WEBVTT
1
00:00:03.000 --> 00:00:17.000
Hello Oregonians and welcome to Life and Legislation with Lucetta. If you've ever wanted to get to know your politicians personally or understand what it is they are actually doing, then you're in the right place. I'm your podcast host, Jessica Campbell.
2
00:00:17.000 --> 00:00:21.000
I'm your state representative for House District 24, Lucetta Elmer.
3
00:00:21.000 --> 00:00:27.000
This podcast is a place for you to get to know Representative Elmer both personally and professionally.
4
00:00:27.000 --> 00:00:41.000
We want Oregonians to feel connected with and educated politics. So we're so glad you've joined us on a fresh new podcast adventure as we cover all things about life and legislation.
5
00:00:41.000 --> 00:01:45.000
Hi Oregonians, and welcome back to Life and Legislation with Lucetta. I imagine many of you have heard about this new thing called artificial intelligence and we get to talk all about AI on this episode of Life and Legislation because this is certainly something that is affecting not just my life, not just your life, everybody's life, every and starting to affect legislation. Or should it? Could it? How does that have anything to do with our state government? And so we get to have a very special guest for this episode. I'm so excited to welcome Joe Sessions. Thank you so much for being willing to come here and talk about all things AI. And welcome back, of course, Rep. Elmer. We're so glad we get to have this conversation. A lot of people have been very excited to hear your thoughts on this. What does this have to do with legislation and what do we need? I know there's a lot of excitement around AI. There's also a lot of uncertainty, maybe even some fear. So let's start right away though, Joe, and just let you introduce yourself to everybody who you are and why you're even qualified to talk on this topic.
6
00:01:45.000 --> 00:02:27.000
Thank you very much. It's an honor to be here. Very thankful to be here. Why am I qualified? I think mostly because I'm human. I did work at Microsoft and their responsible AI division for a contract. Upon that contract ending, it was sort of in a hiring freeze time. It left me thinking like, okay, there's a lot on this topic of ethics. It's kind of been a pet topic of mine through my life anyway. And I've been a teacher and thinking about the future and next generations and what do they need for success. And I definitely puts a wrinkle in all that. And so I got interested in the topic and I, at some point I decided writing a book would be a good thing to do. I'm going to show my book.
7
00:02:27.000 --> 00:02:28.000
Yes.
8
00:02:28.000 --> 00:02:42.000
Okay, good. So AI and ethics. Adopting AI technology without losing. Losing our humanity. And then kind of a second little splash line there. A guide to scalability without an apocalypse.
9
00:02:42.000 --> 00:02:57.000
For anybody listening who has heard of AI but is still going, what exactly is this and how does it exactly work? I guess if you were to try to describe this to a 5 year old or a 95 year old.
10
00:02:57.000 --> 00:02:59.000
Yeah.
11
00:02:59.000 --> 00:03:02.000
What would you say of AI is
12
00:03:02.000 --> 00:04:57.000
so it's a bunch of math. The only answer is you actually need to log on to one of the things. Claude or ChatGPT or Grok or Gemini, like you just need to go get an account. I honestly think everyone actually needs to do that. Even if you think where your opinion is and will stay is like anti AI, AI is not going to talk you out of that, but you do need judgment on what it is. And I have to say, when I started working with it, I mean I have a mechanical engineering degree, I'm very kind of engineer, crawl under a car and change a transmission kind of a guy. And when I started talking with it, I'm like, this is math. I'm really just talking to a program like a computer has never responded this way to me. And I think for everybody that that's going to be an experience. It's just not a time to hide one's head in the sand. So that's the shortest answer I can say. It's just like the weather app on your phone makes all these predictions about every day, coming up every hour in those days. What's the temp going to be? That's not a human making all those changes. That's basically a math trained function that says, well, here's what weather has been for a long time, here's what weather will be. These tools are trained on tons of writing, tons and tons of writing on the Internet, lots of it copyrighted, which is a whole topic, but trained on all this writing. And it says, okay, these are the words that you just said to me. These are the words that came before. So what words probably would you want to come after that? In a nutshell, it's just going. Words generally go from here to here. You just gave me these words, so I'll give you these words. Just like Your phone for 10 years has predicted what next word you kind of want to do. This is predicting the next thousand words or 400 words depending on how big of an answer you you asked for, does that answer it?
13
00:04:57.000 --> 00:05:43.000
And at the same time it's taking all of the data from all over. And so for me, you know, I understand the math part because it's all this data. But then I Think that next layer. So maybe this is higher level math, but to be able to make those predictions is pretty remarkable. But then also, I don't know if the correct word is empathy, but how whatever it has been programmed to, then you know, like we'll say, well Lucetta, based on the fact that you like to be, I would say you should probably do this action. And you're like, whoa, really? Okay, so I mean there's a personalization there that we haven't seen before as consumers, which makes it so much more powerful and also addictive.
14
00:05:43.000 --> 00:06:41.000
Yeah, that's right. I mean for those of us who have gotten into it and use it for everything, as you were saying for I've got two ripe bananas, what can I make with becomes unreal. That there's people that aren't like that's very easy for it to do. I don't know how accurate this is. I asked my Claude to do some actual web research in present time and take a guess on how many people could we say have used ChatGPT or any of those type assistants at all. And it was like 33%. And I was like, wow, that's just hard for me. I'm walking around in a world with 2/3 of the people who don't do this yet don't know what this thing is. And then there's levels of usage way beyond just the chatting with it. Right. That I sometimes help businesses with and stuff like that. And this is coming, this is here. Like this is a thing.
15
00:06:41.000 --> 00:06:56.000
Well, and that's a great transition. It is here. This is not something that is just a trend or a fad that's going to go away. And whether you love it or hate it or understand it or are scared of it, it's part of our lives now, right?
16
00:06:56.000 --> 00:06:57.000
Yes.
17
00:06:57.000 --> 00:07:29.000
And I think that's where the conversation really falls into the life and legislation. This is part of our lives and is it going to be, should it be what's going to happen in terms of legislation, in terms of government, should there be laws around this? And how do you even do that? I know some people can go all the way to the other extreme of it's Terminator and it's going to take over and they're going to a gonna rule us as humans but for where we are right now. What are you seeing? Are people talking about this at the Capitol?
18
00:07:29.000 --> 00:09:07.000
Yeah. Just real quick to go back to what you were saying there. You know, I think of the 90s or whenever. I mean it was probably before the 90s, but I think the 90s is really when the computer phase for everyday consumers, you know, came on the scene and then the cell phones, and that was the next big one where it became a household thing. And so when back at how we have legislated computer software and how we have legislated cell phones, what does that look like? So AI is still so new, but there are several states that have, I want to say, almost every state proposed legislation in 2025 around AI. Some adopted a few things. For Oregon, our governor Kotak had an advisory council that she enacted to report back to her in 2020. Yes, there are lots of conversations. We haven't done any hardcore policies around it, but there are lots of conversations. And probably the most significant thing that we've had come through is we had a public records request that came into the legislature for my caucus and wanting to know how many times, every time we had ever used AI and to turn all of those records over to this requester, and you're talking hundreds, if not thousands of hours of research to figure that out. That's huge, and that's costly. So conversations are happening. Legislation hasn't really transpired yet, but definite conversations in the building. Yeah.
19
00:09:07.000 --> 00:09:31.000
Yeah. Because you think of all the things that could happen. I mean, I found a lot of things with AI, like, oh, this is so great. It can help me, you know, understand a medical record from, you know, my dad's cancer diagnosis. It can help me find a really quick recipe. It can do all these really nice things. But then you see, what it also can do is create stuff that's fake. And you see stuff on social media. It's like, that's not a real video.
20
00:09:31.000 --> 00:09:32.000
Right.
21
00:09:32.000 --> 00:09:36.000
That's not a real photo. I mean, talk about fake news to a whole new degree.
22
00:09:36.000 --> 00:09:37.000
No, I know.
23
00:09:37.000 --> 00:09:40.000
And you think of, gosh, what could that do in the political world?
24
00:09:40.000 --> 00:09:42.000
Well, it also can give you wrong information.
25
00:09:42.000 --> 00:09:43.000
Yeah.
26
00:09:42.000 --> 00:10:26.000
And that's where, I mean, not just my job, I think any professional job, if you're just relying on this as, like, your Wikipedia or something, and then you go, and I stand on the House floor before all of my members, and I fight for a bill based on this. And then you're like, chat, you failed me. You know, if it's wrong. Luckily, that has not happened to me. But it did fail me with an airport story when it told me that I had access to a certain lounge and I went up to give my ticket, and they're, like, denied. So I told chat, like, you, you were wrong. And then I love the response I was, I was completely wrong and, you know, moving on. I want my money back.
27
00:10:26.000 --> 00:11:01.000
Yeah, yeah. I mean, when you think of what kinds of. We've said this before, I mentioned in a couple of episodes ago that the, the first grade teacher in me wants there to be laws of, if you can't say something nice, don't say anything at all. Like, there should be a, A law that says you can't do mean commercials. You can only say nice things about what you're going to do. You can't say mean things about other people. But you think, gosh, what could AI do with that? I mean, there's already so much slander around, and now you could create these videos and photos of other people during election campaigns and all of that. Would that be.
28
00:11:01.000 --> 00:11:02.000
It's happening.
29
00:11:02.000 --> 00:11:08.000
I mean, would somebody propose a bill that says, we're going to make that illegal? Can you.
30
00:11:08.000 --> 00:11:47.000
You can propose a bill for anything, literally. You know, the thing is. And maybe you can speak to this, too. I mean, there's not enough time to really see how this plays out in the courts. And so, you know, when you're proposing legislation, like, we don't have data to draw on, like we do on a lot of other things to say. Well, you know, in the courts, in this case, this is how the court's ruled. So therefore, this is how we interpret the law. You know, we don't. We're still in the newbie stages, so we don't have that yet, but certainly we can. Certainly we're having the conversations. We're seeing it play out in real time. And I know there will be legislation around this.
31
00:11:47.000 --> 00:12:04.000
Yeah, well, and pausing on the legislation piece for a minute, Joe, right now, where is AI in its development? We've talked a little bit already of some of the benefits, but then there's also maybe some risks and fears. I imagine some of that's in your book.
32
00:12:04.000 --> 00:12:05.000
Sure.
33
00:12:05.000 --> 00:12:17.000
So if you want to take a moment or two just to address that, where are we and what are the challenges? What do we need to be careful of compared to how we can use this to benefit humanity?
34
00:12:17.000 --> 00:13:56.000
Yeah. Yeah, good. We'll see how I do it. Keeping it short, it's such a big topic. And I wrote a whole book, I mean, on the legislation thought, you know, writing a book clarifies a lot of thoughts for the author. As you write things, you go, okay, well, what is bad relative to AI? What could come in and be against what we want? And I honestly just kind of kept landing on a very American idea of life, liberty and the pursuit of happiness. That's the basic state of things that we want to preserve. You know, the country was started to try to prevent weird things that happen in government that make it so the average citizen can't just go about his life and make a life out of it, you know, and do life things and pursue that happiness. You're not guaranteed happiness, but you have a freedom and ability within the place you live to go pursuing happiness. That's one thought of it. As far as it being scary. I mean, there's, you know, to sound very like these tech guys that talk about it. There's many failure modes, there's many way. There's, there's a handful of ways that we could see this going wrong. You asked, where is AI right now? So right now, this week, Claude released a model they call Fable. I asked this new model, okay, what do you see with the future of work with this? And here are some of my thoughts. But don't try to please me, just tell me. Like, I just want, I want, I'm telling you all my thoughts just to let you know. I want the deep answer, not the shallow one. Like, what do you really think about this?
35
00:13:56.000 --> 00:13:58.000
You're asking AI about AI?
36
00:13:58.000 --> 00:15:11.000
Yes, exactly. What's the society going to roll out and do with AI in the picture as it gets smarter and everything else? And, you know, I won't try to summarize that answer too much, but I'll just say, woof. I'm talking to a higher IQ person now. This, this thing now, right. It's not a person. It has no skin in the game. It's very detached from what's happening. It's, you know, basically instructed to give like, social good answers. And it's really trying to do me that, to do that for me, you know, to the fears that it talks about, you know, like, a lot of things become very easy to make, a lot of things especially when you go. Robotics is coming along too. Robots and AI could get pizza to your door. It could bring you a drink at a restaurant table. And with probably robots that exist right now, it wouldn't fly people. It's not acceptable. But there are cities. Have you been to a city where you've seen no driver, car, A Waymo. Yeah, that's now. And a lot of people don't think it is, but it's like, no. In cities there's driverless cars have been going around for two or three years.
37
00:15:11.000 --> 00:15:12.000
In Portland, they're being tested.
38
00:15:12.000 --> 00:17:09.000
Yeah, I think the last thing I'd say about where it Is right now the big companies that are making these big models, their biggest and best and newest model, they call these the frontier models. Those are now a bit training themselves. They largely use the AI to train the next AI, which that just allows a bit of this hockey stick type thing. So that becomes sort of the Skynet question, which I don't want to go too deep in that. This is Oregon. Like it's a bigger topic, honestly than Oregon. You know, it actually, actually kind of needs to be a planet wide cooperation against the thing that is about to get smarter than any of us at the Oregon level. I mean businesses I think are wise to get going. I mean we're not going to just cancel it all. Like, you know, I don't see a 2030 where oh, we all decided no more AI, utterly not going to happen. Right. That's zero chance of happening. Then the legislation I think goes to specific use cases. You go, hey, you can't do that. You can't make, you can't fake intimate images of other real people. And there's laws being made about these particular things. But businesses, you know, are going to adopt what's legal. It becomes a bit of a playing field of like, okay, well if you sell pizza and it's cheaper to get it to somebody's house with little robot things, how do you say no to that? Well, if you make it illegal, well then we could still have our young people driving old cars bringing pizzas to people's house. Well, that becomes a thought. It's just a thought experiment I have in my book. But how quickly we let jobs change is one kind, I mean hostile and destructive actions. Yeah, you have to name them and make laws about those individually. So as usual I have big thoughts and they take a while to say thank you for your patience.
39
00:17:09.000 --> 00:18:23.000
Yeah, well, and you mentioned so many things within that and that kind of turns back to the laws. And I think at some level, probably everybody I would hope would agree that we would at a bare minimum want laws around AI doing terrible things that would harm people. 100% basic foundational level there. But then you start getting into some of those things where, okay, AI could make it better for this group, but in doing so it makes it worse for maybe this group. And now you're. Which is kind of the same with almost every law you put out. Okay, we're making something better here, but that comes with a new consequence. Talking about businesses, I know business is your passion and if there's a way for businesses to use AI and save a lot of Money and do things better, but at what cost does that mean now all these people don't have jobs? And that's something I think everyone's a little bit afraid of. Is AI going to replace all of our jobs? And you think of all the youth going into college and starting to do their trainings of should I even get this degree? Is this job even going to be a job in three years or five years? So then those get a little tricky.
40
00:18:23.000 --> 00:20:00.000
Yeah, absolutely. I went down a couple years ago to California and there was a conference hosted by Google and Apple and ebay. And at the time they were requesting legislation around sort of the things that they were developing that could, within the phones, that could go to some of our countries that are enemies and then be used against us. So like if we were going to say, you know, you have the app, you know, you have Gemini or you have ChatGPT or whatever, could it get into the wrong hands of someone that says, you know, create a bomb that would take out Oregon, you know, and so there was this hold and this sort of black, you know, they didn't go into certain areas, it wasn't allowed to be sold into certain countries that were our enemies. So that they couldn't. So that their product, Apple and Google's products, so that they couldn't then in turn take this, the product that we had made this AI and use it against us. Easy to get around that for them. They can buy it on the black market, they can do all kinds of different things. But it was really eye opening to see the magnitude of what we're talking about and really how this can be detrimental and also all of the positive things that we can do with AI and just how massive this conversation is and how we handle it. Because their point was this is growing at such a rate that if we don't have some sort of boundaries around this, it will be too late. And then what? How do we bring it back?
41
00:20:00.000 --> 00:20:15.000
Well, on that point, Joe, if you were to ask our state government to make one or two laws that you think are critical right now, from what you're seeing, what do you think the first two laws should be around AI
42
00:20:15.000 --> 00:22:22.000
at the state level? The first one that comes to mind, this is a great question. The first one that comes to mind honestly is around education. I was at the town hall reading and really understanding English and beyond that, literature, reading lots of stories of quality that talk about the human experience and get one very literate. We're going to be talking to computers, you know, Oregon, suffering in our education performance as an educator in my life, there's lots of wrong with how we measure success. And you know, my kids have been in some private and some public and some really high end public education and they've seen different things to have different judgment. And it's like, you know, dad, in this one school system, if you have kids read out loud, it's just painful. Like. And they're talking middle school, early high school age stuff. They're like, they just, they can't. This is an area that's been allowed to be polluted by new thoughts about education. And it's like, this isn't hard. Guys. Teach kids the phonics, teach what the letters mean and get them to build up words. That might sound like a funny answer to your question, but at the state level humans are going to have to. We have the chance to either rise to greater states of sort of enlightenment is the best way I know to say it. Where people have landed on their values by age 18 and they've like worked out what's important and what works long range. But we're such a consumer video game. Like kids on the phones babysat via movies or the Internet, TikTok, whatever. We are missing the mark on creating this enlightened next generation that has brighter and longer thinking, more judgment about what's right and wrong than we've had. And that would be the first thing at a state level that I think people can do.
43
00:22:22.000 --> 00:22:33.000
And you're preaching to the choir. As a former first grade teacher who is all phonics all the time. How does AI play into that?
44
00:22:33.000 --> 00:23:58.000
The first thing I would say is it just emphasizes the need. A great thought piece that lots of people know is Wall E, the movie. And for some people that just sticks. They like the movie is like three quarters about the robot going around on the ground on a deserted earth. But one little part of it is about the humans who've left Earth because they messed it up. And then they're living kind of in a big spaceship that's like a cruise ship. Like they're all floating around in chairs and they literally can't walk anymore when they get back to Earth because their bones and muscles have atrophy and they're just floating around and drinking sodas and watching the next video entertainment and like they're doing nothing. You know, in a way we're heading that like that's what our culture is moved toward. We don't actually, we try to protect people from risk so much that, you know, kids don't ride bikes. I Had my kid riding to a dance class and I was so proud, like, you know, 90s dad. What's it have to do with AI is that we have such an open gate to going in that Wall E direction, going so much further than we already are. Social media, I think a great speaker on the topic these days is Tristan Harris, who used to be an ethicist at Google. Basically quit because they weren't listening to him and he was part of that fantastic documentary, the Social.
45
00:23:58.000 --> 00:23:59.000
Social Dilemma.
46
00:23:58.000 --> 00:24:01.000
Exactly. Like every American needs to see that.
47
00:24:01.000 --> 00:24:02.000
Yes.
48
00:24:02.000 --> 00:24:42.000
And he says, look at how we basically failed with social media. I think when you just come back to the idea of like something that's negative or a hostile action against your countrymen, let's say, right? It's like, okay, but we can make money by having kids get all this fomo. And oh, Julie got more likes than me from my photo. I guess she's prettier. And like literally young teenage female suicide triples. And we don't rush in with that with legislation. Our sort of enlightened humanity to me is where my mind goes with what our choices are.
49
00:24:42.000 --> 00:25:07.000
Well, I love the Wall E example and what you said about this opportunity for humans to rise and use something to make everything better. And we have all of these things, whether it's tv, cell phones, social media, you name anything, even a certain type of food, people can use it to make their lives better, they can use it to make lives worse.
50
00:25:07.000 --> 00:25:10.000
There's short term better versus long term better.
51
00:25:10.000 --> 00:26:04.000
So you just look at all of the things the same thing with AI, but it's just on a magnitude that we've never seen before where you're thinking at the best, oh my gosh, could AI cure cancer? Right, but then what if an enemy country has AI and destroys us all? I mean, you're looking at these extremes of the best of humanity, the worst of humanity. And so kind of as we were getting ready to wrap up here, and this could go on for so many episodes, right? And it probably will because we probably will need you back in a month because it'll have changed so much right then. But as you, Rep. Elmer, are looking ahead to this next long session, which feels far away but will blink and it'll be here. Then what is the next step? I mean, putting, I imagine almost everybody is thinking of some bills to put forward around AI, Is there one that you have in particular in mind?
52
00:26:04.000 --> 00:26:33.000
No, not yet. As you know, it's an election year and so most of the brain space right now is going towards elections. But we did just get out of a policy meeting this week to where we're starting to prep and look at all the different pieces of policy and, you know, each different area and what we want to make sure that we're touching on and hitting on and not duplicating and getting everyone going in the same direction. So starting those conversations. But I don't have anything specific yet.
53
00:26:33.000 --> 00:26:34.000
Right.
54
00:26:34.000 --> 00:26:48.000
What would you most want to see if you were in charge and it wasn't this whole long process of making a bill become a law and you could just pick a law tomorrow around AI? Is there anything that you would like to just make happen?
55
00:26:48.000 --> 00:28:17.000
That's such a big question, but it's a great question. I think it would be easy to pull on all of the wonderful things that AI can do and then use that to mandate that we use it so that we all can be the best versions that we can be. So taking your example within education, I mean, yeah, use AI to find out the best research out there, the one that gets the most kids reading in the quickest amount of time. That is time tested and it's true. And you implement that into the schools to bring those standards up and then repeat that in all the areas. Math, science. That's an easy one to say. We can see the good come out of something quickly using AI. Yeah, look at the healthcare. I mean I could ask it 50 things right now to please solve and if it could do that in the 11 seconds, think time that it takes. That would be amazing. But not just curing cancer and curing diseases, but also how do we handle the budget deficits and how do we handle Medicaid and Medicare and all of the questions around there and how do we get people to have doctors in a more timely manner? And can we streamline this not just within one state, but throughout the nation? And I mean we have telehealth, but what more could we do to improve upon our healthcare system and then go on to the next topic and the next topic. Yeah, for sure.
56
00:28:17.000 --> 00:28:29.000
That's such an interesting question. I wonder if you were to put in the most recent revenue forecast into ChatGPT or any of the other servers and say, what should Oregon do?
57
00:28:29.000 --> 00:28:33.000
Well, believe me, that's been done. And there are responses.
58
00:28:33.000 --> 00:28:51.000
Yes, that's so fascinating to see what could come by using AI with legislation. And I'm sure we'll have so many follow up episodes about what bills are being put forward, if they're making it through. Will things go to votes? Perhaps. And just.
59
00:28:51.000 --> 00:30:07.000
I will comment on that, you know, you asked about the revenue. That definitely has been done. You get an answer back. And half of your population doesn't want to hear that. And also the answer that you're getting that you are receiving back has bias based on the way in which you ask the question. And so from my perspective as a legislator, one of the I find to be extremely helpful is that my chat knows and will correct me. I don't know if that's real or not, but I feel corrected when I start to rant a little too much or I ask a question that maybe I didn't realize was so hardcore biased. And it will call that out because I've asked enough for information. And then I will say, now what would my Democratic colleague say to this? How would they poke holes in this? And then. And I'm like, whoa. And my eyes are open to my own biases or my own things that I was taking as fact and maybe they weren't fact. And then I will use it to fact check, well, is this fact or is this fact? And what are my sources? Show me this. So it's been extremely useful for me to truly look at legislation in a more unbiased way. And that's a huge positive for me.
60
00:30:07.000 --> 00:30:08.000
Yeah, that's beautiful.
61
00:30:08.000 --> 00:30:09.000
Wow.
62
00:30:09.000 --> 00:32:16.000
I mean, moving to reason is part of a thing I mentioned in the book. It's like we can't wish for a repeat of any previous decade that you happen to like, in American history. I feel like I went through this little microcosm in the Dalles, Oregon of just a beautiful high school experience. I had a great time, you know, a nice balance of getting in trouble and not. And like it was just a happy place to be. And that's not coming back. We have to think of something better than that. But just throw AI at it doesn't always handle it. But this that you're saying of like, okay, let's. One thing about it is it doesn't really have skin in the game except for to whatever degree that its developers want to keep you hooked. Right. I was reading the Claude instructions to this newest model that was leaked and I was reading it's very nice. And basically it said don't just ask them a question at the end to keep them going. Like, actually they had made the choice to tell their AI to not just keep you hooked. So they all sort of have that skin in the game. Like, how much do you stay on it and use it? But really, if they help with reasoning, it's not going to benefit the model. Itself isn't going to benefit or not benefit based on some organ statute that goes through. So it's something, you sit there and you talk out the pros and cons and you're thoughtful to have it, give it the counterpoint. That's probably, that's probably going to be in my next book. That's really brilliant that we reason out these things further and look at the long term. You know, people talk about second or third order consequences to a decision. The more and more we can do that, you know, your first grade teacher point of like, can't say mean things about each other. That's generally not helpful to society at large. And if we can, if we can like talk ourselves out of that in our little chatgpt, then that's a beautiful thing.
63
00:32:16.000 --> 00:32:35.000
Well, I just want to commend you for using ChatGPT the way you have been because truly I think there's a lot of leaders that they're going to use it a different way. The goal is how do I defeat the other side as opposed to help me understand the other side so that we can move forward all in the
64
00:32:35.000 --> 00:32:43.000
name of good governance. Maybe my idea isn't the best, but help me understand it so then I can advocate for it.
65
00:32:43.000 --> 00:33:02.000
I think that is a beautiful place to close. And thank you for your leadership and thank you, Joe, so much for taking time to come talk about AI with us again. I think we'll have to to have you back in the future as more bills get proposed and get your thoughts on that and more about your book. You want to hold that up one more time for everybody?
66
00:33:02.000 --> 00:33:03.000
Yeah, sure.
67
00:33:03.000 --> 00:33:09.000
We'll include a link in the episode notes for anybody who's interested in wanting to read a little bit more about AI and ethics.
68
00:33:09.000 --> 00:33:17.000
Yes, thank you. Yeah, I mean, the goal with the book is to invite everybody into the conversation because the future is what we will build.
69
00:33:17.000 --> 00:33:19.000
Right. And it's not going away.
70
00:33:19.000 --> 00:33:20.000
That's right.
71
00:33:20.000 --> 00:33:59.000
So let's not be the people in Wall E. Let's take a different direction and see what we can do with that. So I appreciate your time so much. Thank you both for being here and part of this conversation. And thank you so much, listeners for tuning in to this episode of Life and Legislation. Hopefully you found some interesting points and got your curiosity piqued a little bit more and that you're gonna follow along and hear what's going on in Legislation. So we hope you come back and join us next time for Life and Legislation with Lucetta.
1
00:00:03.000 --> 00:00:17.000
Hello Oregonians and welcome to Life and Legislation with Lucetta. If you've ever wanted to get to know your politicians personally or understand what it is they are actually doing, then you're in the right place. I'm your podcast host, Jessica Campbell.
2
00:00:17.000 --> 00:00:21.000
I'm your state representative for House District 24, Lucetta Elmer.
3
00:00:21.000 --> 00:00:27.000
This podcast is a place for you to get to know Representative Elmer both personally and professionally.
4
00:00:27.000 --> 00:00:41.000
We want Oregonians to feel connected with and educated politics. So we're so glad you've joined us on a fresh new podcast adventure as we cover all things about life and legislation.
5
00:00:41.000 --> 00:01:45.000
Hi Oregonians, and welcome back to Life and Legislation with Lucetta. I imagine many of you have heard about this new thing called artificial intelligence and we get to talk all about AI on this episode of Life and Legislation because this is certainly something that is affecting not just my life, not just your life, everybody's life, every and starting to affect legislation. Or should it? Could it? How does that have anything to do with our state government? And so we get to have a very special guest for this episode. I'm so excited to welcome Joe Sessions. Thank you so much for being willing to come here and talk about all things AI. And welcome back, of course, Rep. Elmer. We're so glad we get to have this conversation. A lot of people have been very excited to hear your thoughts on this. What does this have to do with legislation and what do we need? I know there's a lot of excitement around AI. There's also a lot of uncertainty, maybe even some fear. So let's start right away though, Joe, and just let you introduce yourself to everybody who you are and why you're even qualified to talk on this topic.
6
00:01:45.000 --> 00:02:27.000
Thank you very much. It's an honor to be here. Very thankful to be here. Why am I qualified? I think mostly because I'm human. I did work at Microsoft and their responsible AI division for a contract. Upon that contract ending, it was sort of in a hiring freeze time. It left me thinking like, okay, there's a lot on this topic of ethics. It's kind of been a pet topic of mine through my life anyway. And I've been a teacher and thinking about the future and next generations and what do they need for success. And I definitely puts a wrinkle in all that. And so I got interested in the topic and I, at some point I decided writing a book would be a good thing to do. I'm going to show my book.
7
00:02:27.000 --> 00:02:28.000
Yes.
8
00:02:28.000 --> 00:02:42.000
Okay, good. So AI and ethics. Adopting AI technology without losing. Losing our humanity. And then kind of a second little splash line there. A guide to scalability without an apocalypse.
9
00:02:42.000 --> 00:02:57.000
For anybody listening who has heard of AI but is still going, what exactly is this and how does it exactly work? I guess if you were to try to describe this to a 5 year old or a 95 year old.
10
00:02:57.000 --> 00:02:59.000
Yeah.
11
00:02:59.000 --> 00:03:02.000
What would you say of AI is
12
00:03:02.000 --> 00:04:57.000
so it's a bunch of math. The only answer is you actually need to log on to one of the things. Claude or ChatGPT or Grok or Gemini, like you just need to go get an account. I honestly think everyone actually needs to do that. Even if you think where your opinion is and will stay is like anti AI, AI is not going to talk you out of that, but you do need judgment on what it is. And I have to say, when I started working with it, I mean I have a mechanical engineering degree, I'm very kind of engineer, crawl under a car and change a transmission kind of a guy. And when I started talking with it, I'm like, this is math. I'm really just talking to a program like a computer has never responded this way to me. And I think for everybody that that's going to be an experience. It's just not a time to hide one's head in the sand. So that's the shortest answer I can say. It's just like the weather app on your phone makes all these predictions about every day, coming up every hour in those days. What's the temp going to be? That's not a human making all those changes. That's basically a math trained function that says, well, here's what weather has been for a long time, here's what weather will be. These tools are trained on tons of writing, tons and tons of writing on the Internet, lots of it copyrighted, which is a whole topic, but trained on all this writing. And it says, okay, these are the words that you just said to me. These are the words that came before. So what words probably would you want to come after that? In a nutshell, it's just going. Words generally go from here to here. You just gave me these words, so I'll give you these words. Just like Your phone for 10 years has predicted what next word you kind of want to do. This is predicting the next thousand words or 400 words depending on how big of an answer you you asked for, does that answer it?
13
00:04:57.000 --> 00:05:43.000
And at the same time it's taking all of the data from all over. And so for me, you know, I understand the math part because it's all this data. But then I Think that next layer. So maybe this is higher level math, but to be able to make those predictions is pretty remarkable. But then also, I don't know if the correct word is empathy, but how whatever it has been programmed to, then you know, like we'll say, well Lucetta, based on the fact that you like to be, I would say you should probably do this action. And you're like, whoa, really? Okay, so I mean there's a personalization there that we haven't seen before as consumers, which makes it so much more powerful and also addictive.
14
00:05:43.000 --> 00:06:41.000
Yeah, that's right. I mean for those of us who have gotten into it and use it for everything, as you were saying for I've got two ripe bananas, what can I make with becomes unreal. That there's people that aren't like that's very easy for it to do. I don't know how accurate this is. I asked my Claude to do some actual web research in present time and take a guess on how many people could we say have used ChatGPT or any of those type assistants at all. And it was like 33%. And I was like, wow, that's just hard for me. I'm walking around in a world with 2/3 of the people who don't do this yet don't know what this thing is. And then there's levels of usage way beyond just the chatting with it. Right. That I sometimes help businesses with and stuff like that. And this is coming, this is here. Like this is a thing.
15
00:06:41.000 --> 00:06:56.000
Well, and that's a great transition. It is here. This is not something that is just a trend or a fad that's going to go away. And whether you love it or hate it or understand it or are scared of it, it's part of our lives now, right?
16
00:06:56.000 --> 00:06:57.000
Yes.
17
00:06:57.000 --> 00:07:29.000
And I think that's where the conversation really falls into the life and legislation. This is part of our lives and is it going to be, should it be what's going to happen in terms of legislation, in terms of government, should there be laws around this? And how do you even do that? I know some people can go all the way to the other extreme of it's Terminator and it's going to take over and they're going to a gonna rule us as humans but for where we are right now. What are you seeing? Are people talking about this at the Capitol?
18
00:07:29.000 --> 00:09:07.000
Yeah. Just real quick to go back to what you were saying there. You know, I think of the 90s or whenever. I mean it was probably before the 90s, but I think the 90s is really when the computer phase for everyday consumers, you know, came on the scene and then the cell phones, and that was the next big one where it became a household thing. And so when back at how we have legislated computer software and how we have legislated cell phones, what does that look like? So AI is still so new, but there are several states that have, I want to say, almost every state proposed legislation in 2025 around AI. Some adopted a few things. For Oregon, our governor Kotak had an advisory council that she enacted to report back to her in 2020. Yes, there are lots of conversations. We haven't done any hardcore policies around it, but there are lots of conversations. And probably the most significant thing that we've had come through is we had a public records request that came into the legislature for my caucus and wanting to know how many times, every time we had ever used AI and to turn all of those records over to this requester, and you're talking hundreds, if not thousands of hours of research to figure that out. That's huge, and that's costly. So conversations are happening. Legislation hasn't really transpired yet, but definite conversations in the building. Yeah.
19
00:09:07.000 --> 00:09:31.000
Yeah. Because you think of all the things that could happen. I mean, I found a lot of things with AI, like, oh, this is so great. It can help me, you know, understand a medical record from, you know, my dad's cancer diagnosis. It can help me find a really quick recipe. It can do all these really nice things. But then you see, what it also can do is create stuff that's fake. And you see stuff on social media. It's like, that's not a real video.
20
00:09:31.000 --> 00:09:32.000
Right.
21
00:09:32.000 --> 00:09:36.000
That's not a real photo. I mean, talk about fake news to a whole new degree.
22
00:09:36.000 --> 00:09:37.000
No, I know.
23
00:09:37.000 --> 00:09:40.000
And you think of, gosh, what could that do in the political world?
24
00:09:40.000 --> 00:09:42.000
Well, it also can give you wrong information.
25
00:09:42.000 --> 00:09:43.000
Yeah.
26
00:09:42.000 --> 00:10:26.000
And that's where, I mean, not just my job, I think any professional job, if you're just relying on this as, like, your Wikipedia or something, and then you go, and I stand on the House floor before all of my members, and I fight for a bill based on this. And then you're like, chat, you failed me. You know, if it's wrong. Luckily, that has not happened to me. But it did fail me with an airport story when it told me that I had access to a certain lounge and I went up to give my ticket, and they're, like, denied. So I told chat, like, you, you were wrong. And then I love the response I was, I was completely wrong and, you know, moving on. I want my money back.
27
00:10:26.000 --> 00:11:01.000
Yeah, yeah. I mean, when you think of what kinds of. We've said this before, I mentioned in a couple of episodes ago that the, the first grade teacher in me wants there to be laws of, if you can't say something nice, don't say anything at all. Like, there should be a, A law that says you can't do mean commercials. You can only say nice things about what you're going to do. You can't say mean things about other people. But you think, gosh, what could AI do with that? I mean, there's already so much slander around, and now you could create these videos and photos of other people during election campaigns and all of that. Would that be.
28
00:11:01.000 --> 00:11:02.000
It's happening.
29
00:11:02.000 --> 00:11:08.000
I mean, would somebody propose a bill that says, we're going to make that illegal? Can you.
30
00:11:08.000 --> 00:11:47.000
You can propose a bill for anything, literally. You know, the thing is. And maybe you can speak to this, too. I mean, there's not enough time to really see how this plays out in the courts. And so, you know, when you're proposing legislation, like, we don't have data to draw on, like we do on a lot of other things to say. Well, you know, in the courts, in this case, this is how the court's ruled. So therefore, this is how we interpret the law. You know, we don't. We're still in the newbie stages, so we don't have that yet, but certainly we can. Certainly we're having the conversations. We're seeing it play out in real time. And I know there will be legislation around this.
31
00:11:47.000 --> 00:12:04.000
Yeah, well, and pausing on the legislation piece for a minute, Joe, right now, where is AI in its development? We've talked a little bit already of some of the benefits, but then there's also maybe some risks and fears. I imagine some of that's in your book.
32
00:12:04.000 --> 00:12:05.000
Sure.
33
00:12:05.000 --> 00:12:17.000
So if you want to take a moment or two just to address that, where are we and what are the challenges? What do we need to be careful of compared to how we can use this to benefit humanity?
34
00:12:17.000 --> 00:13:56.000
Yeah. Yeah, good. We'll see how I do it. Keeping it short, it's such a big topic. And I wrote a whole book, I mean, on the legislation thought, you know, writing a book clarifies a lot of thoughts for the author. As you write things, you go, okay, well, what is bad relative to AI? What could come in and be against what we want? And I honestly just kind of kept landing on a very American idea of life, liberty and the pursuit of happiness. That's the basic state of things that we want to preserve. You know, the country was started to try to prevent weird things that happen in government that make it so the average citizen can't just go about his life and make a life out of it, you know, and do life things and pursue that happiness. You're not guaranteed happiness, but you have a freedom and ability within the place you live to go pursuing happiness. That's one thought of it. As far as it being scary. I mean, there's, you know, to sound very like these tech guys that talk about it. There's many failure modes, there's many way. There's, there's a handful of ways that we could see this going wrong. You asked, where is AI right now? So right now, this week, Claude released a model they call Fable. I asked this new model, okay, what do you see with the future of work with this? And here are some of my thoughts. But don't try to please me, just tell me. Like, I just want, I want, I'm telling you all my thoughts just to let you know. I want the deep answer, not the shallow one. Like, what do you really think about this?
35
00:13:56.000 --> 00:13:58.000
You're asking AI about AI?
36
00:13:58.000 --> 00:15:11.000
Yes, exactly. What's the society going to roll out and do with AI in the picture as it gets smarter and everything else? And, you know, I won't try to summarize that answer too much, but I'll just say, woof. I'm talking to a higher IQ person now. This, this thing now, right. It's not a person. It has no skin in the game. It's very detached from what's happening. It's, you know, basically instructed to give like, social good answers. And it's really trying to do me that, to do that for me, you know, to the fears that it talks about, you know, like, a lot of things become very easy to make, a lot of things especially when you go. Robotics is coming along too. Robots and AI could get pizza to your door. It could bring you a drink at a restaurant table. And with probably robots that exist right now, it wouldn't fly people. It's not acceptable. But there are cities. Have you been to a city where you've seen no driver, car, A Waymo. Yeah, that's now. And a lot of people don't think it is, but it's like, no. In cities there's driverless cars have been going around for two or three years.
37
00:15:11.000 --> 00:15:12.000
In Portland, they're being tested.
38
00:15:12.000 --> 00:17:09.000
Yeah, I think the last thing I'd say about where it Is right now the big companies that are making these big models, their biggest and best and newest model, they call these the frontier models. Those are now a bit training themselves. They largely use the AI to train the next AI, which that just allows a bit of this hockey stick type thing. So that becomes sort of the Skynet question, which I don't want to go too deep in that. This is Oregon. Like it's a bigger topic, honestly than Oregon. You know, it actually, actually kind of needs to be a planet wide cooperation against the thing that is about to get smarter than any of us at the Oregon level. I mean businesses I think are wise to get going. I mean we're not going to just cancel it all. Like, you know, I don't see a 2030 where oh, we all decided no more AI, utterly not going to happen. Right. That's zero chance of happening. Then the legislation I think goes to specific use cases. You go, hey, you can't do that. You can't make, you can't fake intimate images of other real people. And there's laws being made about these particular things. But businesses, you know, are going to adopt what's legal. It becomes a bit of a playing field of like, okay, well if you sell pizza and it's cheaper to get it to somebody's house with little robot things, how do you say no to that? Well, if you make it illegal, well then we could still have our young people driving old cars bringing pizzas to people's house. Well, that becomes a thought. It's just a thought experiment I have in my book. But how quickly we let jobs change is one kind, I mean hostile and destructive actions. Yeah, you have to name them and make laws about those individually. So as usual I have big thoughts and they take a while to say thank you for your patience.
39
00:17:09.000 --> 00:18:23.000
Yeah, well, and you mentioned so many things within that and that kind of turns back to the laws. And I think at some level, probably everybody I would hope would agree that we would at a bare minimum want laws around AI doing terrible things that would harm people. 100% basic foundational level there. But then you start getting into some of those things where, okay, AI could make it better for this group, but in doing so it makes it worse for maybe this group. And now you're. Which is kind of the same with almost every law you put out. Okay, we're making something better here, but that comes with a new consequence. Talking about businesses, I know business is your passion and if there's a way for businesses to use AI and save a lot of Money and do things better, but at what cost does that mean now all these people don't have jobs? And that's something I think everyone's a little bit afraid of. Is AI going to replace all of our jobs? And you think of all the youth going into college and starting to do their trainings of should I even get this degree? Is this job even going to be a job in three years or five years? So then those get a little tricky.
40
00:18:23.000 --> 00:20:00.000
Yeah, absolutely. I went down a couple years ago to California and there was a conference hosted by Google and Apple and ebay. And at the time they were requesting legislation around sort of the things that they were developing that could, within the phones, that could go to some of our countries that are enemies and then be used against us. So like if we were going to say, you know, you have the app, you know, you have Gemini or you have ChatGPT or whatever, could it get into the wrong hands of someone that says, you know, create a bomb that would take out Oregon, you know, and so there was this hold and this sort of black, you know, they didn't go into certain areas, it wasn't allowed to be sold into certain countries that were our enemies. So that they couldn't. So that their product, Apple and Google's products, so that they couldn't then in turn take this, the product that we had made this AI and use it against us. Easy to get around that for them. They can buy it on the black market, they can do all kinds of different things. But it was really eye opening to see the magnitude of what we're talking about and really how this can be detrimental and also all of the positive things that we can do with AI and just how massive this conversation is and how we handle it. Because their point was this is growing at such a rate that if we don't have some sort of boundaries around this, it will be too late. And then what? How do we bring it back?
41
00:20:00.000 --> 00:20:15.000
Well, on that point, Joe, if you were to ask our state government to make one or two laws that you think are critical right now, from what you're seeing, what do you think the first two laws should be around AI
42
00:20:15.000 --> 00:22:22.000
at the state level? The first one that comes to mind, this is a great question. The first one that comes to mind honestly is around education. I was at the town hall reading and really understanding English and beyond that, literature, reading lots of stories of quality that talk about the human experience and get one very literate. We're going to be talking to computers, you know, Oregon, suffering in our education performance as an educator in my life, there's lots of wrong with how we measure success. And you know, my kids have been in some private and some public and some really high end public education and they've seen different things to have different judgment. And it's like, you know, dad, in this one school system, if you have kids read out loud, it's just painful. Like. And they're talking middle school, early high school age stuff. They're like, they just, they can't. This is an area that's been allowed to be polluted by new thoughts about education. And it's like, this isn't hard. Guys. Teach kids the phonics, teach what the letters mean and get them to build up words. That might sound like a funny answer to your question, but at the state level humans are going to have to. We have the chance to either rise to greater states of sort of enlightenment is the best way I know to say it. Where people have landed on their values by age 18 and they've like worked out what's important and what works long range. But we're such a consumer video game. Like kids on the phones babysat via movies or the Internet, TikTok, whatever. We are missing the mark on creating this enlightened next generation that has brighter and longer thinking, more judgment about what's right and wrong than we've had. And that would be the first thing at a state level that I think people can do.
43
00:22:22.000 --> 00:22:33.000
And you're preaching to the choir. As a former first grade teacher who is all phonics all the time. How does AI play into that?
44
00:22:33.000 --> 00:23:58.000
The first thing I would say is it just emphasizes the need. A great thought piece that lots of people know is Wall E, the movie. And for some people that just sticks. They like the movie is like three quarters about the robot going around on the ground on a deserted earth. But one little part of it is about the humans who've left Earth because they messed it up. And then they're living kind of in a big spaceship that's like a cruise ship. Like they're all floating around in chairs and they literally can't walk anymore when they get back to Earth because their bones and muscles have atrophy and they're just floating around and drinking sodas and watching the next video entertainment and like they're doing nothing. You know, in a way we're heading that like that's what our culture is moved toward. We don't actually, we try to protect people from risk so much that, you know, kids don't ride bikes. I Had my kid riding to a dance class and I was so proud, like, you know, 90s dad. What's it have to do with AI is that we have such an open gate to going in that Wall E direction, going so much further than we already are. Social media, I think a great speaker on the topic these days is Tristan Harris, who used to be an ethicist at Google. Basically quit because they weren't listening to him and he was part of that fantastic documentary, the Social.
45
00:23:58.000 --> 00:23:59.000
Social Dilemma.
46
00:23:58.000 --> 00:24:01.000
Exactly. Like every American needs to see that.
47
00:24:01.000 --> 00:24:02.000
Yes.
48
00:24:02.000 --> 00:24:42.000
And he says, look at how we basically failed with social media. I think when you just come back to the idea of like something that's negative or a hostile action against your countrymen, let's say, right? It's like, okay, but we can make money by having kids get all this fomo. And oh, Julie got more likes than me from my photo. I guess she's prettier. And like literally young teenage female suicide triples. And we don't rush in with that with legislation. Our sort of enlightened humanity to me is where my mind goes with what our choices are.
49
00:24:42.000 --> 00:25:07.000
Well, I love the Wall E example and what you said about this opportunity for humans to rise and use something to make everything better. And we have all of these things, whether it's tv, cell phones, social media, you name anything, even a certain type of food, people can use it to make their lives better, they can use it to make lives worse.
50
00:25:07.000 --> 00:25:10.000
There's short term better versus long term better.
51
00:25:10.000 --> 00:26:04.000
So you just look at all of the things the same thing with AI, but it's just on a magnitude that we've never seen before where you're thinking at the best, oh my gosh, could AI cure cancer? Right, but then what if an enemy country has AI and destroys us all? I mean, you're looking at these extremes of the best of humanity, the worst of humanity. And so kind of as we were getting ready to wrap up here, and this could go on for so many episodes, right? And it probably will because we probably will need you back in a month because it'll have changed so much right then. But as you, Rep. Elmer, are looking ahead to this next long session, which feels far away but will blink and it'll be here. Then what is the next step? I mean, putting, I imagine almost everybody is thinking of some bills to put forward around AI, Is there one that you have in particular in mind?
52
00:26:04.000 --> 00:26:33.000
No, not yet. As you know, it's an election year and so most of the brain space right now is going towards elections. But we did just get out of a policy meeting this week to where we're starting to prep and look at all the different pieces of policy and, you know, each different area and what we want to make sure that we're touching on and hitting on and not duplicating and getting everyone going in the same direction. So starting those conversations. But I don't have anything specific yet.
53
00:26:33.000 --> 00:26:34.000
Right.
54
00:26:34.000 --> 00:26:48.000
What would you most want to see if you were in charge and it wasn't this whole long process of making a bill become a law and you could just pick a law tomorrow around AI? Is there anything that you would like to just make happen?
55
00:26:48.000 --> 00:28:17.000
That's such a big question, but it's a great question. I think it would be easy to pull on all of the wonderful things that AI can do and then use that to mandate that we use it so that we all can be the best versions that we can be. So taking your example within education, I mean, yeah, use AI to find out the best research out there, the one that gets the most kids reading in the quickest amount of time. That is time tested and it's true. And you implement that into the schools to bring those standards up and then repeat that in all the areas. Math, science. That's an easy one to say. We can see the good come out of something quickly using AI. Yeah, look at the healthcare. I mean I could ask it 50 things right now to please solve and if it could do that in the 11 seconds, think time that it takes. That would be amazing. But not just curing cancer and curing diseases, but also how do we handle the budget deficits and how do we handle Medicaid and Medicare and all of the questions around there and how do we get people to have doctors in a more timely manner? And can we streamline this not just within one state, but throughout the nation? And I mean we have telehealth, but what more could we do to improve upon our healthcare system and then go on to the next topic and the next topic. Yeah, for sure.
56
00:28:17.000 --> 00:28:29.000
That's such an interesting question. I wonder if you were to put in the most recent revenue forecast into ChatGPT or any of the other servers and say, what should Oregon do?
57
00:28:29.000 --> 00:28:33.000
Well, believe me, that's been done. And there are responses.
58
00:28:33.000 --> 00:28:51.000
Yes, that's so fascinating to see what could come by using AI with legislation. And I'm sure we'll have so many follow up episodes about what bills are being put forward, if they're making it through. Will things go to votes? Perhaps. And just.
59
00:28:51.000 --> 00:30:07.000
I will comment on that, you know, you asked about the revenue. That definitely has been done. You get an answer back. And half of your population doesn't want to hear that. And also the answer that you're getting that you are receiving back has bias based on the way in which you ask the question. And so from my perspective as a legislator, one of the I find to be extremely helpful is that my chat knows and will correct me. I don't know if that's real or not, but I feel corrected when I start to rant a little too much or I ask a question that maybe I didn't realize was so hardcore biased. And it will call that out because I've asked enough for information. And then I will say, now what would my Democratic colleague say to this? How would they poke holes in this? And then. And I'm like, whoa. And my eyes are open to my own biases or my own things that I was taking as fact and maybe they weren't fact. And then I will use it to fact check, well, is this fact or is this fact? And what are my sources? Show me this. So it's been extremely useful for me to truly look at legislation in a more unbiased way. And that's a huge positive for me.
60
00:30:07.000 --> 00:30:08.000
Yeah, that's beautiful.
61
00:30:08.000 --> 00:30:09.000
Wow.
62
00:30:09.000 --> 00:32:16.000
I mean, moving to reason is part of a thing I mentioned in the book. It's like we can't wish for a repeat of any previous decade that you happen to like, in American history. I feel like I went through this little microcosm in the Dalles, Oregon of just a beautiful high school experience. I had a great time, you know, a nice balance of getting in trouble and not. And like it was just a happy place to be. And that's not coming back. We have to think of something better than that. But just throw AI at it doesn't always handle it. But this that you're saying of like, okay, let's. One thing about it is it doesn't really have skin in the game except for to whatever degree that its developers want to keep you hooked. Right. I was reading the Claude instructions to this newest model that was leaked and I was reading it's very nice. And basically it said don't just ask them a question at the end to keep them going. Like, actually they had made the choice to tell their AI to not just keep you hooked. So they all sort of have that skin in the game. Like, how much do you stay on it and use it? But really, if they help with reasoning, it's not going to benefit the model. Itself isn't going to benefit or not benefit based on some organ statute that goes through. So it's something, you sit there and you talk out the pros and cons and you're thoughtful to have it, give it the counterpoint. That's probably, that's probably going to be in my next book. That's really brilliant that we reason out these things further and look at the long term. You know, people talk about second or third order consequences to a decision. The more and more we can do that, you know, your first grade teacher point of like, can't say mean things about each other. That's generally not helpful to society at large. And if we can, if we can like talk ourselves out of that in our little chatgpt, then that's a beautiful thing.
63
00:32:16.000 --> 00:32:35.000
Well, I just want to commend you for using ChatGPT the way you have been because truly I think there's a lot of leaders that they're going to use it a different way. The goal is how do I defeat the other side as opposed to help me understand the other side so that we can move forward all in the
64
00:32:35.000 --> 00:32:43.000
name of good governance. Maybe my idea isn't the best, but help me understand it so then I can advocate for it.
65
00:32:43.000 --> 00:33:02.000
I think that is a beautiful place to close. And thank you for your leadership and thank you, Joe, so much for taking time to come talk about AI with us again. I think we'll have to to have you back in the future as more bills get proposed and get your thoughts on that and more about your book. You want to hold that up one more time for everybody?
66
00:33:02.000 --> 00:33:03.000
Yeah, sure.
67
00:33:03.000 --> 00:33:09.000
We'll include a link in the episode notes for anybody who's interested in wanting to read a little bit more about AI and ethics.
68
00:33:09.000 --> 00:33:17.000
Yes, thank you. Yeah, I mean, the goal with the book is to invite everybody into the conversation because the future is what we will build.
69
00:33:17.000 --> 00:33:19.000
Right. And it's not going away.
70
00:33:19.000 --> 00:33:20.000
That's right.
71
00:33:20.000 --> 00:33:59.000
So let's not be the people in Wall E. Let's take a different direction and see what we can do with that. So I appreciate your time so much. Thank you both for being here and part of this conversation. And thank you so much, listeners for tuning in to this episode of Life and Legislation. Hopefully you found some interesting points and got your curiosity piqued a little bit more and that you're gonna follow along and hear what's going on in Legislation. So we hope you come back and join us next time for Life and Legislation with Lucetta.
