CEO of Liquid AI Ramin Hasani Says a Worm is Changing the Future of AI

Mar 17, 2025

57:03

forum

Ask episode

view_agenda

Chapters

auto_awesome

Transcript

info_circle

Episode notes

A new era of intelligent technology is here.

Today on Digital Disruption, we’re joined by Ramin Hasani, co-founder and CEO of Liquid AI and a machine learning Scientist.

Ramin Hasani is the co-founder and CEO of Liquid AI and a machine learning scientist at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL). Previously, he was a Principal AI & ML Scientist at Vanguard and a Research Affiliate at MIT. His work focuses on robust deep learning and decision-making in complex dynamical systems. He earned his Ph.D. in Computer Science from Vienna University of Technology, with his research on Liquid Neural Networks receiving global recognition with numerous nominations and awards such as TÜV Austria Dissertation Award nomination in 2020, and HPC Innovation Excellence Award in 2022. Ramin is also a frequent TEDx speaker.

Ramin sits down with Geoff to share his new and unique approach to AI, inspired by biological and physical systems that could redefine how businesses use machine learning.

Ramin shares insights from biology, specifically the neural structures of simple organisms that have led to the development of Liquid Neural Networks (LNNs). Unlike traditional neural networks, LNNs use differential equations to model dynamic decision-making processes, making them more adaptable for real-world applications. They discuss the advantages of these approaches and the scaling of these models. They also talk about the future of AI and how it may not just be about bigger models but smarter, biologically inspired intelligence capable of transforming how we interact with technology today and into the future.

In this episode:

0:00 Intro

0:38 The story of Liquid AI technology

6:13 Transparent worm vs. traditional neural networks

14:15 Liquid AI’s mission to build powerful AI systems

18:12 Liquid AI will help you to develop the best kind ofmodels

20:03 Compelling use cases

22:12 Agentic behavior

26:30 The most surprising use cases

29:35 This is a better learning system

32:01 Efficiency of liquid foundation models as opposed toGPT

35:33 Is this tech ready for the market?