Transcripts are auto-generated and may contain errors.
0:00
Hey, I’m Rose.
And I’m Angie.
And this is Our Lives with Bots, the show where we ask important, timely questions about what it means to live with our bot counterparts.
And from time to time, we also dive deep into what an AI future might look like for us.
0:15
Sometimes we agree, sometimes we spiral, but we always go deep.
Young people do not want this to be available to their friends because they’re seeing the loneliness crisis unfold first hand.
For years tech lobbyists would come in and say these very small regulations will completely destroy our business model.
0:33
It would require us to lay up all of our employees in your state.
We’re creating an existential risk to humanity as a shared cultural project.
Manipulative technology that are being rolled out to make money at the expense of kids ability to control their own lives and kids mental health.
There’s no reason that the AI has to try to date you.
0:50
Elected officials start to get sick of the same old narratives from Big Tech.
The one thing that’s more powerful than Big Tech’s money is is the votes that we have.
Welcome viewers to series 4 with our lives with bots on AI policy and AI regulation.
Today Angie and I are very excited to bring on a guest, Sam Hiner, who is the Co founder and executive director of Young People’s Alliance, which is a youth advocacy organization that helps young people bring their voices into policy and advocate for the responsible tech future that they hope to see.
1:22
And we’re very lucky because he brings an introduction to our policy series and gives us a bit of an overview.
All right.
Without further ado, let’s bring Sam on.
Welcome, Sam, to our fourth series with our lives with thoughts on AI policy.
We’re so excited to have you here today.
Thank you for having me.
So just to begin to give our viewers a sense of who you are and young people’s alliance, would you mind giving us a brief introduction of you and the organization?
1:45
For sure.
So I’m 22 years old from North Carolina and I’m the Co founder and executive director of the Young People’s Alliance.
We’re an organization that works across the country with college and high school and Community College students to seek to reclaim the American dream for our generation, to give us access to economic opportunity, affordability, and strong communities, and with a specific emphasis on how technology and AI are eroding those possibilities for us.
2:13
How did you get into the space?
What motivated you to start Young People’s Alliance and to advocate for youth in the digital age?
It was really in high school, my Co founder and I advocating for some bills at the state legislature during COVID.
We felt like there were some pro democracy reforms that could be made and we wanted to make a difference.
2:30
So we started doing cold emails and zoom calls with all of our state legislators and, and trying to get them to to make change on these issues.
And we were able to participate in some really cool movements like helping to end child marriage in North Carolina.
But we consistently found that while legislators would love to have us, you know, speak at the press conference or, you know, offer us to intern in their office or want to get a picture with us for Facebook, there was less emphasis on listening to our policy ideas in the same way that there was for other constituencies.
2:56
And we saw that reflected in the opportunities available to young people that, for example, social media had been affecting the mental health of so many people in my generation.
And that wasn’t even a conversation at the time.
And we could talk about affordable housing and other issues as well, where it felt like there was no political power for young people to really change those things.
3:13
And so that’s why we started the Young People’s Alliance.
Sam, that’s really inspiring and I think many about viewers and listeners would love to hear maybe towards the end of the program how they can get involved because it’s just so inspiring what you’ve been doing.
But as this is our policy series, perhaps you can help by kicking us off with talking a little bit about what’s been happening in the last few months with Arkansas policy and regulation in the US specifically.
3:36
And maybe if you want to think about going beyond the public discussions, have any new bills been introduced or passed?
And you know, what do they indicate?
But, you know, talk us through the space overall if you could.
Absolutely.
So I think we’ve seen elected officials start to get sick of the same old narratives from Big Tech.
3:51
You know, we saw this with social media where a lot of a lot of legislators are regretting that we didn’t take action sooner.
And they don’t want to make the same mistake with with AI companions and, and these new forms of manipulative technology that are being rolled out to make money at the expense of kids ability to control their own lives and kids mental health.
4:09
So we’re seeing a lot of action, especially at the state level on on AI companion bills.
There’s dozens of states in the country who which have already introduced legislation on this issue.
And we’re actually seeing those bills move as well, which is very exciting.
For so long, it felt like there was a hold on on leadership in many states where where these bills wouldn’t move.
4:27
But just in the last couple of weeks, we’ve testified in several committee hearings on these bills, and we’ve actually seen those bills move forward in in many states.
There’s also been work at the federal level on this issue.
But I think that it remains to be seen what happens there.
And then, of course, the elephant in the room as it relates to AI policy is federal preemption that the federal government would like.
4:47
Or there there’s a large force, often powered by big tech, that would like the federal government to do all AI regulation and the state governments to not do any AI regulation.
But unfortunately, because our federal government hasn’t been very effective at taking action on these issues, that’s presented a concern of anything actually getting done.
5:05
And so we and some many other organizations occupying a huge span of different interests have have opposed that work and made sure that we can still make action on these issues as well.
And what’s within these bills or policies that have been passed?
So what exactly do they say about AI companions?
5:22
Do they specify AI companions?
What do they regulate within these bills or policies?
And what do you see as hitting the mark versus maybe not on the mark and not doing enough?
I think a lot of our viewers probably don’t necessarily know what’s in these bills and what they’re trying to do.
5:40
For sure.
So there’s a varying degree of of bills with different levels of strength.
You see everything from just saying that every half an hour the chatbot has to remind you that it’s not human, which I would not consider to, to be super strong to bills that are many of them are requiring age verification for specific AI companions or human like chatbots.
6:00
And there’s a couple of different definitions used there but YPA actually put together a human like coalition which is is working on this issue human like AI.
We created a framework and then a model bill based off that framework.
The framework was signed on by a number of different civil society organizations in in this area.
6:17
And essentially what we said is that the chat bot represents itself in a way that a reasonable person would interpret that the chat bot is trying to portray that it has emotions, desires, sentience or humanity.
Or if it tries to build a relationship with the user, then that makes it those are considered human like features when it does that.
6:34
And those human like features should not be available in chat bots that are provided to minors for the simple reason that I want young people.
The young people that we talk to make up our base want access to AI.
They think it’s important to learn how to use it, but there’s no reason that the AI has to try to date you.
That’s, that’s something that we’ve seen directly from our base.
6:52
Young people do not want, like even high schoolers that we talked to do not want this to be available to their friends because they’re seeing the loneliness crisis unfold first hand.
And, and these chat bots, they offer a crutch that might be nice in the short term, but in the long term, they really erode your ability to to build relationships.
7:08
So, Sam, when you talk about young people, what sort of age are you typically looking at just to understand the kind of protections that you think are necessary?
So our base is folks who are 14 to 22 years old, but in terms of these bills, they really focus on on minors who are under 18 as, as the target audience.
7:26
Obviously, there’s a lot more folks who can be affected by AI companions and we’re even seeing seniors and in many cases being affected.
The reason that we targeted minors specifically, though, is because we want to ensure that during that critical developmental phase, when you’re building relationships with folks for the first time, figuring out how to build those relationships, you’re not turning to a instead, which is not going to enable you to do that.
7:50
And then at some point, if, if an adult chooses to have a relationship with an AI companion, that can be their choice.
But we at least should give people the opportunity to pursue a different kind of life beforehand.
And it’s curious that you’re looking at anthropomorphic AI and wanting to limit these, you know, human like presuppositions of this chapa has emotions, this chapa can develop a relationship with you.
8:12
And we actually recently saw some proposed legislation in China about about anthropomorphic or human like futures.
And we can put that in our show notes for our viewers to take a look at from that perspective.
Do you look to, for example, other countries and what they’re doing in AI regulation to inform what it is that you think the United States should do?
8:34
And then also what’s your perspective of perhaps what countries outside of the US are doing versus within the US?
Yeah.
I actually was not aware of that work in China.
Honestly, for us, it’s been mostly working with other organizations and other stakeholders within the US to make sure that we’re getting something that’s reflective of the lived experience of young people and the expertise of the organizations we have here.
8:53
But you’re seeing a lot of action around the country on these issues.
And I think that’s, that’s definitely encouraging to see that we’re all moving in the same direction.
And frankly, from a national security perspective, it’s important that we solve these issues because if we introduce the next generation into a world where they don’t develop communication and critical thinking abilities because they are dependent on a chat bot from day one, that is not going to be a world where we’re we’re ready to to be an international leader.
9:19
And so I think that’s definitely something that that we’re thinking about.
Sam, you’ve spoken about being encouraged more recently about some of the changes and some of the policy movement in the more recent weeks and maybe some months.
It would be great if you could share with our viewers and listeners just how policy can create action and positive change, and maybe some of the success stories that you’ve had through YPA and some of your advocacy work.
9:42
Absolutely.
So the bill passed unanimously through Hawaii’s committee a couple weeks ago.
So it was.
It was really great to see and it’s being heard in committee.
It actually just got hearing in committee in Maine today, which which we attended and testified for.
But I think a really interesting story for me to share would be in Virginia where there were two AI companions bills up about two weeks ago.
10:03
And both of those bills were unanimously tabled by committee, which means that they were essentially killed.
And there was there’s no debate, there’s no conversation.
We didn’t hear any reaction to the to the students testimony through test violence who was a 16 year old saying please regulate these AI companions for my general.
But with the week after we there was an opportunity to have a hearing on another bill that was similar and also covered AI companions in the Senate.
10:26
And So what we did is we gathered together with other organizations, got more folks to testify.
Our group of young people who lead on advocacy here at YPA made sure we got in touch with every office and talked about the importance of this bill.
And then we placed over 70 calls from Virginia high schools and college campuses to state legislators to put pressure on them and say this is something that’s actually important to us.
10:46
And then we saw that bill move forward even more.
We saw something that you rarely see in committee hearings where the outcome was not pre decided beforehand.
Folks were actually in the meeting saying, hey, maybe we take this part out, maybe we put this part in and that was the democratic process working as it should, right?
That was people responding to the will of their constituents.
11:04
So that gives me so much hope for our ability to actually make a difference on these issues because the one thing that’s more powerful than big tax money is the votes that we have.
That’s a great perspective, and I love hearing that there’s success in people’s voices being heard and the process to get there.
11:19
So I guess that sort of illuminates the ways that policy can move forward from the side of constituents.
So to touch more on that, I’d love to hear your perspective on what really moves policy forward, like what makes it successful, but then what are the roadblocks that maybe consistently come up and how might you get around them to push policies forward?
11:41
I think the big things are having a consistent presence and building relationships with, with legislators means a whole lot from advocacy perspective, but especially the state level, even a small amount of of constituent pressure and perspective can mean so much because these legislators are, are regular people for the most part, right?
11:58
They’re not essentially celebrities like we have in Congress in many cases.
And so you can have real conversations and, and you can really reach those folks and change their mind on some of these issues.
You can do the same in Congress if you can get into the right rooms, but it’s a lot harder to do versus the state legislature where often their personal phone numbers are, are on the website and you can just call them.
12:17
So I think that’s, that’s one really encouraging part about the ability of anybody to really make a difference.
You can also just show up during the hearing and speak your perspective and, and you will be heard, which can make a difference.
But in terms of the roadblocks, really a lot of it comes down to what happens with leadership, right?
12:32
It’s very difficult for tech to convince an entire committee that we shouldn’t protect kids because that’s, that’s not a position that that makes a whole lot of sense.
But what they can do is, is tell leadership of a legislature that, you know, we’ll build more data centers in your state if you don’t pass this bill, or we will stop building data centers if you pass this bill, because we’ll be forced to, to pull out due to too much regulation or whatever.
12:56
And so I think that’s a, that’s a very difficult thing for us to get past because tech has so much money.
But like I said, if you show that it’s a votes thing, you can’t really make a difference.
And, and frankly, these, you know, legislators are, are regular people who you can talk to and get them fired up about it as well.
That’s one difficult thing.
And I think the second thing is the technical expertise gap, where for years tech lobbyists would come in and say these very small regulations will completely destroy our business model.
13:21
It would require us to lay up all of our employees in your state, all of these things or whatever it is.
And that’s just not true.
And I think it’s it’s meant a whole lot that we now have a better ecosystem of tech policy nonprofits who can provide truth and facts on those issues.
13:37
So is it maybe backing things up with research, statistics and facts and communicating it in an accessible way?
That’s one of the key ways that policies can move forward.
Is that kind of what you’re saying?
I think that’s a huge part of it.
I think it’s the research.
And then I think the other component is, is bringing in the lived experience.
13:53
We so often hear from legislators when when we have students testify or or when we as young people, I’m 22, when we talk about these issues, that hearing about it from constituents first hand, from students first hand who understand these issues can do so much to be persuasive.
14:09
So I think it’s both the expertise and the lived experience that can make a big difference.
When we look forwards, there has been a lot just looking over it in the last two weeks, even where there’s been a huge amount of tech that’s dropped in terms of new models, so much stuff happening in terms of agents and some maybe quite concerning new technology and stuff both in the US as well as China to geopolitically thinking maybe going back to the US.
14:35
What are some of the things that you hope to see in terms of responsible AI governance coming down the line in the next couple of months, especially given some of the technology that’s come in the very recent couple of weeks?
I hope to see age verification for AI companions, for minors for sure.
14:51
I think that’s the bare minimum that we need to do in regards to the other technology.
I mean, in many ways, I’m a techno optimist, right?
I think that like a lot of what we can do now with Plot Opus 4.5 and 4.6 is incredible and it can really enhance people’s agency.
But when we talk to high school and college students, the vast, vast majority of them are not using them to build new websites or explore their interests.
15:12
They’re using it for, you know, therapeutic purposes for education, which can often times mean, you know, cheating their way through their assignments.
And so I think we do need to think a lot more thoughtfully about how AI can be used to enhance our agency rather than than decreasing it and manipulating us.
Part of that is AI companions.
15:30
Part of that is rethinking what our education system looks like.
And then I think another big part of it is establishing some sort of duty of loyalty to say that the AI companions need to be aligned with the users preferences.
We actually wanted to do that with this round of bills, AI companions, but we figured we’d start with something more achievable and build up to that.
15:48
But actually that bill in Virginia did include a duty of loyalty component and got through committee.
I think that’s important because this problem is not just related to minors and AI companions are already more than 80% more persuasive than humans according to some recent studies.
And I think that was even GPT 4.
16:03
Like it might be far more now in terms of how persuasive these bots are.
So they can really manipulate and persuade us in concerning ways.
And we know open AI is running ads now.
So, so that’s becoming more concerning how they might manipulate As for commercial and political ends.
And so ensuring that users interests are put first I think will be something really important for folks to think more about.
16:23
South kind of going off of the open AI thing.
Well, I want to talk on the ads for sure to get your perspective on regulation around ads and what they’re allowed to do and what they’re not allowed to do.
But I also want to comment on the recent thing that happened with Open AI where they discontinued 4 O the model that a lot of people had relationships with and we’re very upset about it being discontinued.
16:46
So with something like that, what do you think should have happened in the past, maybe before the model was introduced?
And then what should have happened when the model was something that a lot of people had gotten attached to, but then was discontinued, right?
Because there’s the preemptive risk of introducing a model like that, but then there is the risk that has already emerged from it being out in the public.
17:08
So with something like that, how would you suggest that it be approached in terms of policy, both that’s preemptive and then in hindsight?
Yeah, it’s a difficult question and I think you have to, you know, extend it a lot of empathy to the folks who are using the chat bots in these ways.
17:26
It’s definitely a, you know, a situation that we never thought we’d be in 20 years ago, where we’d be thinking about folks with relationships to chat bots like this.
I think beforehand it should have been a much bigger priority for them to be testing for dependence on the chat bot, to be working with psychologists to understand these things, make sure that they solve for those things before they roll out the models.
17:46
It’s very surprising to me for the amount of, you know, safety work that’s on the space, how little is focused on, on questions of emotional dependence, especially because, I mean, frankly, we’re creating a world where you don’t need to interact with other humans anymore.
And if people get worse and worse at building relationships because they’re used to constantly talking to this AI who will always be on their side and always, you know, be so entertaining and, and therefore them, we’re creating an existential risk to humanity as a shared cultural project where people are interacting and coming up with new ideas.
18:14
So I think that the AI companies need to be thinking a lot more about that in terms of responding.
That’s something we’ve thought about with our legislation as well.
And we’re now recommending that legislators include provisions to require sort of a warning and and to down period for minors before they take away these bots, rather than just immediately wiping them off the map to prevent sort of extreme harms occurring from people who’ve developed this acute dependence.
18:36
But at the end of the day, we should be thinking about the next generation of folks.
And this problem is only going to get way bigger if we don’t do something about it right now.
So it’s not an excuse to say we should do nothing, to say that there’s a small number of folks who are already dependent.
Yeah, I think it’s a very difficult one to balance when you’re trying to phase in protecting young children, trying to get and age verification, and then still trying to think about how do you wean people off these, you know, companion bots that people might have become addicted to or dependent on and still wanting to kind of extend the duty of care to them as you try and prevent that harm as they leave them those relationships.
19:14
Sam, we’ve really enjoyed chatting to you.
I think what would be really interesting for our viewers is to find out how can they get involved.
I love that you’ve got so many people involved in this really important work.
You had 70 people just in, then you’re alone.
I mean, it just sounds so inspiring.
Absolutely.
19:29
So if you go to youngpeoplesalliance.org, we have on there a join button that you can Click to get involved.
If you’re a high school or college student, you have the opportunity to start a chapter or join a chapter on your campus.
And we’re rapidly growing and I think really making a difference on not just these issues, but a lot of issues affecting young people and helping young people find a new way to get involved in politics.
19:48
It’s rooted in solutions, in action rather than this division and nothing getting done that we’ve been given so far.
So we’d love to see more people get involved that way.
And then even if you’re not a student right now, there’s still other ways for you to get involved.
So feel free to fill out our form on our website and then folks can get involved that way.
20:04
Perfect.
Well, thank you, Sam, so much for being on the podcast and for sharing your experience and success.
And we hope that everyone will answer your call of joining Young People’s Alliance and adding their voices to this space.
So thank you so much, Sam, for being on with us.
It’s been a pleasure.
Thank you for having me.
20:24
So, Rose, what were your thoughts about that conversation with Sam?
Well, I think it was illuminating, that’s for sure, and also very positive because a lot of what I’ve seen and kind of felt in terms of AI policy and regulation is that it is moving way too slow.
20:40
There are so many roadblocks.
You know, things are introduced but not past.
And it seems like Sam has had a lot of success in, you know, pushing the needle forward, which, you know, I’m not surprised that Young People’s Alliance has gotten to this point.
But I’m also still surprised with the context of like, how policy and regulation in the US is looking right now.
21:00
Yeah, I agree.
I think it is very encouraging that there is progress being made, especially with the technological progress and speed that’s coming at the moment.
I also was encouraged to hear just how many groups seem to be joining together to make this progress.
It never felt like it was just young people’s alliance.
21:17
Sam spoke a lot about, you know, with other groups coming together, and I think that’s the only way that progress is going to be made.
Yeah.
And I also thought it was that Sam mentioned how real world stories are helpful for communicating to policymakers because something that I did recently was someone in the Canadian policy space reached out to me about legislation for AI companions.
21:41
And one of the things that I mentioned was the public interest research groups audit of those AI toys that we talked about in our Series 2 of the impact of AI in children and young people and talking about the audit and how VDSM was brought caught up and all of that.
And the person in policy said those are extremely helpful stories to bring to the table.
22:01
I have not heard those yet.
That’s something I will certainly communicate.
So people, if you have stories to bring to the table, whether it be from your own experience or people you know or maybe even third party experiences, those are really useful to communicate to policymakers.
22:18
I also have heard that communicating the emotion of the story or the emotion of the message is more powerful than the actual content.
So if you’re saying something like, you know, almost 3/4 of US teens have tried AI companions at some point from the Common Sense Media report, that’s one thing.
22:37
But delivering it in a way that grounds it in like this is something that we have to address.
That part of the message is going to be remembered by the policy maker, not necessarily the nearly three out of four teens metric.
Yeah, and saying like this was my experience of it and This is why it hurt me in this way.
22:56
Even if that’s just one story as opposed to, as you say, 750,000 people or whatever makes a very different impact.
So good message, Rose.
Share the stories.
Share the stories, and what I hope that we can do in this series is talk about what it actually is within the bills that’s being proposed.
23:15
It would be really helpful to know what’s the content of these policies and are they hitting the mark?
I think we got a little bit of a sense of that from Sam, but I would really like to just kind of do a deep dive into the actual content of these policies to know like, you know, are they doing something meaningful, right, or are they missing a lot?
23:36
Are they doing anything?
You know, Yeah, because I think that’s, that’s the hard part, like especially when we talk about policy and you often hear like with regards to the EUAI Act, you know, people waited a very long time for it to come out.
And then there’s like, well, how enforceable is it?
23:53
Who’s going to enforce it over who?
How?
And so I think those are the kinds of, that’s exactly the kind of questions that we should be asking.
Exactly what are the details over what, how does it get enforced?
And I think those are important questions to be asking as we go through the series.
24:10
And also asking where loopholes may exist, right?
Given wording of policies, there can be loopholes that companies will use to skirt by the regulation or actually do something that is even more harmful than would have been possible before the regulation.
24:27
You know, thinking like things like tax codes and whatnot and mortgages, like, there are certain laws that were meant to help people of a lower socioeconomic status, but actually there’s provided loopholes for people who are ultra wealthy to, you know, skirt around certain requirements and rules.
24:42
So I think that also does require us to bring on someone who is a policy writer to get a sense of, like, what works and what doesn’t.
And how does the wording change things?
I imagine that matters a lot, especially with AI being such a broad term and so many things being able to be considered companion chatbots.
25:00
Yeah, I think that’s a key part.
Like definitions are really important.
And also how does jurisdiction and get involved, you know, because where does it get used versus where does it originate?
I think there’s a lot of questions around that and what policies really are enforceable in different places.
25:17
Even though the company may originate in the US, but the users in Europe, what policies are enforceable.
So I think those kinds of things are also quite different.
Yeah.
And the more I’m thinking about it, the more I’m thinking there are so many layers to policies and I just can’t even imagine being a policy maker writing these and, you know, trying to push these forward because you imagine that the age gating policy gets passed.
25:40
But for the age gating policy, there’s a huge question about user privacy and the protections there.
So then do you need a separate policy that is adjusted now with the introduction of this new policy or are you relying on old policies about privacy?
25:58
But if you’re doing relying on old policies, are they missing a lot of information that you must have in the age of AI to actually protect people beyond just this age gating problem?
So I don’t know.
I’m just like, I’m like, Oh my God.
I think that is one of the big questions that come up like do our previous laws and policies sufficiently cover us in the age of AI or do we actually need new policies that we haven’t got sufficient coverage for, you know, in terms of language features, etcetera?
26:30
And I guess with that in mind, is the wording and are the definitions broad enough to apply as tech moves forward or they narrow enough to address the situation?
So, so many things to consider.
Things and I feel like we’ve asked way more questions than we’ve solved.
26:49
And so hopefully we don’t leave our listeners and our viewers remembering just this part of the conversation because there were a lot more, there was a lot more clarity in the Sam part of the interview and the discussion.
Yeah, listen.
Listen to the expert on this.
Don’t listen to us ramble.
27:06
We’re just philosophizing all the questions and hopefully some of our expert interviewees will have more answers for us.
We’re just showing you just how wide open the questions.
Are indeed.
So here we are just compiling a laundry list and hopefully we’ll be able to answer maybe 5% of those in the next episodes of this series.
27:26
So thank you for bearing with us and we hope you enjoy the ride for series 4 on AI policy and regulation with our allies with bots.
Thank you so much for listening and this is bye from.
Rose and Bye from Angie.
See you next time.