AI risks, RLHF, and inner alignment: GPT stands to give the business world a major boost. But with everyone racing either to develop products that incorporate GPT or use it to carry out critical tasks, what dangers could lie ahead in working with a tool that applies essentially unknowable means (inner alignments) to reach its goals? This week’s guest Jérémie Harris speaks with Jon Krohn about the essential need for anyone working with GPT to understand the impact of a system comprising inner alignments that cannot – and may never – be fully understood.

Additional materials: www.superdatascience.com/668

Interested in sponsoring a SuperDataScience Podcast episode? Visit JonKrohn.com/podcast for sponsorship information.

The Super Data Science Podcast: Machine Learning | AI | Success

668: GPT-4: Apocalyptic stepping stone?

What happens when AI starts mimicking not just our intelligence, but our irrationality? In this episode, we dive into the strange new world of “semantic confabulations”, - those eerily human-like errors large language models make when meaning and truth part ways. From hallucinated headlines to cognitive dissonance at machine scale, we explore how today’s AI systems reflect our mental quirks and what it means for trust, leadership, and decision-making. Tune in for sharp insights, uncanny examples, and a look at whether the real problem with AI… is how much it sounds like us.

Bit & Being - Semantic Dissonance at the AI-Human Intersection

P1, S1: Discussing the research paper - AI Alignment: A Comprehensive Survey (2024) It's an extensive survey on AI alignment, which focuses on ensuring AI systems operate according to human intentions and values. It introduces four core objectives—Robustness, Interpretability, Controllability, and Ethicality (RICE)—that guide alignment efforts.

1-AI_Alignment__Navigating_the_Future_of_Intelligent_Systems

P2, S1: Second pod discussing the research paper - AI Alignment: A Comprehensive Survey (2024). Key areas explored include learning from feedback, addressing distribution shifts, assurance techniques like safety evaluations and interpretability, and governance practices by various stakeholders. The text emphasizes the socio-technical aspects of AI alignment, highlighting the evolving challenges as AI becomes more integrated into society.

2-Navigating_AI_Alignment__Risks,_Research,_and_the_Future_of_Human_Control

The Aligned Sovereign Intel. Core cast - we cover all things aligned AI

The latest machine learning, A.I., and data career topics from across both academia and industry are brought to you by host Dr. Jon Krohn on the Super Data Science Podcast.

As the quantity of data on our planet doubles every couple of years and with this trend set to continue for decades to come, there's an unprecedented opportunity for you to make a meaningful impact in your lifetime. In conversation with the biggest names in the data science industry, Jon cuts through hype to fuel that professional impact.

Whether you're curious about getting started in a data career or you're a deep technical expert, whether you'd like to understand what A.I. is or you'd like to integrate more data-driven processes into your business, we have inspiring guests and lighthearted conversation for you to enjoy.

We cover tools, techniques, and implementation tricks across data collection, databases, analytics, predictive modeling, visualization, software engineering, real-world applications, commercialization, and entrepreneurship − everything you need to crush it with data science.

New tracks tagged #AI alignment