Q&A: Ex-Google lead Jad Tarifi on avoiding dogmatic AGI

In partnership with

A quick note: Some of you may know that my brother lost his battle with mental health last week. This piece was written before that. I hope you enjoy it—Jad Tarifi is an intelligent and creative guy with a strong care for the world, and we need more people like him.

At this point, I plan to continue posting on my existing cadence (the 1st and 3rd Tuesdays of each month). It’s not an overwhelming schedule, but my productivity is currently quite slow and intermittent, so my overall workload has temporarily decreased. There are a few ways you can support me:

1️⃣ Most helpful would be to share this newsletter with your network. The more subscribers I get, the more likely I am to get more lucrative sponsorships. 2️⃣ You can also send me a “tip” at the link below if you’re financially willing and able. I understand the pressures of the current economy, so please don’t feel like this is a precursor to reading my work. You are welcome here regardless. 3️⃣ At the end of this article is an ad for a cool news platform. Feel free to click it if it interests you. I earn a few bucks per click.

Now on to the story…

The forecasted future of AGI (artificial general intelligence, or “AI with capabilities that rival those of a human,” according to McKinsey) continues to progress, even as the U.S. faces a radical shift in governance that, to many, feels like a good enough reason to stop the world.

Jad Tarifi is one such person working towards AGI, but he’s doing it from a markedly moral (but not dogmatic) standpoint.

Tarifi is CEO and co-founder at Integral AI, a company working on the building blocks of AGI. He previously worked at Google for nine years, where he started their first generative AI team.

“We've been working at the intersection of new AI models and robotics, and inching closer towards my lifelong dream of building AGI and using that to change the world,” said Tarifi.

Jad wrote a book called “The Rise of Superintelligence: A guidebook for human emancipation in the age of AGI” that emphasizes how this technology could—ideally—serve as the launchpad for a more free and just global society.

“The future is completely open-ended. We could all die tomorrow. And it's really quantum mechanics that tells you [that],” said Tarifi. “We have a role to play in our decisions and how we embody and choose the future. We need to paint a positive vision to know where we want to go. If we're gonna try to fight for something, let's try to fight for something positive. But it's far from guaranteed…If we make poor choices, I don't think we're going to be around for too long, and the universe would recycle us.”

If you’ll forgive me for a brief editorialization: Frankly, it makes me feel more comfortable that someone like Tarifi is working on AGI when I’m sure there are others with a foggier moral compass taking on the beast.

Ideals vs. political power play

In his book, Jad encouraged readers to seek out diverse perspectives, especially those that challenge their existing beliefs. But he hardly sees this practice in mainstream tech.

“Currently in Silicon Valley, a lot of tech leaders tell themselves they’re embracing diverse perspectives, but in a deeper sense, they're not,” said Tarifi. He isn’t sure whether the claims of open-mindedness are meant to be genuine or just a recruiting tool, but he knows they generally don’t succeed. “Can I see some leaders that have been doing that? I generally am very critical, but back when Larry Page was running Google, I think that was a little bit more positive. It has completely gone downhill since then.”

Larry Page, co-founder and former CEO of Google

Jad didn’t leave Google to start a company. He just wanted to change the world, and he thought AI was the best path for that. Unfortunately, he found a lot of “organizational inertia” at Google.

To achieve AGI, he also needed a hand in robotics, which was outside of Google’s scope at the time. But it went beyond that.

“Google was not as interested in the ideals it claimed to be interested in as much as political power play,” said Tarifi. “I’m very disappointed in Google. We’ve pioneered a lot of the early work in AI. Now, if you use Gemini…it’s completely incompetent.”

For its part, Integral AI is more focused on creating models that can predict the future states of the world under various contexts. A world model is “literally the lens by which we see the world,” said Tarifi. “We're all embedded within our world model.” In other words, what we see and experience is our entire reality. To generate a version of that via AI enables us to foresee different possible futures based on our freedom to choose different paths. This could definitely change the world, whether in regards to global climate change, international relations or something on a smaller scale.

For Jad, the pathway to that capability is not rigid. 

“I’m not dogmatic,” he said. Instead, he works through the lens of the “deep sense of humility that you get when you realize that you're not smart enough to know everything and the other person across the call from you might have a lot to teach you.”

Thanks,

The Daily Newsletter for Intellectually Curious Readers

Join over 4 million Americans who start their day with 1440 – your daily digest for unbiased, fact-centric news. From politics to sports, we cover it all by analyzing over 100 sources. Our concise, 5-minute read lands in your inbox each morning at no cost. Experience news without the noise; let 1440 help you make up your own mind. Sign up now and invite your friends and family to be part of the informed.

Join for free today!