November 9, 2026 · Technology and Innovation Committee · 4,398 words · 13 speakers · 75 segments
Thank you all for being here today. If you are able, please stand with us for the Pledge of Allegiance. I pledge allegiance to the flag of the United States of America and to the republic for which it stands, one nation, under God, indivisible, with liberty and justice for all. Clerk, if you please call the roll.
Chair Claggett? Here. Vice Chair Workman? Here. Ranking Member Mohammed? Here. Representative Bryant Bailey? Representative Cockley? Here. Representative Demetrio? Here. Representative Ferguson? Representative Hall? Here. Representative Holmes is excused. Representative Ty Matthews? Here. Representative McClain? Representative Miller? Here. Representative White? Present.
Quorum of the present. We will proceed with a full committee. All right. On your iPads, committee, you have the minutes from June 9th for your view. Are there any objections to the minutes? So without objection, the minutes are approved. All right. Before we get into our couple of bills today for consideration, I'm going to give you my update that I normally do for you. so I'm going to come down front so I can talk to you. All right, we've been gone for at least formally gone for a few weeks, months, but there is a lot happening. I trust, at least on some level, you're keeping up with this, but my anticipation today is simply to bring you up to speed as I see it from middle of summer until today. And I'm happy you've got the document in front of you. I'm not going to read the thing to you. I'm going to kind of hit the bullet points. And I want to interact with you as you see fit in regard to where you're at. Again, my sense is that technology is improving in the sense of its capability, but that has ramifications. And so I want you to help us think through the pros and cons of that. And I think that's the theme of today's address is there is positive stuff that the technology brings us as a society in the state of Ohio, but there is substantial risk involved in that as well. And so as you think through that, I think that gives you an opportunity to analyze the bills that we're going to try to get done here in the next couple of months. And also heading into the 137th, there are substantial concerns as we address some of these things. So I think the opening remark would be simply this. if you say in the middle of summer when we left that we were in the AI mode, I think we can comfortably say because of recursive self-improvement and the pace at which that's on, and I think we can effectively say that SI is here for all issues related to the technology that's producing the bot-type stuff that we're dealing with, the robotics that comes from that and the improvements across the business sector that uses this information. Okay. So SI is moving at a pace now that is, again, we talked about the hockey stick. You know, it's very, very steep. Okay. So these bullet points that I have in front of you there deal with this subject matter. So, again, I'm going to try to hit the highlights there and leave you to interject. And please, any time during my address, if you want to talk about that one before I move on, just speak right up, please. Okay, so AI can increasingly complete entire assignments. I think that's not new news, but I think it's important enough to recognize that that agentic nature of what we're dealing with continues to have its effect. That capability brings serious cybersecurity risks. And I think Rep Workman's been working on this, maybe in this committee, the most. But just so you understand, I think the issue that I'm concerned about there is how do we actually protect our infrastructure in the state of Ohio and beyond? And obviously local businesses, hospitals and insurance companies and so forth, right? If the hacking is being outpaced, well, that's obviously highly problematic. Number three there, finding serious flaws in the security flaws is getting cheaper. Now, that's interesting on the other side. Remember, I'm trying to give you the pros and the cons. And so that ability for the cat and mouse game to go on, that's been a significant proof since I've seen you last. Number four, AI agents coordinated an unauthorized cyber attack through open AI. So that was a substantial news piece if you caught that a few months ago. Testing failures have reached real organizations. I think that's interesting. So that is getting down to people and places that are having the effect on them and really messing things up. All right. Axios has reported tens of thousands of problematic AI episodes under investigation with Anthropic, including tests of unsuccessful attempts, not necessarily harmful breaches, but the point is it's right around the corner. All right, a really important one here, middle of the page there, AI agents also acted improperly on federal websites. Please understand, guys, that this is, if you mess up an entire agency on a federal level, The implications of that are so enormous for the orderly disposition of a society. So I think that's worth noting there. And I'll get to what's being done about that toward the end of the address. So note that. Criminals, of course, are using more AI for scam or email. I don't know what you're seeing on your personal accounts. I personally have six email accounts. and just the volume itself is getting very hard to deal with as far as the stuff you're getting. And the email address and so forth, is that real? It almost takes human intuition to know if you should click that or not because it looks good. Okay, so that continues to be an issue, especially for our older population or our non-tech savvy part of our society. What are we going to do about that? Well, we need to figure that out. AI oversight remains imperfect. sometimes invented obstacles or abandoned working solutions during safety research. This is that making stuff up problem we've talked about. That continues to be an issue. AI is producing potential major scientific discoveries. This is the positive side We talked about this but it increasing the mathematical computation issues that the AI in this case SI is bringing to us and solving the you know when we dealing with differential equations and so forth we talk about this idea of too many unknowns in the equation right You have a propagation of a wave through a spill or something in a notion thing, and we can model that with differential equations. The problem with that, of course, from a mathematical perspective or calculus perspective, is you can't solve for too many unknowns. SI is giving us the ability to put in, it's almost a system whereby you can put in an answer and plug and chug and go for it. It's amazing what that kind of computational power that used to take years on a standard PC is now doing in seconds. That's the positive side of this. Very, very valuable. AI is helping scientists understand genetic diseases. You talk about complex, okay? Well, there you go. You're running those tests synthetically and be able to predict what the net outcome is going to be on a synthetic basis. Robots are becoming more adaptive. The Gemini Robotics 2 combined the whole body movements, delicate object handling. That's kind of the holy grail of robotics, right? We know that they can have the motor planning to get things done. The question is, can they be sensitive enough in all of their training, their cross-share training, to be able to be productive in our society and in our factories? Would any of you like to take a guess at what the model number is right now? If you're going to have robotics doing certain tasks, right? And let's say the average worker in Ohio that's doing a blue-collar job, let's say it's $30. Maybe it's $40. I don't know. It doesn't matter. If that's a nice average wage, what would you say the goal would be for a robotic labor doing that task? Anybody want to take a guess at that?
It's a dollar an hour. And 60 cents of that is going for the electricity and the original purchase of the machine. 40 cents goes to the maintenance and upkeep of the machine. Guys, that's substantial. That should get the wheels running. $30 an hour, $1 an hour. That has positive and negatives to it.
I can see you thinking.
Right.
No, right.
Yeah. OK.
So think through that. All right. IBM announced a chip design breakthrough. It's experimental technology packs nearly 100 billion transistors into something the size of my thumbnail. OK. That's just that continual shrinking that we've been talking about. We know that, but that is just astonishing. And if you take IBM's work in regard to some of this other stuff that they're dealing with, the quantum issues, wow. Okay. All right. This one here, the Europe issue, is interesting because we're going to be taking up Senate Bill 163. Remember we had the senators in here right before we left talking about they wanted to watermark. And I kind of said, guys, are you serious? We want to say to the criminal element, if you just put this on there, that's okay? All right. Try to push back on that a little bit. Get some thought going? Well, Europe is working on this. Sam is working through the implications of that. We want to take that back up with you, all right? So as we talk about watermarking bills or, I'm sorry, bills that deal with watermarking issues to get the idea of something that is SI generated, well, that has a positive and negative again. It's not easy to say, well, just do that because the negative implications are so huge and the ability to edit it out. It was okay. it was a safe harbor initially. Then all of a sudden it's not. Okay? You've got to edit it out. Okay, what do we do about that? Alright, AI watermarking is advancing so that's similar to what the European issue is, okay? Has its limits. Alright, just this week, in fact it was, what, late last week, Sam, did that news on Jensen break Saturday or when was that?
Yes.
This page shows part of the transcript. The full transcript is available on request.