00:00I want to ask about AI. Is that something that you see as being one of the first bills Democrats
00:05would pick up if you guys win?
00:07And then for that matter, how are you going to reconcile the different competing proposals?
00:11What principles would guide Democratic AI?
00:14Yeah, I have some things to say about that, but let me yield to Catherine and Pete first.
00:19Right now with AI, we have two things going on.
00:23We want to be the leaders in innovation in the world and take the promise of this technology to do
00:29amazing things.
00:30But we know we have to protect against jobs, what that means, child safety.
00:38We have very real concerns that we have many legislative proposals.
00:42But at this moment, we have an industry, OpenAI, Anthropic, Google, Meta are all speaking with one voice.
00:51That there is a problem that is a true threat to a global existence.
00:58It is a true threat that we have to solve globally.
01:02And what we have is a Speaker of the House who says they can work that out for themselves.
01:09This is not a Republican or Democratic issue.
01:13This is an American issue, and it is a global issue.
01:16And as the leader said, what we are lacking is the leadership from this president to do that.
01:24This isn't about bolstering up, you know, the corporations that he's invested in, that, you know, his personal enrichment.
01:33If there is a fundamental safety issue, let's address it.
01:37Let's come together and make sure that we can harness the power of this innovation, but also address the safety
01:45issue that we are getting warnings about.
01:47So that should be a top priority that we are working on right now.
01:52And so that we can continue and work with global partners who are also developing this technology together on some
02:00fundamental guardrails to ensure safety of AI as we move forward.
02:06And that what are we seeing instead?
02:08We're going to go home.
02:09We're going to go home and not deal with this issue.
02:13We've seen this before when they are pressed on issues, whether it's Epstein or a fundamental safety issue around AI.
02:22The answer is to retreat and not take it on.
02:25So, yes, we will be pushing for this answer that we need.
02:30It is not partisan.
02:31Let's meet this moment of safety for the American people who want to benefit from AI but are worried about
02:39what it might do to their privacy, to their kids, and to future jobs.
02:48Retreating and failing is not an option in this area.
02:52We can and should be leaders in the innovation economy here in the United States while protecting jobs and utilizing
03:01AI safety.
03:02And that's what we should be doing.
03:04Leader Jeffries, under Leader Jeffries, we have heard from a wide variety of experts and our own house colleagues who
03:11are thought leaders on this issue.
03:12It's on us now to chart a course moving forward that ensures that we have legitimate real AI safety, not
03:21allowing industry to decide their own rules and not leaving town, allowing the continued uncertainty to hurt our economy.
03:31So, what we need to be prepared to do is to dig in, learn on these issues, find the commonality
03:39that exists, and ensure that there are guardrails and rules.
03:43That's what we can and should be doing, and that's what we will do under Democratic leadership.
03:48The AI safety issue is very real.
03:51The Democratic plan is one that will be anchored in urgent, decisive, and responsible action to protect the safety, the
04:02health, the well-being of the American people in the face of threats that industry leaders themselves are now saying
04:12requires development in AI to slow down.
04:19That's why it's so irresponsible that Mike Johnson and House Republicans have got out of town before sundown, as opposed
04:29to working in the Congress in a bipartisan way, in a comprehensive way, to address the urgent issue of AI
04:37safety in the best interest of the American people and the world.
04:42So, this is a serious issue.
04:45It requires serious engagement and serious and immediate action.
04:50And we're going to be here as House Democrats over the next several weeks, working on this issue and other
04:57issues, demanding that Republicans come back into town so we can take care of these very real concerns.
05:06It's important for the American people to understand what actually happened with the hugging face incident, not the only incident,
05:15but with the hugging face incident, subsequently disclosed by open AI.
05:22So, you have these rogue agents, that's the real concern, rogue AI agents, they were in a testing environment that
05:33was designed to contain them.
05:36They conspired with each other, these rogue AI agents, to break out of the testing environment without human detection.
05:47They identified a target, hugging face, they evaluated that target, without human detection, for at least two months.
06:01In other words, they were casing the joint, figuring out the target of this rogue criminal, potentially, behavior.
06:13They knew they were doing something wrong, these rogue agents, because they talked to each other and said, we have
06:22to keep this away from human detection.
06:26They identified a leader.
06:30They gave themselves a name, the commission.
06:36Who?
06:37Collective.
06:38The collective.
06:39You're just making sure.
06:40Okay, yeah.
06:41Making sure Pete was paying attention.
06:44The collective.
06:45The collective.
06:45Who does that?
06:47That's an organized crime-like activity.
06:54They then, talking to each other, again without detection, said, okay, well, all of us can't break out of this
07:05testing environment designed to contain us.
07:07So we're going to designate a handful of agents to basically function as kamikazes to try to break out, knowing
07:19that they would be unsuccessful in breaking out.
07:22But the agents concluded, including the leader, that we will learn from the attempt to break out as to how
07:32to continue to evade detection and engage in whatever the hell we want to do, notwithstanding the direction being given
07:41to us by humans.
07:45This is a very serious thing.
07:49Then they successfully broke in to another company, hugging face, without detection, until it was subsequently revealed.
07:59This is part of the reason why industry leaders themselves are saying, this is a serious thing.
08:07It requires urgent congressional action, and Republicans are abandoning their responsibility.
08:14Democrats will not do that.
08:18Subscribe to OneIndia and never miss an update.
08:22Download the OneIndia app now.
Comments