How to Raise Your Agent Episode 5 | Daniel Solove, Bernard Professor of Intellectual Property & Technology Law at George Washington University Law School

And one of the things that I keep hearing from our guests on this series is that AI and AI agents are really exacerbating existing problems within organizations and shining a light on many of those.

And when I was talking with Michelle about it, she was telling me, Alec, you've spent 25 years studying privacy and technology.

And she's heard you say that AI doesn't create entirely new problems so much as it remixes existing ones in complex ways.

I wanted to start off by asking you, what sort of remix of existing problems are you seeing?

Yeah, so I think that when AI came on the scene, there was everyone pronouncing like it's incredibly new and everything is different and we knew new laws and it's a whole new world.

And actually, it's an old world.

AI has been around For a very long time, the technologies behind AI are rebranded older technologies that are working a lot better because there's a lot more data and there's a lot more compute.

And they're working in ways that are essentially not opening up new problems as much as they are transforming old problems, making them worse by adding speed, adding scale, adding size, and various other things to these problems, and mixing different problems together in different ways.

So there's a remix, there's amplification.

There are a few new problems that are emerging.

But a lot of them I traced back going back a while.

And I think it was important to point this out in the article I wrote because legislatures and policymakers were rushing to create new AI laws, which I think I'm not necessarily opposed to.

But I think that what was getting lost in the mix was that we need to think about the existing laws and how to retrofit those laws to address AI.

In certain ways, the problems with privacy law that I had been complaining about for more than 20 years, actually more than 25 years, these problems, AI really demonstrate emphatically why the laws don't work.

and why they didn't work.

So in essence, I think instead of turning to a new law, why not look back at the existing laws, which are trying to address a lot of these problems and failing? And maybe we retrofit and fix those laws to better address the problems.

And so that was the general gist of my article, Artificial Intelligence and Privacy, that we shouldn't neglect the opportunity that AI is presenting us to rethink the existing law and see that it just doesn't get the job done and AI is really making it crystal clear.

And one of the things that I keep hearing from our guests on this series is that AI and AI agents are really exacerbating existing problems within organizations and shining a light on many of those.

And when I was talking with Michelle about it, she was telling me, Alec, you've spent 25 years studying privacy and technology.

And she's heard you say that AI doesn't create entirely new problems so much as it remixes existing ones in complex ways.

I wanted to start off by asking you, what sort of remix of existing problems are you seeing?

Yeah, so I think that when AI came on the scene, there was everyone pronouncing like it's incredibly new and everything is different and we knew new laws and it's a whole new world.

And actually, it's an old world.

AI has been around For a very long time, the technologies behind AI are rebranded older technologies that are working a lot better because there's a lot more data and there's a lot more compute.

And they're working in ways that are essentially not opening up new problems as much as they are transforming old problems, making them worse by adding speed, adding scale, adding size, and various other things to these problems, and mixing different problems together in different ways.

So there's a remix, there's amplification.

There are a few new problems that are emerging.

But a lot of them I traced back going back a while.

And I think it was important to point this out in the article I wrote because legislatures and policymakers were rushing to create new AI laws, which I think I'm not necessarily opposed to.

But I think that what was getting lost in the mix was that we need to think about the existing laws and how to retrofit those laws to address AI.

In certain ways, the problems with privacy law that I had been complaining about for more than 20 years, actually more than 25 years, these problems, AI really demonstrate emphatically why the laws don't work.

and why they didn't work.

So in essence, I think instead of turning to a new law, why not look back at the existing laws, which are trying to address a lot of these problems and failing? And maybe we retrofit and fix those laws to better address the problems.

And so that was the general gist of my article, Artificial Intelligence and Privacy, that we shouldn't neglect the opportunity that AI is presenting us to rethink the existing law and see that it just doesn't get the job done and AI is really making it crystal clear.