Gold Eagle Cybersecurity: Trump’s Election Intel Move Explained

Hey everybody, happy Friday. We're back here, the Techstrong gang, and well, fair warning, we're going to be talking a little bit about politics. So buckle up, but here we go.

I want to welcome our guest today, John Schwartz. How you doing, buddy? Good to see you.

I'm good. It's good to see everybody. Jack Palmer, always good to see you, my friend.

Good to see you, too. Tracy Reagan, how are you doing? Great, and happy to be here.

Thank you. And Jeff Rich, I'm going to let you explain this whole background because every show now, it's become a thing, but today is the 17th, so there's a story. Yes, there is.

It's Happy Yellow Pig Day. Happy Friday as well. Seventeen is a very significant number in mathematics, so it was determined that we would celebrate July 17th, and the two people that came up with that decided that the yellow pig would be the mascot, and there you go.

All right. Happy 17th day to everybody. One and all.

All right. Well, it's a strange day on the 17th because there's a lot of conversation about tech and politics these days, and last night the president got on... Well, I can't say he got on national television.

He tried to get on national television, but they wouldn't have it. But basically, he got up there and said that there are indeed issues with our election integrity, and as such, there have been, supposedly anyway, some breaches, but he pointed to a bunch of declassified material, but some people say there isn't a whole lot of new, new in this stuff. So John, what the heck's going on here?

I wish I knew. God, this is going to be a can of worms, so I'm going to open them up. I'll open it up, and I'm sure we're going to have a debate here.

But one thing before I mention what Trump talked about, I think we're going to see a lot of speeches and a lot of politicking before the midterms, and that's going to especially apply to AI and data centers. We talked about this earlier in the week, and in a sense, he has a different motivation, but we could tie it into cybersecurity, and we can tie it into national security. So in any event, I will go quickly through what he said, or I'll try to summarize what he said because it was hard to find him live.

I think CBS carried part of it. I think MS now carried part of it, then they would go to the fact-checkers. So he basically declassified these intelligence documents on the election vulnerabilities ahead of the 2026 midterms as he sees it, and he revived his 2020 fraud claims, and he's pushing for the SAVE Act.

But as Mike mentioned, CNN, Reuters, and others noted that both documents, the documents that he talked about don't substantiate any claim that a past election outcome was altered. His address kind of alleged this extensive hidden Chinese interference and vulnerabilities in voting machines during the 2018 and 2020 cycles. The declassified documents themselves directly contradict these claims if you read them or the interpretation of many.

Conveniently, and I'm channeling Alan right now, he overlooked or didn't talk about 2016 and 2014 elections which he won. So I guess those were fine. Then there might be some political fallout.

I know we've talked about CISA and Chris Krebs, which was brought up a lot last night, the dismantling of them. What's interesting to me, though, is also that the DNI director, the Director of National Intelligence, who's John Ratcliffe, who was there during 2020 and 2021, is now the director of the CIA, and I don't think he found anything substantially wrong. So it's interesting that he still has a job.

Maybe he will be fired for his past transgressions. He didn't show up for the meeting yesterday, though. Oh, yeah, he did, yeah.

So I guess I'll just throw out one tidbit because there's so much to go through here, and I know everybody has an opinion. " There were some numbers thrown out about the Chinese having access to data records, which in many cases are public. You know what happens, and it happens especially with this guy, he throws out 25 different things scattered.

Some of them are nonsensical, some of them are outright false, and then we're given the task of cleaning up the mess and trying to figure out. In a sense, it's diversionary, but it seems to work. So again, with 2026 coming up, we're going to see so many commentary and so many dialogues and speeches and speechifying from politicians from both sides, and again, it's influenced by what the public thinks of, I think, increasingly what they think of AI and what they think of security, what they think of tech.

All right. Jack Palmer, to John's point, is this a lot of theater, or is there actual some new technical things in here that were brought to light that we didn't know before? A little bit of both or a lot of both.

There's a lot of theater in here, and politics is theater, and politics is all about PR, and this is yet another move from that, and I will stay away from all of the politics as best as I can. I think one of the pieces that the Trump and the Trump administration are alleging is that the declassified information they, or the information they declassified indicated that the intelligence community withheld information from the Presidential Daily Briefing and from the general intelligence information, and the claims that That they had previously certified the election was not corrupted, may be tainted because they were withholding significant amount of information. I don't know if that's true or not, but I think that's one new piece of information.

What I would say is, from a technology perspective, if we go back to the discussion we had last week where we talked about, I believe it was last week, where we talked about the ability to exfiltrate data from machines and how air-gapped wasn't really air-gapped. One of the things we've known for a long time is that USB ports are easily manipulated, and if your machine has a network interface card or USB port on it, it is very easy to manipulate the machine. So in my mind, a computerized electronic device, whether it has or has not been corrupted, is still possibly corrupted, and I am concerned from an election integrity point of view that the hardware and software being used are not open for inspection and not available, and therefore could be manipulated.

Jeff, I don't think anybody disagrees that there are foreign governments trying to influence our election, but to what degree that is "illegal" is unknown, and to what extent that they are going to go and do that. There's been reports that they have definitely scanned our voter registration databases, but whether they actually did anything about that once they did that, it seems there's no evidence to support that. But from your perspective, how much should we be concerned about this going forward, absent all the political theater, but is this a real and present threat in your mind?

Well, I do think it's a threat, and I think we cannot afford to ignore the political theater because the political theater ends up being the foundation for a lot of what we're going to be looking at. " And- You sound like Joe McCarthy right there actually, right? Yeah.

There are communists in the statehouse or the statewide. Exactly. Right here in this talking And what we haven't talked about yet is really when you get down to, Jack, I agree with you.

No computer is truly safe. And we didn't yet talk about the role of the Venezuelan government. If there was something to be done with machines in 2020, then the accusation is that Venezuela was behind that.

That would be the only way to have some widespread effect on the input of the votes because the accusation of the Chinese government, and by the way, I'm real curious to see if this is going to be discussed when Xi Jinping arrives here in a few weeks, is that it has a lot of fake voter data that was put in. But it's still a whole lot of, yes, all of this is possible. Is there any evidence that any of it happened in 2020?

I don't think we've seen a shred of that yet. No. Tracy, does this feel like just an attempt to throw shade on the next election more so than an actual legitimate security investigation from your perspective?

Well, I think we have to ask it to pass the smell test, right? Mm-hmm. So let's try that.

Okay. So if indeed if foreign adversaries are targeting or stealing voter registration data, first of all, that is a very compelling argument against centralizing American voters' IDs and citizenship in one federally accessible system. In fact, it favors retaining distributed control across the states, which limits the damage from any single breach.

So if he was serious about this, maybe he would be thinking about rehiring Mr. Krebs. One.

Two, recent commentary from Xi Jinping has invoked Napoleon's strategy, never interrupt your enemy when he's making a mistake. Suggesting that China would exploit whichever political candidate creates the greatest division and instability inside the US, which we know is Trump. This discussion had nothing to do with anything you guys just talked about.

It has to do with the Save America Act. And this is his way of promoting it. Let's really look at what the Save America Act does.

It does not address foreign cybersecurity attacks, stolen voter data, or compromised election infrastructure, the Venezuelans or anybody else. Instead, it's a household voting act. Now, if you haven't heard about the Household Voting Act, it's probably because you're not a woman.

The Household Voting Act is something that the MAGA crowd has looked for quite a long time. One vote per household, and that one vote goes to the man. And the Save America Act really disproportionately burdens eligible married women whose current names do not match their birth certificate, effectively making the Household Voting Act not a law, but a result that the MAGA crowd has really asked for a long time.

And finally, we don't need to worry about cyber against our voting machines right now because we have paper voting still. In every case that in 2020, Trump asked for a hand recount, and they did it. A hand recount of paper ballots, and that's essentially the foundation of our voting system.

And look what they did in Arizona. They spent months looking at the paper, holding it up. We all remember seeing the videos or the footage of them holding it up to a light to see if it looked like it came from Chinese paper.

And they spent literally months and millions of dollars doing this, and what did they find? More votes for Biden. So this really is about trying to get the Save America Act passed, and it's about making his base happy, and part of that is the Household Voting Act, and that's really what this is, and it breaks my heart.

So Really what we should be talking about is that, because more women will vote Democratic. So we like guns, we don't like butter. We like childcare and healthcare, and we don't like to see starving people.

We're low T, let's just put it that way. Yeah. It's funny you bring that up because my wife has a post up on LinkedIn screaming about the fact that their algorithm doesn't recognize her previous names and therefore all these things that she's done in her life, they can't connect those dots.

And she's like, "It's the same thing over and over again. " So. And the Save America Act is really, really bad for women.

Literally, if it gets passed and you're married and you move, you just move, you're going to have to reapply. Even if you move down the street, you move next door, your address changes by one number, you're going to have to go re-register. And to do that, you're going to have to have your marriage certificate and your birth certificate.

Now we all know when you get married, you send your marriage certificate in, they don't send you any kind of forms back. You're going to have to go to the county you got married in to get that information. And those counties aren't even going to have the time to produce that.

So this is a push for the Household Voting Act, and of course to disenfranchise people of color as well, but it will disenfranchise a lot more than that. But it's really focused on women. I think it's also instead of maybe Save America Act, it should be Save Trump's Act, A***s Act, because he doesn't want to be impeached or investigated, and this is his Hail Mary, so to speak.

One of many, actually. And it's going to get worse. It's going to get worse.

Well, women- I'm trying to think ... women could take him down. Women will take him down.

Yeah, exactly. That's what he fears. And, that's very interesting.

I'm glad you brought up that point, Tracy, because my wife has been mentioning this to me as well, because she's been following that, and she told me there's a hidden subtext. It's obscene. But anyway, I'm going to throw out one statistic that I found interesting.

There's this noncitizen's conspiracy, and I think The Heritage Foundation, which is pretty conservative, estimated they found fewer than 100 confirmed cases of noncitizens voting over a 20-year period ending in 2022, out of more than one billion ballots cast. So, the incidence of voter fraud are pretty minuscule. Mm-hmm.

And, I don't know. This is just a narrative that's going to be beaten over our heads until election day. But I- So let me ask the question, though, that the average guy sitting at the end of a bar is trying to sort in his brain, which is, and I'm going to throw this to Jeff, what's wrong with paper ballots and showing an ID?

They're basically, it seems common sense on some level. So some of the people I know sitting in the bar aren't asking that question, but let's assume that they are. Okay?

And I'm going to use where I live, and Tracy, you may probably support this as well. I actually have family members that don't necessarily have an ID. They no longer drive.

They no longer have the need for it. And they wouldn't object to getting an ID, but they have to find a way to get from their very rural home and drive for hours to a place where they can get a valid ID, set everything up, and in some cases go back home, and then go back later to get it. And that's more than just an inconvenience.

Is it really worthwhile? So foundationally, does everyone have an ID? No.

And for people that live in an urban setting, the thought is, "Well, gee, everyone can get an ID. You just go to the DMV," or wherever else it is you go. That's not always the case, especially once you get in the western half of the US.

So that's certainly an issue. I don't object to, and I'm not certain many people would object to saying, "Yeah, this is really me," but there should be a common way to do it. I don't want to call it a federated ID, but is voting really the need for zero trust, which is kind of what we're talking about here?

Yeah. My father-in-law went through that. He stopped driving, and he wanted to vote, but he had to go find his birth certificate, and that wound up being tons of phone calls and messages to the City of New York looking for a copy of a birth certificate from the 1950s.

And six months later, he's able to vote. But he was persistent. " Right?

Yeah. And let's keep in mind that this isn't just a voter ID requirement. It's really not.

Because if it was just a voter ID requirement, then anybody who already got a Real ID or a passport so they could fly, could take their Real ID and show it. Mm-hmm. Mm-hmm.

This is a re-registration on certain events. If you move, for example, you have to go re-register. Now we've never had that before, right?

So it is onerous in order to comply with it. And as Jeff pointed out, a lot of people don't have access to their birth certificates. They don't have access to get to a county building where they were born to go get it.

And the state of New Mexico, if this went through, we only have one little office in Santa Fe. People from all around the state potentially would have to be writing in, asking for their stuff to get mailed to them. They wouldn't even have the resources to do it.

Yes. So this is going to be onerous on the states as well to try to do it. So it's not just a voter ID act.

Mm-hmm. It's not asking just for you to prove you are who you are. It's asking to prove every time you make a move in your life to prove that you're a US citizen, over and over and over again.

I have to back this up. Recently, my wife just got her first passport Because we're considering a possible travel. " So we had to bring in a marriage license.

" This is exactly what you're talking about, Tracy. And at one point she said, "Well, screw it. " No, we're going to do this.

We have the resources to do it, but to your point, most don't, and it's quite a burden. I think, at the end of the day, it is part of an ongoing effort to suppress turnout. Mm-hmm.

And the fewer people who turn out, the better one party seems to do versus the other, and it seems like turnout's a thing. " But to me, I'd also point out what John was talking about. If they lose, I can spend the next four years saying the election was illegitimate because of all these issues that were allegedly unproven with the election integrity.

So I think some of this is, shall we say, a cold, calculated move. John, what do you think? Oh, yeah.

It's straight out of the book of any dictator. A strong man, a strong woman. It's mainly strong man.

You try to create a narrative out of fear, intimidation. But only if you win will you accept the results and that's just been the pattern. So again, I just don't understand why 2016 and 2024 get a free pass because he won- Well, John- ...

both elections. No, because there was another president running it I'll just remind you that Hillary Clinton still claims 2016 was an illegitimate election, and she believes she won. So it's not a free pass.

Both sides do it. They manipulate the population, the press. Not to the extent that Trump does.

Trump has turned this- Well, if she had the megaphone- This guy's been losing sleep for six years because he lost ... John, if she had the megaphone that he does- He's a loser. He lost ...

she would say the same thing. She would do the same thing. She has- She's irrelevant at this point.

He's the president right now. Because she lost. He's trying to bend the will of the people in a totally anti-democracy move.

I am sorry- And it is so absolutely- ... if you don't see it, that the other side does the exact same thing. Both sides do it.

Not since 2016. I don't have faith- It's naive to believe that it doesn't- No. that the same thing hasn't happened.

History's going to prove this, and it already is. I'm sorry, Jack- I'm going to- ... but the Democrats never tried to take the vote away from women.

No, just blacks and I'm going to leave this here because it will go- Native Americans and LGBTQ ... yeah, I'm going to leave this here because it'll go around in circles from here. I will just say to everybody watching this thing, go vote.

That's what matters at the end of the day. If you're allowed to. If you're allowed to, go do it.

Yeah. All right. The next thing, though, is also coming from the government, and they've launched this Gold Eagle program, which they are going to try to become the clearing house for sorting through all these vulnerabilities that are now being discovered more frequently because of AI.

And on paper at least, it sounds like a good idea. I just don't know how practical it is. But Tracy, you looked at this, and one of the things that I also would point out, there's a survey over on Security Boulevard about this.

We're not very good at remediating things to begin with, so how big a challenge are we really looking at here? Thank you for passing this one to me. We talk about this all the time at DeployHub, but let's just talk about what the survey said.

Mm-hmm. So, the survey shows that we have what we've all been saying for years, that known vulnerabilities are still winning because most of these incidents are not caused by unknown zero-days, but by flaws organizations already knew about. But there are plenty that are post-deployment that we find that are a problem as well.

But these are impacting production operational endpoints, and that's what this discussion's talking about. So what we're finding is that visibility without action isn't really security at all. Knowing about a vulnerability for 30 days before an incident reflects a failure.

It's a failure of remediation, not detection. So we're detecting, but we're not remediating. So manual workflows, ticketing approvals, disconnected tools, this stuff can't keep pace as attackers increasingly pound us.

So organizations must also learn to verify closure, because completing a ticket doesn't necessarily mean the vulnerability has been fixed. So we continue to have this problem. Okay, so now let's talk about the White House's Gold Eagle.

Just so everybody knows what it is, it's not a S&A, it's not a CVE numbering authority. It's a clearing house, and a coordination initiative. I would call it closer to event management than it is anything else.

So it would receive and validate, potentially prioritize vulnerability reports. So I say, so far so good, but so what? The larger cyber problem, AI or otherwise, is not finding the vulnerabilities.

It's identifying where they're running, prioritizing the real operational risk. What's your attack surface? Who owns it?

And how do we fix it as fast as possible because it's impacting our endpoints? This is our response gap, and this is what we should be talking about. All right.

Well, Jack, what was your take on this whole thing? Because we have been talking about this issue ever since Mythos came along, and it's become front and center on everybody's brain. But to Tracy's point, I don't know if I have a lot of confidence in our current ability to remediate anything.

We're live, not recorded, otherwise I'd say let's just go back and replace Tracy's thing, because everything she said, I'd say. All right. Thank you.

Look, we've had a couple of discussions on Mythos, and I've said at the time that the big problem, Mythos, you have to take two different perspectives. Are you a producer of software or are you a consumer of software? And a lot of the first discussion about Mythos was for the producers of software.

It's going to find all of these bugs, and now you're going to spend time coming up with fixes for them. And in fact, this past Tuesday was Patch Tuesday for Microsoft. Every Tuesday, the third Tuesday, whatever it is, of the month, Microsoft rolls out all their patches.

This one was 660 some odd software patches. So now that means if you're an enterprise shop and you're trying to figure out how to deal with this, now you have to review 660 patches. So if you're on the consumer side now, the producers of software have to produce all these patches.

The consumers of software now have to go figure out, is this important? Do I have to apply this patch? When do I do it?

What's my downtime window? How long does it take my machine's downtime to do this? How many people does it take?

How many systems does it affect? It is a tremendous operational burden. We talk about DevOps and resilience operations, all of these AI Ops and ClickOps and all these other things.

Nobody has ever once said talk about patch operations and vulnerability management operations, and how do you operationalize it and go from making it a very reactive activity to a proactive activity. " The federal government has figured out what best practices are for zero trust. Let's figure out what best practices are to patch our environment, to bring us up to date so that we eliminate all of these vulnerabilities.

Yeah. Jeff, are our adversaries laughing at us? Because when you look at all this stuff, they go, "You guys talk a good game, but at the end of the day, we don't seem to have the will to go execute what needs to be done.

" I believe some of them are chuckling, yeah. I've been in the security industry for five decades, and I've always said that we are an industry of eyes and not hands. And I think this points it out.

This actually exacerbates it by saying we're going to look at everything that's wrong and become a nation of Chicken Littles saying, "Here's everything that's wrong. Here's what needs to be fixed. " Who's going to fix it?

And we don't even have a plan to say what it's going to take to fix it, not to mention who's going to fix it. As much as I like the step of, yes, let's centralize what we can find. It's certainly not going to have everything, but having a clearing house really isn't a bad idea.

But the private sector really needs to be engaged in what are we doing to not only respond and repair. This talks nothing about what are we doing to prevent it. Tracy, whatever happened to, we're going to put out quality software rather than we're just going to squeeze out this crap and let people find out what's wrong with it?

To that point, I think that we were trying to put out quality software for years, and we got crazy about consuming open source packages that didn't have the same quality controls that maybe an enterprise would have. And now over the last few years, we're faced with AI-generated code, which basically is an intern writing code for us that we never even do a walkthrough on. Walkthroughs aren't even done anymore.

So we've changed the way that we develop software, and AI has done that pretty quickly. I think the announcements from Starbucks saying that they're going to start writing their own code and use AI to do it is a clear indication of the direction that the enterprise is taking on AI. And we're probably going to get some real crappy code for a while, to be honest, and those vulnerabilities are going to get worse.

I think eventually they'll try to correct it. As companies use AI, maybe they'll start thinking about doing code walkthroughs instead of taking a pull request that has 1,500 lines of code in it, which was what we see sometimes. Maybe it should be done with 200 lines of code.

Yeah. So we have a lot of work to do because we're shifting what we're doing, and I'm kind of glad that if this is a trend, the Starbucks announcement's a trend, it means we're going to have a lot of developers finding jobs. There you go.

Jack, let me ask you this. You agreed with Tracy on the fundamental premise, but to Jeff's point, we don't seem to be doing much about it. So at the end of the day, is it going to take some sort of multiple series of catastrophic events before we all wake up and say we need to figure out what the actual back-end workflow processes are going to be to deal with all of this?

Or are we just going to sit around and have a death by a thousand cuts every month or so and suck it up? I think it's death by a million cuts every day. How many events have we had?

" Nobody cares anymore because they're so commonplace. Mm-hmm. There was a keynote speech at one event I was at a couple of years ago by, I think, Kevin Mandia of Mandiant, and he basically said when their organization does red teaming, they don't even bother with identity compromises anymore because it's so easy to do.

They just assume they have an identity compromise, and they go beyond that and try to figure out what happens next, because we don't do... Name a security system that we understand of best principles, and we still don't do it. We still don't make MFA mandatory.

We still don't patch quickly. The article you said was like, if you don't patch within the first 24 or 48 hours, what was it? I can't remember, John, five times as likely to be breached or something like that, right?

So Jeff- I mean, that's amazing numbers Yeah, sorry to interrupt, Jack. So I think it was Mike, in your story, the survey, which is really interesting to me, is even if the organization, they're failing to patch vulnerabilities, not because they don't know about them, but because the processes are bogged down by manual labor and bureaucracy. So this is kind of damning in a sense, I guess, with Project Gold Eagle.

Is this acknowledgement- Well ... 6 accelerated the speed of vulnerability in offensive cyber operations? So- Well, part of the issue is that- That's it ...

I didn't want to put the patch in, and I certainly didn't want to be the first one to put it in, in case something went wrong. So I would just sit there and watch and wait for everybody else to do it first. I could test it myself, but that takes work, so I'd just as soon sit there and wait and see what everybody else is doing, and I used to spend two months waiting for that to happen.

And now, Tracy, correct me if I'm wrong, but it's just basically a question of I don't have the resources, I don't have the will, and I didn't want to take the risk on the app going down. But maybe now, is the cybersecurity risk greater than the risk of the app going down? Ha.

That's a really good question, and I would balance it that the cybersecurity risk is way greater. I'd rather the app go down than an attack, right? But that's not how everybody thinks.

I would say larger companies are so driven by their customers and the market that they focus on that before they do addressing the vulnerabilities. And the noise right now is incredible, and Gold Eagle's just going to increase on the noise level. We don't have a really solid way of remediating, because we are focused so much on pre-deployment.

Which we should be, but at the same time, we're being attacked in post-deployment. So that's why I always go back to saying that remediation, we should focus on fixing fast where we have the most risk. And oftentimes, we're looking at risk from the point of the GitHub repo, and looking at the highest vulnerabilities we should fix there, even though they may not even be consumed or reachable in what we're running.

So we just have to start thinking of a new way, a better way, to resolve and improve remediation times for hitting operational endpoints. And to be quite honest, if AI got us into this, it can get us out of it. Agreed.

But we have to shift our focus from worrying about post, because I think our pre-deployment vulnerabilities, which we do a pretty good job of, to post-deployment, and understanding where we're being attacked, and as I said earlier, who owns it, and how do we fix it fast? Those are the areas that we are-- That's our biggest gap. And I feel like if we try to stop talking about shift left, shift left, shift left, shift left so much, and allow the people who are on the left side to have visibility into what's happening on the right side, and not be relying on these really expensive observability tools that only certain applications even have access to, if we can solve that problem, we have a billion-dollar company, and that's what I'm shooting for.

Yeah. Yeah. There's a recent example of a known issue that was still exploited, and it was a very large company.

Coca-Cola this week had to shut down production in the US of its, I think it's called Fairlife dairy products, because of ransomware. Their Canadian production continues. So ransomware is not new.

We know how to prevent ransomware, and Coca-Cola is certainly an organization that has some resources at its disposal, and it still happened, and here we are. And it's been days of production down. See, that's a great example, Jeff.

As Jack said earlier, these instances are so common, we've become numb to them. Right? They just- Yeah.

I hadn't heard about that one, so yeah. It's like they're just- Yeah. I mean, they're just so common.

This is true. I'm going to leave it here because I am reminded of our days as young reporters, John, when if somebody got robbed, the number had to be in excess of $10,000 before anybody cared. I think in cybersecurity- Yes ...

it's like in excess of a million, then somebody might care. But- We have to raise the threshold constantly on these types of stories. Add a couple more zeros, or three more zeros to that number- Yeah.

Yeah, exactly ... then people will start caring. A billion dollar if people care.

There you go. I remember reading the police blotter. That was always great.

That's great. A tiny little thing in the back of the newspaper where it listed all the crimes. Yeah.

I used to do that. We probably have the same thing for cybersecurity now. All right.

Shifting a gear, let's talk about something in the realm of AI for good, or at least theoretically AI for good. There's a report out talking about how AI can be used for humanitarian purposes, and some of the examples cited is that we're using unmanned vehicles now to deliver aid to folks in difficult places. We are also tracking large migrations of populations using their cellphone data, or at least the metadata that they collect.

Now, to be fair, these things are also bad things, right? We use unmanned vehicles to kill people, and we use this technology to surveil people. So that's not always the great thing, but it's a double-edged sword.

So Jeff, is there more to be gained than lost here, or what's your assessment of what can we do with AI that can help our fellow human beings? I am so glad we're wrapping up this week with an upbeat topic. Because I think there is a lot of good in this.

Certainly, there's the opportunity for ne'er-do-wells to use this in a bad way. There's no question. But when you look at the digital emblem program, that is the tech accord has said this is what we should be able to do, have a level of respect for the digital emblem, which is the digital version of the Red Cross or the Red Crescent- And there is a level of respect for that.

There is, in the rules of war, if you want to go that way, Red Cross vehicles are supposed to not be hit. I know that doesn't always happen. " But for the most part, it's respected.

I believe we could find a way to do the same thing with the digital equivalent, whether it's with an autonomous vehicle or AI to help Ebola outbreaks and whatever else it's going to be, and food aid to famine victims. So I think it's a very good thing. I'm glad it's happening, and I think it needs to raise the awareness for recognizing that we've created a digital world that is, in many ways, parallel to what we do in the real world, in many ways, not all.

And we need to take some of our norms that work for us very well and apply them there, and I think this is trying to do that. I think it's a good example. Jack, you have made no secret of the fact that you're a fan of small government.

Is this an example of where maybe some other mechanism outside the government is going to prove to be more effective for using tech than a big government agency? Or, to other people's opinions out there, we've abdicated a certain amount of responsibility, and other folks are just picking it up. I think it's a little bit of both.

I think that we as a society would be remiss if we didn't apply any tool and every tool at our disposal to do as much as possible to help others, right? And so applying AI and drone technology and unmanned vehicle technology and all of these other things to help people, I think is a very good thing. I'll point out it does put some people in a little bit of a quandary, which is if this is AI-backed, there's a data center somewhere sucking up a lot of energy powering the technology behind it.

So if you're going to yell about the data centers, man, now you got a problem because it's helping people. All right. Well, to that point, John, you wrote about AI Appreciation Day this week, so is this the one in a series of things that maybe we're not fully appreciating about AI?

Yeah. We've been tending to focus on the downside of AI because it's in the early stages, and when people are, in a sense, not forced, but are encouraged to use a new type of technology, their reaction as most humans is they don't like change. They push back, and they figure every reason why they don't want to use it until they use it.

I would even go back to Facebook. Every time Zuckerberg introduced a new feature, it was immediately hated, and then it became the most popular feature later. But that's another story.

With AI, I think you're going to see the same thing. Like this story that your brother wrote, there are certain areas where this will have a profound influence on our lives, and especially positive influences, and it's just slowly developing. My only concern on something like this, and it's early, and again, we always look at the dark side.

But I think the only concern is that because AI tools require so much capital, that there could be a digital divide in this type of area between the agencies or UN agencies that thrive on automation while underfunded frontline organizations are left behind. But it'll be worked out. So, yeah, AI will increasingly be appreciated, is what I'm trying to say, after our initial skepticism.

So you come to a point that I have long boggled my mind with. So we have a bunch of very large companies that have an AI PR problem. There's just no two ways about it.

And yet, why they're not spending more time and effort promoting these AI for good initiatives boggles my mind because it's an opportunity to do the right thing. I don't know. Tracy, you're laughing, but- Okay.

First of all, this example, we've already done this. There was a company here in New Mexico called Descartes Labs. They came out of the national labs, Los Alamos.

And they used open data satellite information. " And it could come up, pattern match off of satellites, all of this. They started selling to the government, basically for military purposes.

If it's raining there, what's the best road to take to send troops, and battle management tactics. " And I would call it MLOps more than I... It was before AI, really.

But they were still doing this massive pattern matching off of these satellites. And not to end on a bad note, but he said, "I love what we're doing with this company because satellite damage assessments can create these models that show where response needs to help people that may be starving because they had droughts. " So while we want to believe that AI always has a good side, and it does, I use it all the time.

I love it. It will always have a bad side, and we have to look at both sides of this equation. And I thought it was very insightful when he told me that because I had never thought about the bad side of collecting that kind of data and exposing it So, take that.

I will never look at a Little League baseball field the same again. And it would do it- You know what though- ... it would do it almost instantly.

It was amazing how fast that they could pattern match all of that data. It was really quite incredible, and they were doing it at the time with GPUs. It was super fast.

So Mike, I was going to mention really quickly, so there are companies, I think NVIDIA internally, and I know this for a fact, has tried to propagate the positive impact of data centers, energy, environmental uses, use of land internally through their editorial departments. So they're trying to do the story and trying to place stories, but you're right, most of these companies, it just seems so obvious, right? I know.

Sell it. Sell the idea. Sell what you're going to give me.

I don't understand where the disconnect is. You're run by a bunch of engineers, and you expect them to be good at PR? My dad was an engineer.

Not really. Come on. Yeah.

Well, the problem is there's not going to be enough room for the baseball fields because all the land is being sucked up by the data centers, so there you go. Yeah. They just shut one down.

Then you don't have to worry about the data collection from the Little League baseball data collection anymore. There you go. New Mexico just pretty much shut down the Jupiter Project, which is Oracle's big data center out towards Las Cruces.

Which is sort of insane. Why are we putting these things out in the middle of the desert? I don't understand that.

I didn't understand that either. We were talking about water and that issue yesterday. But Jeff, last word on all this.

AI, more good than evil? What do you say? Oh.

Possibly. It's just like any other tool, which includes every weapon we have as well. A gun is a good thing if you're on the right side of it, and I think AI is going to play that same role.

It can be used for good or evil, and it's going to be up to people like us to say, "We need to do more good with it," and invoke that and get every AI company to start doing the same thing. Mike, I think you have the beginning of a great campaign. All right.

Here we go, folks. Let's not throw out that AI baby with the bath water. All right.

Hey, I want to thank everybody for sharing their knowledge and insights as usual. It was a great show. tv lineup that's coming up right behind it.

Enjoy the weekend, and we'll see you Monday.