Artificial Intelligence and You

In all the news about and focus on startups like OpenAI and Anthropic, there’s a long-established player that doesn’t get as many soundbites but which deserves our attention: IBM. It’s important to understand what their current strategy and thinking is. And I have just the person to do that: Ruchir Puri, the Chief Scientist of IBM Research, and Vice-President of IBM Corporate Technology and Technical Community.

He led IBM Watson as its CTO and Chief Architect from 2016-19 and is a Fellow of the IEEE, and has been an ACM Distinguished Speaker, an IEEE Distinguished Lecturer, and was honored with John Von-Neumann Chair at Institute of Discrete Mathematics at Bonn University, Germany. Notably from our point of view, he is an inventor of over 70 US patents and has authored over 120 scientific papers, which lands him squarely in the technical perspective that we value so much on this show. 

We talk about the continuing importance of the foundation of computer science, which is algorithms, the tension between the deterministic nature of algorithms and the more touchy-feely variability of generative AI interfaces, what that means for reliability and verifiability, how best to think about Gen AI, the history of IBM’s big AI projects like Watson and Deep Blue, and IBM’s ambitions in quantum computing and the role of AI in that.

All this plus our usual look at today's AI headlines!

Transcript and URLs referenced at HumanCusp Blog.

I have been talking with Maurice Chiodo, Assistant Research Professor at the Centre for the Study of Existential Risk at the University of Cambridge, about his new paper titled Reframing AI Loss of Control: What Control Is, How to Have It, How to Lose It and everything that control of AI and loss of it means. Maurice is the principal investigator of the Ethics in Mathematics Project, holds two PhDs in mathematics, from Cambridge and the University of Melbourne, and has over a decade of experience working as an academic mathematician. He has direct industry experience with over 30 startups, having been a member of the Ethics Advisory Group at Machine Intelligence Garage UK for over 2 years.

In this conclusion of the interview we talk about AI in wartime decision making, the mechanics of real-time detection of loss of control, the problems with putting a human in the loop, and some of the most serious scenarios of loss of control.

All this plus our usual look at today's AI headlines!

Transcript and URLs referenced at HumanCusp Blog.

With so much talk about the need to control AI, and notable incidents of losing control of it, like the OpenAI/HuggingFace incident among others, we will do a deep dive on what control or loss of control of AI means with the help of Maurice Chiodo, Assistant Research Professor at the Centre for the Study of Existential Risk at the University of Cambridge, and the principal investigator of the Ethics in Mathematics Project, where he addresses the ethical challenges and risks posed by mathematics, mathematicians, and mathematically-powered technologies. Maurice holds two PhDs in mathematics, from the University of Cambridge and the University of Melbourne, and has over a decade of experience working as an academic mathematician on problems in algebra and computability theory. He has direct industry experience with over 30 startups, having been a member of the Ethics Advisory Group at Machine Intelligence Garage UK for over 2 years. But his recent paper titled Reframing AI Loss of Control: What Control Is, How to Have It, How to Lose It, forms the main reason for our interview.

We talk about ethical issues in mathematics and whether there should be a Hippocratic oath for mathematicians, AI’s impact on mathematics in solving previously unsolved problems, how to define control and losing it (using sharks as an example), and the difference between operational, tactical, and strategic goals. 

All this plus our usual look at today's AI headlines!

Transcript and URLs referenced at HumanCusp Blog.

We are back on the consciousness train again, trying to understand what that thing is. So I've been talking with Joscha Bach, renowned cognitive scientist who uses the lens of artificial intelligence to understand the human mind. He has taught and researched at Humboldt University of Berlin, the Institute for Cognitive Science in Osnabrück, the Berlin School of Mind and Brain, MIT’s Media Lab and the Harvard Program for Evolutionary Dynamics. He’s the creator of the MicroPsi cognitive architecture, the author of Principles of Synthetic Intelligence, and the executive director of the California Institute for Machine Consciousness.

We conclude by talking about how to know if something is conscious, the role of memory, what Joscha thinks of Integrated Information Theory, applying the Turing Test to LLMs and whether they’re entitled to rights, substrate independence, or the portability of consciousness from one platform to another, the possibility of making a conscious AI, and the geopolitics of doing so. 

All this plus our usual look at today's AI headlines!

Transcript and URLs referenced at HumanCusp Blog.

We are back on the consciousness train again, trying to understand what that thing is. Most of us would cop to being conscious, but how many of us have any idea how to describe it, let alone test for it? Here to help with that is Joscha Bach, cognitive scientist, AI researcher, and philosopher who uses the lens of artificial intelligence to understand the human mind. He has taught and researched at Humboldt University of Berlin, the Institute for Cognitive Science in Osnabrück, the Berlin School of Mind and Brain, MIT Media Lab and the Harvard Program for Evolutionary Dynamics. He’s the creator of the MicroPsi cognitive architecture and the author of Principles of Synthetic Intelligence. At Intel Labs, he led a team for cognitive AI. He is currently the executive director of the California Institute for Machine Consciousness where he draws on computer science, neuroscience, psychology, and philosophy to ask how minds construct reality, how consciousness might emerge from computation, and what AI can teach us about ourselves.

We talk about the Institute’s mission, the machine consciousness hypothesis, why we actually are in a particular kind of simulation, dreaming, the role of perceptions and predictions, and a fascinating take on the Chinese Room.

All this plus our usual look at today's AI headlines!

Transcript and URLs referenced at HumanCusp Blog.

The idea of AI posing an existential risk to the human race is kind of like the Terminator itself – it just keeps coming back. We'll look at some of the history of that issue, how attitudes towards it have waxed and waned over the years, and recent events where AIs have escaped their containers and the reactions - or lack thereof - to those incidents. What should we take away from this slow-motion train wreck and what should we do about it?

All this plus our usual look at today's AI headlines!

Transcript and URLs referenced at HumanCusp Blog.

What is AI doing to medicine? Specifically, in molecular biology, and protein folding: figuring out the 3-D structure of proteins. Remember DeepMind's AlphaFold and the announcement of it predicting 200 million protein structures?

I'm talking with one of the top researchers in the world, Professor Mohammed AlQuraishi of Columbia University’s Department of Systems Biology, and a member of their Program for Mathematical Genomics. He heads a lab focused on biological perspectives at the molecular and systems levels. He has an MS in statistics and a PhD in genetics from Stanford University and was a Departmental Fellow at Harvard Medical School.

We conclude by talking about how the advent of AlphaFold landed in the protein folding community; how AI is about to transform molecular biology and scientific experimentation; the possibility of simulating a cell; OpenFold, Mohammed’s answer to the closed source parts of AlphaFold; and his vision for the next 10 years of the field.

All this plus our usual look at today's AI headlines!

Transcript and URLs referenced at HumanCusp Blog.

What is AI doing to medicine? Specifically, in molecular biology, and protein folding: figuring out the 3-D structure of proteins, which look like the output of a pasta-making machine on LSD. Remember DeepMind's AlphaFold and the announcement of it predicting 200 million protein structures?

I have questions. And here to answer them is one of the top researchers in the world, Professor Mohammed AlQuraishi of Columbia University’s Department of Systems Biology, and a member of their Program for Mathematical Genomics. He heads a lab focused on biological perspectives at the molecular and systems levels. He has an MS in statistics and a PhD in genetics from Stanford University and was a Departmental Fellow at Harvard Medical School. As if that wasn’t enough, before this storied academic career, he founded two startups in the mobile computing space. 

We talk about Mohammed’s background in molecular biology, the big goals of that field and his North Star of understanding biology well enough to simulate it; why I haven’t yet seen the kind of revolutionary impact I was expecting from AlphaFold, what it means for a protein to fold and why it’s a big deal and how AlphaFold and similar models work at figuring that out.

All this plus our usual look at today's AI headlines!

Transcript and URLs referenced at HumanCusp Blog.