Rethinking Intelligence - #27 with Borhane Blili-Hamelin

Episode #27 of The Only Constant is out now - and I am extremely excited for this one. Recorded just before the summer, I relistened to this episode during my vacation in Japan, and it is one of my all time favourites. Borhane dives skillfully into the tricky waters of what intelligence is, how to balance risk and opportunity, and how auditing processes underscore value and trust. This is an episode you can't miss if you are in the AI, risk or innovation space.

Read more about Borhane Blili-Hamelin at the bottom of this newsletter.

As always, I am very happy that we once again managed to make an episode that explores more than it explains ??

Listen to the episode here:

Apple: The Only Constant on Apple Podcasts

Spotify: The Only Constant | Podcast on Spotify

Google: The Only Constant ( google.com )

Spreaker: The Only Constant ( spreaker.com )

What made this episode special to me?:

I was introduced to Borhane, or Bo for short, by Sree Sreenivasan after having recorded episode 2 of the podcast. Sree himself had met Bo somewhere at an event in NYC and though he and I should get in touch. Since then, we've had something like 3 deep dive virtual coffees that always went beyond and above the usual chats. Bo is razor sharp, speaks his mind, is inquisitive and just raises the bar and I loved our conversations so much I had to have him on the podcast as well.

And am I glad we got this one recorded! Bo tackles some of the more delicate subjects of AI with an intellectual grace I have seldom encountered. Practical definitions of intelligence and bias are coupled with equally pragmatic approaches to risk and innovation, while never losing precision or reverting to banalities. In some ways, I feel episodes like this was exactly why I decided I wanted to do the podcast in the first place - but perhaps that is exactly also due to my very own bias, being a historian with a knack for philosophy in a world of technology.

My key post-recording deliberations are:

• Are we done imagining intelligence? Do we know what it is and can we even say what isn't intelligence with certainty? Bo makes it brillianty relative: it is all about what we deem desireable and nothing else. This also means it is hyper normative and that intelligence in some ways is always retrospectively evaluated based on its desireability.
• Innovation and risk are two sides of the same coin and success is balancing risk with innovation. This realisation has to come to the fore and break down the hype that is currently creating two AI camps: one that only sees opportunity and one that only talks about risks, ethics, problems.
• Foundational models are not platforms, but solutions to AI supply chain. How many enterprises have implemented it as a platform expecting it to solve their productivity! Foundational models can provide AI quickly and out of the box, but you have to cater it to your use case, problem, people.

This episode made me think on so many levels, and I hope everyone listening in will feel the same way I did about it.

About Borhane Blili-Hamelin, PhD :

Dr. Borhane Blili-Hamelin is an AI ethics researcher and consultant dedicated to mitigating sociotechnical AI harm. He develops practical strategies for AI governance, risk assessments, responsible disclosure, and audits. His research explores leveraging red-teaming to empower communities, creates AI auditing and governance frameworks, and critiques practices like benchmarks and concepts like AGI. His workshops to empower cross-disciplinary collaboration in responsible AI have been featured at FAccT, RightsCon, and Mozilla Festival.

He holds a PhD in philosophy from Columbia University. Borhane is an officer at the AI Risk and Vulnerability Alliance, an affiliate at Data & Society, and a senior consultant at BABL AI.