What Does Generative AI Mean to Society; Learnings from Reid Hoffman and Eric Schmidt at The Milken Global Conference - May 2, 2023

I attended the Milken Global Conference on a mission to gain insights into generative AI from industry leaders.?I am interested in learning about this subset of AI that creates new content or ideas from existing data.?What are the positives and negatives of generative AI to society? What should we as society do to maximize the good outcomes and minimize the bad? I had the privilege of listening to a conversation between two tech titans, Reid Hoffman (Co-Founder of LinkedIn) and Eric Schmidt (Former CEO and Chairman of Google) to help answer my questions.

The Good / The Bad

Everything we do will be affected and amplified by generative AI.?Generative AI is a fundamental accelerator of positive benefits but, on the flip side, asymmetric bad actors (criminals, foreign governments, etc.) can use it for manipulative and even evil purposes.?We did not get it right with social media and AI is infinitely more powerful than social media.?A logical use of AI is to make even social media even stronger, and AI applied by bad actors can target you with such precision whereby the lesson we are taught from birth - that we can trust what we hear and see - is no longer going to be true.?We don’t want to cross the line into manipulation.

The problem is there is nothing stopping the underlying technology from diffusing out globally to those who want to damage democracy, conduct AI cyber-attacks, or use it for biological warfare.?Specifically, regarding its impact on democracy, there is the capability for political leaders to be vilified to the point where nobody will believe a word they are saying, suppressing trust and participation in democracy which needs to be addressed.

Currently there is no answer to the diffusion problem of the underlying technology powering AI. ?Questions have been raised if slowing down AI development is the answer. ?The answer is no, because bad actors are certainly not going to slow down development.?Reid Hoffman wants developers with good intentions to get there first.

How to Respond

Granted, AI offers huge potential but my concerns and additional questions are around the negatives, biases, and unintended consequences.?Tech has been left to regulate and put guardrails in place, and Eric Schmidt believes that the impact of job dislocation and inequality due to the impact of AI “should be addressed through the political processes.”?However, given the potential impact and scale, in my opinion the current parameters of the tech “Safety and Alignment” teams around maximizing good outcomes and minimizing negative outcomes does not go far enough.

There is a difference between freedom of speech vs. freedom of reach.?Eric Schmidt believes people can say whatever they want but that does not mean it needs to be amplified which is relevant to AI because AI algorithms will be particularly good at amplification.?One solution is for platforms to watermark data to know where it comes from.?Eric Schmidt believes we need to stop “random crap” from not knowing where it comes from.

In summary, we don’t want to repeat the same mistakes we made with social media by learning from our mistakes along the way.?We should take those learnings and not just the tech industry, but a larger group should proactively get in front of generative AI to address the negatives, biases, and unintended consequences.?“[AI] is so hard” and this is so important and impactful it should not be left to a few tech titans.

Source: Milken Institute