and white-collar professions, creating a mix of excitement and apprehension for the future.

Dan Adamson highlights the potential for harms when AI systems are deployed without proper testing and regulation, suggesting that the haste in bringing AI products to the market can lead to significant losses, as seen with Google's market cap following their AI launch debacle. Adamson points out that while AI regulation may have 'teeth' to address current harms like bias, future existential risks may require a separate framework.

Discussing the various government and organizational focuses on AI regulation, Phil Dawson comments on the convergence toward AI safety and risk management. With the anticipated EU's AI Act, Dawson envisages a shift toward a more risk-based approach, with potential impacts on short- and medium-term risks of AI deployment.

Karthik Ramakrishnan emphasizes the need for regulation of AI models, mentioning that even though the application to a relatively new technology like AI complicates matters, the core regulatory concerns are not novel. He brings to light the importance of involving existing regulatory authorities to ensure AI is incorporated responsibly within their domain.

Some of the predictions for 2024 shared by the co-founders include:

- An EU AI Law will become a reality.

- Generative AI's significant role in the build-up to the US elections.

- A shift towards parity in spending on internally built models versus purchased models in the enterprise context.

- Resolution to major questions around copyright, privacy, and standards.

The episode displays a mix of technical foresight and policy insights, capturing a comprehensive picture of the possible AI landscape for the coming year, as seen through the lens of industry experts.

Listen on Apple podcasts: https://lnkd.in/ga4t4WuZ

and on Spotify: https://lnkd.in/gBzmKsDE