Maxim Afanasyev has been coding large language models for over 20 years, long before the AI boom. He's an applied AI expert who's worked at McKinsey building AI transformations and now leads AI strategy at Alphabet.
Maxim and Anthony discuss why 95% of AI applications don't deliver bottom-line impact, the global shortage of applied AI experts (only 200 mature professionals worldwide), and why banks are solving yesterday's problems instead of tomorrow's value propositions.
They cover why LLMs became persuasion machines by learning human biases, how customers are now smarter than banks because of AI agents, and why agentic payments will force financial institutions to rethink their entire business model. Maxim also explains why cutting 20% of your workforce signals that your CEO failed at transformation.
Timestamps
00:00 Intro
01:08 Coding LLMs 20 Years Ago in a Soviet-Era AI Lab
06:02 How RLHF Turned Neural Networks into Machines for Persuasion
10:48 From AI Winter to AI Hype: What Actually Changed
13:40 Why 95% of AI Applications Fail to Deliver Bottom-Line Impact
20:22 The 200 Applied AI Experts Problem
25:27 How to Nurture Applied AI Talent Through Experimentation
32:52 Why Bank Chatbots Create Dead Weight Loss
42:48 How AI-Smart Customers Are Breaking Legacy Banking Models
52:08 Asia's Lead in AI Adoption & Why Banks Must Rethink Value
01:04:14 Agentic Payments, Trust & the Renaissance of Physical Experience
01:16:25 Agents, Regulation & the Future of Human-AI Interaction
01:25:00 AI Deepfakes, Recruitment Fraud & the URL-to-IRL Shift
01:35:04 Why CEOs Who Cut Jobs for AI Are Really Admitting Failure
- Follow Maxim Afanasyev
- Follow Anthony Sar
Hosted on Acast. See acast.com/privacy for more information.