On this episode, host Tim Storey dives into the world of artificial intelligence with Alex Wirth, the co-founder and CEO of Quorum, a Washington, D.C.-based public affairs tech platform that organizations use to manage engagement, launch advocacy campaigns and keep track of legislation and policy trends in Washington, all 50 states, thousands of cities and even in Brussels. Storey talked with Wirth about how the platform evolved and how his tech team has increasing used artificial intelligence tools to perform sophisticated analysis. Wirth broke down how he thought AI would affect state government, legislatures and individual legislators. He also offered some advice on managing the use of AI in legislatures, and explained why legislative offices may soon need AI tools to handle all the AI-generated communications they will be receiving.
On this episode, host Tim Storey dives into the world of artificial intelligence with Alex Wirth, the co-founder and CEO of Quorum, a Washington, D.C.-based public affairs tech platform that organizations use to manage engagement, launch advocacy campaigns and keep track of legislation and policy trends in Washington, all 50 states, thousands of cities and even in Brussels.
Storey talked with Wirth about how the platform evolved and how his tech team has increasing used artificial intelligence tools to perform sophisticated analysis. Wirth broke down how he thought AI would affect state government, legislatures and individual legislators. He also offered some advice on managing the use of AI in legislatures, and explained why legislative offices may soon need AI tools to handle all the AI-generated communications they will be receiving.
Resources
TS: This is “Legislatures: The Inside Storey.” Thank you for listening. I am the host Tim Storey, the CEO of The National Conference of State Legislatures, NCSL. My guest is Alex Wirth, the co-founder and CEO of Quorum, a Washington, D.C.-based public affairs software platform that organizations use to manage engagement, launch advocacy campaigns and keep track of legislation and policy trends in Washington, all 50 states, thousands of cities and even the EU and Brussels. I talked with Alex about how the platform evolved and then got into the topic I really wanted to pursue: artificial intelligence, Gen AI, and he had a lot to say. Alex broke down how he thought AI would affect state governments, legislatures and individual legislators. He offered some advice on managing the use of AI in legislatures and explained why legislative offices may soon need AI tools to handle all the AI generated communications that will be coming at them.
Alex also talked about his family roots in the political world and legislatures. He talked about his father Peter Wirth who is the majority leader in the New Mexico Senate.
Hey Alex Wirth, I am just so delighted that you are making time to join me on the podcast today. Thank you for being here.
AW: Awesome. Well thank you for having me Tim.
TS: And as people heard in the brief little intro, you are the founder of Quorum and the CEO of Quorum. So, before we jump into some other stuff, tell us about that. When was Quorum founded? I know a fair amount about it, but I know I’m going to learn a lot more. It has grown tremendously almost exponentially it seems in the last few years so tell us about Quorum.
AW: Happy to. So, I started the company with my co-founder and college roommate in 2014. And we launched it in 2015 focused on bringing a modern public affairs technology platform to help professionals work smarter and move faster. What that means in practice is we focus on legislative tracking across what’s happening in the U.S. Congress and of course all 50 state legislatures and also now at the local and school board level as well. Also have a series of grassroots database tools to help people get in touch with their elected officials and share with them their opinions and what’s on their minds. And then we have a PAC management business as well for organizations that are participating in a political process through campaign contributions. And our focus is bringing all those platforms and workflows together into one modern platform to help people be able to more effectively engage in the political and policy process.
TS: Like I said I know something about this, but I’ve always been curious like you know how much of your work on the state and local side versus the federal side in terms of your clients and what they are focused on.
AW: So, it’s about 50/50 now. I think when we started in 2015 there was a lot probably a little more focus on Washington because people saw it as a very effective pathway to be able to engage and make a lot of things happen. And as the U.S. Congress has gotten more and more gridlocked, we have noticed a very significant shift towards focusing on the states, to the laboratories of democracy, because people realized that both one in many cases states are leading the way on policy issues that then start to affect what the U.S. Congress is considering. And also, there were many more opportunities in a much quicker pace of policymaking to be able to get what they were interested in enacted and passed. And so, we now see a tremendous amount of focus and time at it’s been at the state level and it’s probably equivalent to the focus that both we and our clients end of having with the U.S. Congress.
TS: You echoed talking points that I’ve used many times on this podcast talking about you know this is where the action is, where the innovation is and it’s widespread and it’s thoughtful. You know there’s a lot of thoughtful policy coming from the state. I was just looking this morning at the bills introduced in the past. So far, we are sitting here just at the beginning of June. June is around the corner. This will air in June. It’s been a very active time for legislatures and you are doing a lot of this work so that’s good for you honestly because it means you know a lot of good business and you are where the action is. Where are you from? Tell the listeners where you are from and how you wound up meeting this person and getting into this.
AW: Yeah so, I am from Santa Fe, New Mexico, born and raised. And the fun background is that my dad is a state legislator. Was elected when I was in the 5th grade, in New Mexico statehouse representative at the time and now the state Senate. So, I grew up on the sidelines of the political process where the conversation always at the dinner table was what was going on. What was the policy issue of the day. What happened at the committee hearing or what happened on the study committee. So, I very much grew up in that political eco system. I had the opportunity to do my undergrad at Harvard and met my college roommate at the time. And how we came up with the idea for Quorum is that I was doing some advocacy work on Capitol Hill. We were trying to get a presidential youth council of young Americans to advise the president. We had a fairly vanilla resolution and we were looking for members of Congress to sign onto it. And we ran head-on into the partisanship that is in Washington now of trying to figure out well what Republicans and what Democrats would work together most frequently. I did it the hard way going to literally 60 offices on Capitol Hill and saying would you sign onto this resolution. I got like 10 yesses and 50 nos. My college roommate Jonathan, who is my co-founder at Quorum CTO, studied computational biochemistry and the analytics between protein networks. And what we figured out that if you take the same code that is used to analyze protein networks and apply it to congressional networks, we could map out for every member of Congress who they worked with most frequently.
So that was the spark and the starting place for us and we ended up building the entire Quorum platform from that. And before you ask, we can also do the same thing at the state legislative level where we have the voting records and many cases although not in all states, we also have co-sponsorship information. You can use that and use computer algorithms to map out where the relationships are. Who works with who. Where are the opportunities to work across the aisle. Where are those legislative connections and provide a whole new way of being able to look at how do you engage in a legislative and political system.
(TM): 6:24
TS: That leads me to something we want to talk about today which is GenAI, Generative Artificial Intelligence, and AI written at large. What you did was in some ways you took what happens in the process in the 50 capitols and the territories and which is usually there is a person right. Awe, you got to go talk to, you know, Sally. Sally has been here forever. She can tell you how these networks work. And you were able to accelerate that in a very dramatic way, not to say that Sally’s info isn’t still vital to understanding what makes these very unique ecosystems. I think of the capitol as the really unusual ecosystem and that where humans interact and very different, but also very similar ways to institutions and organizations where people interact. How is it working out?
AW: It’s been great as we’ve been able to really dramatically scale the business and so we now have a couple thousand customers, 350 team members and are working across all 50 states U.S. Congress. We are also doing a lot more work internationally now with an office in Brussels and expanding our international legislative tracking that’s happening. But I think one of the big reasons why we are talking is that we have seen the most significant advance in technology with you know really the advancements that have been made with artificial intelligence and neuro networks. In the 10 years that we’ve been running in the business and honestly, I think it’s the largest technological advancement that we you know as humans have seen or really as kind of modern society has seen is the launch of the internet. Part of I think the key message that I would have today is that we all as a society need to be approaching this with the same level of thoughtfulness and intentionality as if the internet just launched and it’s going to totally fundamentally change both how we interact as humans and also have really significant impacts on the policymaking process on state governments and on the engagements that happen between citizens and their government. And I think those are a lot of interesting areas to chat about from there.
TS: So, let’s get our yellow highlighter, our metaphorical yellow highlighter, out and highlight what you just said. You think this is the biggest technological development even in our world of legislature world of policymaking world since the advent of the internet. Is that that’s what I heard you say.
AW: That is spot on.
TS: We will dive into this a fair amount here so because these are the words of like ah that’s what they say every time something comes along you know it’s always this is the thing that’s going to change it. They probably said that I’m sure there were skeptics of the internet. Aw mosaic or what was the other one Netscape. Remember when Netscape came along. It’s sort of telling you when I was around at my age and experience with this. But you don’t seem to have a ton of skepticism, not to say that you don’t appreciate that there is going to be twists, turns, bumps in this road. You know what makes you think this?
AW: It is the ability to access information in an entirely new way with much greater speed and much greater intelligence than ever possible before. And I say this as someone who runs a technology company that has tried to do it the old way of how do you find like-minded bills that are similar in one way to another series of bills. Or how do you figure out which state legislator works with another most frequently. Or how do you figure out if someone’s press release or tweet is positive or negative. What we’ve seen historically is that there are some ways to figure that out. We have figured out a number of them often through very significant amounts of very painstakingly written code. And there are other things like sentiment analysis that is just really, really hard or even mapping what bills are similar to get right. And what the advancements are with this significant enhancement in AI and neural networks is it makes a lot of those really difficult problems much, much easier and also makes it accessible so that anyone that has access to the functionality can go and be able to both get some of those insights, but also apply those insights onto the datasets and information that they have access to.
And so, I think that that’s one of it. I think the other piece of it is the ability to create texts and snippets and information with a couple of sentences of a prompt and as we think about the policymaking process whether that be a tweet, a press release, a bill, a research report that there is a world that you know basically generative AI is used much more frequently in those items. And so therefore is going to have a massive impact on the policy process because suddenly your chief of state or legislative assistant can be doing things that it used to take 10 different engineers to go try and do and that’s going to really have an impact on the day-to-day work across you know all 50 states and beyond. And that’s also why I think you are hearing about it and reading so much about it in the press because it’s not just state governments that are impacted by this. It’s literally everyone that is impacted by this in a really significant way.
(TM): 11:31
TS: Quorum turns 10 next year that’s exciting and really tremendous growth obviously. You really found a special place and leveraged your expertise in a pretty neat way. We all are kind of familiar with this mile marker of November of 2022 when ChatGPT and open AI releases that lively to everyone. But obviously AI has been around a long time. You know particularly I know the military is very advanced in some of their AI applications. I’m not sure other extremely high-tech operations and midways things that we use all the time are already echoes of AI, but when did it seem different and serious to you. I mean when did you would you say it was a year or 2 ago and why is this moment different than it has been for the last you know five or 10 years.
AW: The direct answer to the question I think it all changed in November with both the launch of ChatGPT and also obviously with Google very soon after releasing Bard and a series of other advancements. You are right. AI has been around for a while and there are multiple different types of AI. I think one of the more easy ways of explaining it is one of the types of AI a machine learning which is essentially using a series of computers to try and learn off a specific dataset and produce meaningful and insights. And so, to give you an example of that, something that we do is we take all 120,000 bills introduced every year across every state legislature and we use machine learning to go in and tag those bills relative to a specific issue area. The way that we do that behind the scenes is that we have written a fairly sophisticated piece of code used to go and take the bill text and compare it to bill text that was actually introduced in the U.S. Congress and tagged to an issue area because the congressional research service goes through and hand tags every single bill through a series of issue areas. And so, we essentially had a training dataset of here is a bill and here is what issue areas are related. So, we could then write code that says hey this bill looks really similar in terms of the words that are used. Education, teachers and school is mentioned a number of times. And the U.S. Congress categorized that as an education bill so this is likely an education bill. And so that’s a piece where we’ve been using AI for years and it has been present in our lives and there is obviously a number of things that we as humans interact with that do have AI affiliated with them. Something as simple as what advertisement you might see on Instagram or on Twitter.
The challenge that we have had is being able to find what is a bill that is most similar to another bill and introduced in one state that might be similar in a different state. And the reason for that is that every state writes their bills a little bit differently and the words end up being totally different. And so, this is the promise of advancements of modern AI is that it is essentially multiple of those machine learning networks on top of each other to create what the computer scientists are calling a neuro network and gives a much stronger, more capability to say hey this bill was introduced in Montana. Can you pull me up similar bills in 50 states all across the country and actually have those be much more accurate and much more precise in saying these are the bills that are similar because what neuro networks are doing is they are essentially looking at all of the language in each of the bills introduced and saying well what are the ones that are most similar in comparison. And it ends of being very difficult to write kind of one machine learning algorithm to be able to do it. And so, the magic of AI is having multiple of those all working together across a multitude of different subjects and that’s why it ends up being such a major significant enhancement. And so, when we look at it, we are able to do things that you know three, four or five years ago we would of dreamed of doing but not had either the computational power nor the machine learning or computer resources to really be able to do. And suddenly those things are now much more accessible and much more accurate and that is the potential that is coming you know both to how individuals interact with government, but also to industries across the spectrum.
TS: So, this is a great example of how it is impacting your particular part of this grand legislative world. Let’s expand it out a little bit because I know you’ve given some thought to the bigger picture of what does this mean for governing. I mean for governing states free at large and that’s a huge question that we could spend you know half a day talking about but and let’s admit it we’re at the beginning of this thing right. I mean we’re not even at the end of the beginning. We are at the beginning and I don’t know how the beginning how long that part is going to last. In fact, we could save this recording and listen to it in a year and we’ll probably sound silly. So where does this go kind of top level you know there is no way to get super deep, but with just governing. You know a state governed maybe the better rim of two.
AW: Yeah so, I actually think you have to look at it in two different places. One, what it means for any of our legislators who are listening for their process of governing. And the story I always tell is with a group of communications professionals and we were talking about AI in a round table session. And one of them mentioned how one of their team members who was a little bit of an underperforming not as strong came back with a draft press release one day that was unbelievable. And everyone kind of scratched their heads and said look at the amazing work that so and so team member did. And everyone read it and said yeah this is really good. And so, the head of the team went over and asked him and said hey what did you do differently? Have you been like doing some coaching like what happened. This is phenomenal. And the answer from the team member was oh well I used ChatGPT and the team went oh wow. Like both very innovative, resourceful, but also then like not sure how to feel about it. And I share that story because for many of the legislators that are listening to this podcast or for anyone else who has teams that are working for them, there is a very good chance that your teams are actively using, experimenting and exploring ChatGPT and that is an open question that you may or may not know about it. And now I share that because I would highly encourage you to have that conversation with your staff at the organizations that support you and start to set some parameters about how you want it to be used. And the experimentation is a very good thing, but I would encourage at minimum that they disclose to you when they are using it. Because for example, one of the challenges with it is that it is not 100% accurate. So, imagine if you have a team member or a staff member go and say hey summarize this bill using ChatGPT, something that ChatGPT or any one of the other neuro networks should be very strong at. They might come back with someone something that is 80% accurate, but what if they miss the 20% thing that you know is super important. And this is one of the examples of the limitations of artificial intelligence. That’s why it is still not going to come replace a policymaking process or replace a lot of the day to day of the importance of human interaction review and analysis on a piece of legislation.
But it actually could have a fairly significant impact. Or two, what if it is using wrong information and so you might ask and say well hey, I need some evidence and information to back this up. Your staff member might go put that into either Bard or ChatGPT and the information could come back wrong. I think many of us have read the story about the lawyer that used it in the airline case and ended up citing cases that weren’t actually accurate or truthful and that didn’t exist and so had to throw the whole case thrown out. So, you have pretty significant factual challenges and then you also have intellectual property challenges. What if it uses information that is not cited or is actually infringing on someone’s intellectual property. And so that’s where there are a lot of items that haven’t been figured out. And what we are seeing folks do at an organizational level is put in place guardrails and restrictions and guidance about when or when not it can be used. And as lawmakers or state policymakers and staff if you either haven’t done that for your team, I would encourage you to start talking about that because that’s going to be one of the first ways that we see it used. And I think there’s a second question about okay well what does it look like for states using it. But why don’t I pause there and see if there are any thoughts that come from that.
TS: I have a number of thoughts about that. One is that I was thinking about NCSL and how we’re this trusted source. One thing that makes us you know so integral to the legislative process in many ways that you know when staff come and get information from us at this like table or an article you know they can take it to the bank that it’s reliable. It’s been read. It’s been researched. I think this sort of notion of the source of the information you know is going to be more important than ever, but then again someone could say it’s NCSL. ChatGPT could say it’s NCSL and it may not be NCSL and so I think you know this need for vetted trusted sources is going to go up, but it’s going to be more difficult to pin down like the information actually come from. And of course, in our hyperpartisan world it’s getting so hard to have anybody that everybody relies on and says how can we know this is accurate. It’s not spun. It’s not biased information. So, there’s that.
And then are you familiar with the Alaska legislature who used ah the ChatGPT to write a bill?
AW: I’m not.
TS: You have to look it up. So, this was probably now about a month or so ago and I read this in the media so I hope it’s accurate. It was, you know, out of Alaska media and I don’t know but I really should look up who it was because I know a bunch of those folks. But he needed to a bill draft and didn’t get it when he needed it is my understanding and it was something somewhat innocuous like allowing gaming on the state ferry system. And asked the ChatGPT to write the bill and it was after submitted rations and that’s kind of what you referred to. Like you may get a version of something that’s 80% accurate or 75% accurate. But that took you five minutes versus maybe to get a first draft might take you three hours in pre-times. So, he gets this thing. He edits it a few times. It was pretty good. I mean I read it. It looked a lot like a legislative bill. Not perfect. But then with a little bit of human touch you get it tuned up. It’s in style as you said. Every legislator has its own style for bill drafts. So, this felt significant to me. This you know somewhat out here on the margins like it felt like a big deal. And yes, you know it will take folks who will have to check it and review it. But on the other hand, it will save them tons of time right that you can get the first draft done in that short amount of time.
(TM): 22:07
AW: Huge. And I think of that as a really significant story and I would encourage every legislator listening to this to think about one what is the potential then with your legislative council service if suddenly there is the ability to draft bills even faster. And yes, totally spot on – there is still going to be a need for the human touch on top to polish it and make it better because it is only going to get it 75%, 80% but it’s still 75%, 80%. And then I think the second interesting one is what happens when any lawmaker in your chamber can with the click of their fingers suddenly have a bill or an amendment ready to go and doesn’t have to take the time to go up and send it through council service. Like how does that change the process season. Like that is a real world that starts to happen quick quickly and like that changes some of the things that happen on the floor.
TS: This is a great example. We could play the what if game you know all afternoon because I think there are a multitude of hats that this could take and this is one right that you are sitting in committee normally. And most states have rules that if not all of them you know you have to send it in for style and substance that comes officially through the bill enactment agencies. But you know a lot of floor minutes are written on the fly usually with the help of a drafter. But now this is going to be a bit of a gamechanger because they could you know, I think. It’s one of many possibilities in this.
AW: And let me take it one step further and this is part where there is still a little bit of uncertainty but what happens to the information that is put into a Bard or a ChatGPT. Now I think Open AI has come out and said hey we are not using your data for training purposes. But it could be a different algorithm provider out there that is and so if you suddenly then have draft legislative texts in the algorithm what if that produces an incorrect answer for someone else. What if that has something that ends up being very confidential to the next step in the process and that’s now suddenly you know part of someone else’s AI answers they are thinking about. What’s happening. And then is what companies are wrestling with the proprietary data is they don’t want someone taking a trade secret or a piece of code or a piece of the actual property and putting it into one of these neuro networks that they don’t know what is exactly you know it’s how it is going to be used. And certainly, from the public reporting opening the eyes planning to release you know a business version that has more of guidance around how and what happens to the data and I assume that on that pathway there is going to be a government version as well. But right now, it is a little bit of the wild west in terms of well what goes on and what is the information in it and that’s what I think is really important that both staff and lawmakers and anyone involved at the government level needs to be thinking about and also why there needs to be a certain level of guidance about how and when you know neuro networks and AI should be used in the day to day workplace.
TS: So, one of the biggest lines on the job description for effective legislatures communicating with constituents, communicating with interest groups, with people who can help them in the policymaking role. So, it’s not just writing legislation. It’s writing communications you know back and forth with people in your district, people in your state, maybe even people around the nation. So how does have you thought about how it changes that part of their job?
AW: Significantly. So, I think a couple of things that we’ve talked about it is great at giving you essentially a rough draft to be able to work from. And so ideally it should reduce some of the work from legislators and their staff as they put together those communications out because you can ask for it to write you a draft that covers these specific things. Gets you 60% of the way there and then you can go and fine tune and touch it up.
I also think in the same level that means that there is going to be a lot more communication out there that is both more succinct, more well written, more to the point because underneath it AI is creating it. So, whether that’s coming from a fellow lawmaker or coming from a business or coming from someone who is trying to sell you know a piece of clothing to someone that’s going to be present and so it’s going to be more important than ever to follow some of those best practices in terms of being able to break through. I also think there is the interesting inverse, which means a lot more of the incoming to an individual lawmaker is going to be more specific, more tailored, more well thought through, more well researched and put together and composed because individuals are also going to be using it as they put together a content to communicate cause that policy briefs and so you are going to have much more information out there that is actually you know in many ways going to help. Also going to raise the bar in terms of some of the sifting that needs to come through. And also, is going to as we talked about earlier raise the importance of trusted sources of information whether it be an NCSL or someone else. It also raises the importance of human relationships because there’s another real component where when you are talking human to human, you know that there is not someone you know sitting there unless there is some sort of earpiece giving them you know the best talking points that Generative AI can help do. But rather that it’s that in person connection to break it down. But this is where I go back to this being the most significant technological development that we’ve seen since the internet is that it is going to really change the way that both legislators can interact with their constituents and also the amount of incoming information that is going to be presented to legislators.
TS: As we’ve ventured into the beginning of the beginning or the middle of the beginning, we’re thinking about you know the various computations of good and not good like you know what’s going to happen. Where will the ground rules have to be established and one of the things, I think is a major concern is the authenticity will just ooze out will be challenged. It’s already hard to determine authenticity in some ways, but legislators are really good at it and then constituents are really good at it in some ways. You know when you get something that was auto pinned from you know hey this is the president, ex-president or President Biden you know. I’m just so happy to hear about your wedding. Well, we know that Joe Biden didn’t sit down and write a note and it still means something to people. But and then when legislators start to get a whole bunch of incoming communication around that issue, they often know it’s sort of fake. It’s AstroTurf grassroots and it’s going to be a whole lot harder to tell what’s fake and what’s authentic. So, I don’t know. I’m just thinking about the role of authenticity in a world where there will be more and is AI generated material fake. Is that even?Is it fair to call it fake if someone has read it, they produced it. And then it gets you into the whole plagiarism questions. How does a legislator cope with that?
(TM): 28:51
AW: Yeah so, it’s something that we’ve thought about a lot as it’s our systems that people will often use to figure out which state legislator or member of Congress represents them and what’s the information that’s coming through. And one of the hard requirements that we have for all of our customers is that it must be an actual human that has a registered address who is a constituent who lives in the district who is ready to submit every single time they are reaching out and contacting their elected official. And that’s one of those important ethical standards that especially in this time of more information, it’s important to continue and there have been cases that we have not been affiliated with of people that essentially are creating you know fake constituents to send information or post fake regulatory comments.
Now these are very, very few, very, very rare, but is certainly something that is you know a possibility of risk that people have reported on and talked about. And so, it speaks to the importance of having that level of verification that it is a constituent and here is their address and they live in your district. Here is what it is that they have to say. But I do think that the piece to be aware of is that there is now a world that that constituent is using a, you know, generated put together their individual pitch to a lawmaker about why they should care about the issue and then click and press submit and sending that. And so it is then coming you know from their thoughts with a lot of backing generative AI of different points of statistical details or insight some of which may or may not be correct or fact checked based on what’s coming out of the AI algorithm and that’s what’s coming in and so that’s where I do think lawmakers are going to need to be aware that there incoming messages from their constituents whether it be part of a mass grassroots advocacy campaign or even a one up are about to get a lot more sophisticated, a lot more specific and even a lot more targeted where the message that you are receiving may look a lot different than the message that your colleague on the other side of the aisle is receiving. And that’s why this is such a major breakthrough in change in the technology that is out there and why legislators need to be thinking about okay what does this mean for me and also have probably a little bit of awareness and skepticism if they get something that says I don’t think that you know this is a constituent put this whole 10 page policy brief together about why X or Y thing is so important especially in the case that they might know the constituent before and realize that that might be something that they are going to run into.
TS: We know that when they start to see you know the boiler plate on an issue, it’s the same you know it’s a copy and paste into a system that’s sending them emails and that kind of thing and there emails you know frankly most of them are just massively overwhelmed. You know it gets to the point where it is impossible to keep up with unless you have a large staff like you are in a handful of large staff states to this is going to change that.
AW: But on that, here is an interesting thing. I’ve talked about one of the benefits of Generative AI is you can take all the emails that you receive from a constituent or from constituents in say a day and submit them into generative AI and say summarize these, count the number of similar messages that there are and let me know what are the topics that are most frequently being corresponded on. And so suddenly then what you know a U.S. Senator would have an entire office dedicated to doing. You as a lawmaker or you as your staff member could do with a few clicks of a button and so you have a way better way of being able to come in. And that’s where we are going to be at a world that you are going to need to use AI to help sort through the AI that is out there. So, you might have a message that was in part generated or started with AI, but you can turnaround and use it to summarize and you get back in the same place which is okay how many messages did I get about this issue. How many messages did I get about Y and what is it that came through and that is an example of one of the timesaving elements that have potential to save a tremendous amount of time for lawmakers and their staff.
TS: There are 7,000 legislators and 30,000 legislative staff, but they are going to start to do this whole thousand flowers bloom thing, right, where they are like oh here is something, I’m using it for. And then they will talk because you just came up with an idea how you could you know really use it to make your job a little easier and more streamlined on a legislative perspective versus that constant sense of being overwhelmed. Constant sense of overwhelm.
AW: Completely. And this is also where I put my policy hat on for a moment and encourage lawmakers to start asking their state governments about what are they doing to build set guidelines around the use of generative AI and also how are they using it to advance a more modern and efficient government. Because the same way that those implications for lawmakers and their staff, those implications for state governments in terms of what they can be doing to either provide more effective constituent services to save tremendous amounts of time to totally change government processes and it’s really excited. But to that end, it’s got to be a focused centralized deferred of different departments and sharing information and best practices and different pieces so that if one agency or branch figures it out, they are able to go and share that with another agency. But you can’t really do that unless you’ve got you know a centralized group or body that’s helping to coordinate that. And I think the key message you know from the conversation today is it’s happening whether you like it or not, you are going to have both your staff. You are going to have folks in the state government like people are out there using it. It is the most faceted piece of adopted technology that has been introduced in the last 20 years and it’s really hard I think to live in a world of oh well it’s not going to impact me. It’s not coming to me because it is already, and it is only going to increase in terms of its adoption.
(TM): 34:26
TS: I’m with you there Alex. I think that’s where we are. I heard a legislature maybe last week we had a meeting, and they were saying that they needed to they wanted to generate like a memorial piece of collateral celebrating let’s say international women’s day. So, they asked ChatGPT to write it from you know based on their circumstances in their state celebrating you know this recognized day. And then they blew it out on their social media you know and through Twitter I guess and Facebook and all of those. And the legislators said you know it saved me all this time. I couldn’t believe it. But then you know then the question is well did you say this was generated by is there any obligation to say I didn’t write this. This was written by GenAI and I don’t want to delve into that question. I just think this is the kind of thing that can happen over and over and over again. People are going to come up. We are going to start having sessions at NCSL’s be sure to register folks for the NCSL Legislative Summit in Indianapolis. We are going to have sessions at all of our meetings about like how are you using it. And then there is the other the opposite side of that of like is this good?Is this bad. Do we need guidelines. Does it need to be identified as coming from an AI tool or not.
I love it. I do I think it is just fascinating. I realize it’s probably gives people a lot of cause for pause and other people they are like where is this going to take us. I mean it does seem to be happening fast and this is not always people’s comfort zone when change happens like this. It happens fast.
AW: Completely. Look I think you are spot on of you know it’s both how does it impact you know anyone who is listening as an individual. But then also for the lawmakers, there is this very significant policy angle on this too. And my encouragement is that you need to know how it works and at least experiment and be enough of a user of it so that when you get asked to vote on a bill or introduce a bill that’s related to it, you understand the implications and can help all of us as a society collectively wrestle with this problem or opportunity.
TS: You make a really good point but most states don’t have most chambers don’t have a technology committee because technology kind of comes into everything. The health committee. Criminal justice committee. The education committee and it allows tech issues. But this person was suggesting we need technology committees that you know are looking at you know privacy laws that a lot of states are looking at and a number have passed. You know this is you need experts. You need people who can step back and you know we know there are the people who know everything there is to know about transportation you know. They’ve been in that world. We put them on the transportation committee. Well, you probably we could be thinking about a technology committee. Well, We should probably wrap it up. This was sort of a terrific conversation and I think the first of many that we are going to have and I’m glad you are excited about it. What else are you excited about Alex as we it’s always a good place to stop. What are you looking forward to?
AW: I think all the exciting things that everyone is going to be able to do with generative AI. It’s going to save a tremendous amount of time. There’s going to be a whole series of new business ideas that are launched from it. We are going to be able to both research and inform policy issues in entirely new ways with more information than ever before. There’s going to be more opportunities for collaboration because you can take new approaches to thinking about a given bill or a given initiative. And I also think it really has a fundamental way to transform how people are interacting with government both as constituents interacting with lawmakers as government providing services to individuals and that you know it has that really fundamental power and so it’s a whole new honestly tool and toolkit that people have to be able to interact and we’re sitting at really the start of it. And so, it’s awesome to be on here with you Tim you know in the early stage of it to talk about what it means. And I totally agree it’s a conversation that very much needs to continue.
TS: We will be talking about a lot more. Honestly Quorum you guys have adapted very well over your 10 years almost 10 years in business. People can learn a whole lot more about everything you are doing at …
AW: Quorum.us and we will also be at NCSL in August so you can find us there as well.
TS: It’s always great to see you so and thank you so much for your time. I love the conversation. Thanks Alex.
AW: Awesome. Thank you, Tim.,
TS: I’ve been talking with Alex Wirth, the co-founder and CEO of Quorum, about how artificial intelligence may affect state governments in general and legislatures in particular in the years to come. Thank you for joining me and Alex on this episode of “Legislatures: The Inside Storey” brought to you by the National Conference of State Legislatures.
Ed: You can check out all the podcasts from the National Conference of State Legislatures by searching for NCSL podcasts wherever you get your podcasts. Tim Storey, NCSL’s CEO hosts “Legislatures:The Inside Storey” where he focuses on leadership and legislatures. The “Our American States” podcast dives into some of the most challenging public policy issues facing legislators. On “Across the Aisle” host Kelley Griffin tells stories of bipartisanship. Also check out our special series “Building Democracy” on the history of legislatures.
(TM): 40:06