On this edition of Wall Street Week, Former US Treasury Secretary Lawrence H. Summers says that the SEC and exchanges should look into the surge in the VIX. DWS Americas CIO David Bianco says that the spike in volatility is evidence of investors overreacting. Harvard professor of economics Greg Mankiw says that the
US is leaving the next generation a negative bequest. WestExec Advisors co-founder Michèle Flournoy says that there is already an AI arms race underway, and SandboxAQ CEO Jack Hidary explains why the larger economy needs large quantitative artificial intelligence models.
This is Bloomberg Wall Street Week. The global push into infrastructure, breaking the IPO logjam in text, the financial stories that shape our work, cutting inflation without losing jobs. Do we need rate cuts and if so? How many? Investing in a time of geopolitical.
Turmoil Through the eyes of the most influential voices.
Ten Rogueff economists of Harvard, former FDIC had Shila bet Ge CEO, Larry Culp, San Francisco fed President Mary Daily Bloomberg.
Wall Street Week with David Weston from Bloomberg Radio.
Market turmoil, Middle East threats, and the presidential race is off and running. This is Bloomberg Wall Street Week. I'm David Weston. This week, Harvard economist Greg Mank on facing up to the federal deficit.
You can't leave your children a negative bequest unless you're doing it through the government.
Former Pentagon Number two Michelle Flornoy on what's needed for the US to face a more dangerous world.
We really have to take a fresh look at what we're spending on and where we're investing.
And Jack Hittery of Sandbox AQ on what's real and what's hype in the world of artificial intelligence.
This is a great moment for a I that takes us beyond the chat GPT.
We start with a wild week in the markets, stocks and bonds responded to weak jobs numbers, a shift in Japanese monetary policy, and geopolitical uncertainty. To take us through what it all means, we turn once again to our special contributor, Larry Summers of Harvard. So, Larry a lot to sort through here. What was your overall take on the terminil the marketplace and the speculation, the speculation that FED maybe should have an emergency rate cut.
I would shay on current facts, given that there has been some recovery, given that volatility has.
Come way down.
It were not out of the woods. You can't be certain. The FED certainly needs to be watching carefully, but I think an emergency response would be launching panicked, overheated and counterproductive on current facts, Which doesn't mean that the FED shouldn't be watching very closely. Before they have another decision to make in September, there's going to be a lot more data that it's going to come in, and I think they should make clear that they're going to be watching all that data and they're going to make a decision reflecting the need to balance the concern of making sure that we have stopped inflation with the concern of maximizing employment, and I think they need to not be pressured into specificity of a kind that is impossible.
Let me pick up on one of the things you mentioned was volatility. The VIC really did spike up, I think over sixty, maybe the highest it had been since the pandemic first hit. How much attention should the FED pay to that volatility index what it might be telling you about the markets are functioning.
I actually think that the SEC and the relevant exchanges may want to pay a bit of attention. My understanding is that because there are some ill liquid instruments that go into the calculation of the VIX, the VICS had a somewhat artificial movement on Monday that if one looks at the vic's futures, which are a somewhat different instrument, they the movements were much much less drum. So I certainly had my attention caught by the vics early in the day on Monday. But as I have looked into it, I think you were learning more about issues around liquidity in the options markets than you were about some profound reassessment of the American economy, and since that is so widely watched an indicator issues of liquidity. Issues around how it settles, I think should be studied by the relevant parties in the industry and the regulator, the SEC. Larry.
We spent much of the week with various people getting advice to the FAT. It included our former President Donald Trump on Thursday and a long news conference and which he addressed it, reiterating he thinks the president should have some say over monetary policy and rate setting, including saying, boy, he'd made an awful lot of money and he thinks he knows better than the chair of the FED. You had a reaction to that. I know you posted on that question. Give us your stance about how dangerous that could be.
I guess I can't say I was surprised by ex President Trump because he had said things like that before, but I sure was appalled at.
How bad an idea it was.
I mean, start with the preposterous arrogance. The Central Bank has nineteen members of the FOMC who spend more or less all their time scrutinizing every economic statistic. President's got a lot of things to do at any given moment, and he's actually much less close to the economy. God knows, the skills associated with being an economic forecaster and the skills associated with being a successful real estate operator are very, very different ones. So I don't think there's any particular reason to think that a president of the United States would have a intellectual contribution to make. And the reason why countries all over the world this isn't just.
An American thing, but countries essentially.
All over the world have moved to independent central banks is that they recognize a profound conflict of interest.
Who's ever an.
Elective or political office always is tempted to put more money lower Intereststrates hit the accelerator hard to get a boost to the economy to make people feel good, But when everybody sees that coming, it doesn't actually make people feel good. It just raises the expectation they have for inflation, so you get more inflation and you don't get any substantial output gain.
Okay, Larry, thank you so very much once again for being with us. That is our special Wall Street Week contributor. He is Larry Summers of Harvard Markets went on a wild ride this week in the wake of the boj moved to titan and the week US jobs numbers from last Friday, but after gapping down when trading began the week with the VIC shooting up to over sixty, things settled down a bit. The S and P five hundred ended the week at fifty three forty four, just about where it was at the end of last week and only slightly below the median year end number for our Bloomberg elves of fifty five sixty eight. The Nasdaq hit a similar air pocket Monday morning, but ended the week up two point six percent from Monday's lows, again just under where it ended last week. The yield on the tenure in the meantime, added about fourteen bases points over the course of the week to end at three point nine to four percent to take us through the week. In the markets, we welcome back now, David Bianco, DWS America's Chief investment Officer. Always great to have you here, David, and this is a perfect week to have you because where there was a lot of action in the market, what did you make of it?
It was wild indeed for in August, and even though the S and P is back to where we were last Friday, let's just for a moment talk a little bit about what the causes were for the volatility around the world. As said, a lot of conversation around the Bank of Japan hiking, poor timing around the hiking, and the unwind of the end carry trade, this legendary en carriage trade. It's real, always difficult to know just how big it is. But what happened last week was that not only did the BOJ hike, the FED put on the table the possibility of a cut in September. But then, don't forget, we had that jobs report on Friday, and the jobs report had a jump of further climb in the unemployment eight to four point three percent, and it triggered what a lot of people are calling the somb rules, some indication of a recession in progress. The job creation numbers were still good one hundred and fourteen thousand on the payroll survey, but a lot of panic broke out, particularly in the bond market, that maybe the Fed's way behind and maybe emergency cuts are necessary. So just as the BOJ what I think they could have done this earlier in the year, decided to do a hike, last week, you have the FED and then data suggesting the FED maybe cutting a lot. It was kind of crossing the streams. It's really dangerous across the streams to major central banks moving in different directions, and that sparked ye appreciation people trying to probably pay back some of their yen loans. That's part of what created all the volativity and made markets, particularly Japan, fall out of bed on Sunday night, Yet appreciated. Japanese stocks down more than twenty percent on one point, led by the big banks. It hurt other global equity markets on Monday. But we've recovered and there's more going on though than Japan during the week and last week.
But David, how much did you think was the markets and how much was the economy? I mean, you referred to recession possibility, and some of the banks started saying, well, we're taking up our probability of recession, something that last year we thought was a real probability, Now not so much. Where are you on the probability of recession?
I think we're still a safe distance from recession. We do dbs. We expect the economy to slow further, but I don't think we're slipping into a recession. Not right now what happened this week late last week, I think is more about investors overreacting, sometimes putting words into policymaker's mouths about what they may be doing, or maybe just overreacting to step in one direction how much further you're going to go, whether it be.
Hikes or cuts.
But there was this background of other concerns. We've had geopolitical risks, We've had investors take an additional concern with what remains a week economy in China and a week manufacturing economy in Europe and still weak in the United States outside of data centers and chips, and then an earning season. Well, the earning season is basically done. Ninety percent of companies have reported, and while the results of coming okay, we're still hearing from the non tech companies that businesses sluggish, manufacturing, consumer businesses. They're cautious about the outlook as well. And I would just say, even though we've got ten percent year on year growth at of SMP earnings about sixty dollars a share in the second quarter, it's all still coming from big cap tech.
It doesn't help it's in August and so the markets are not as liquid as we normally expect, right, But what are you expecting for the rest of the year, given where we are in earnings right now, all the factors, where do you think the SMP five hundreds headed for the rest of the year.
Well, I think market's going to stay volatile. Markets can be valuatle particularly any day any week. By the way, options and futures markets can be even more volatile. I think the volatility is with us. We've been shaken by some of the events that that broke out, and I think that those risks and those concerns, particularly the election and the uncertainty for what that means for taxes and tariffs, and I think investors want to see some election outcome that keeps taxes low and prevents tariffs from going higher. But you know, there's a lot of uncertainty around the elections, and I think markets will stay volatile at least until then.
David, thank you so much for being here. That's David Bianco of DWS coming up. Despite the market volatility, the US economy continues its relatively strong performance. We ask carbon economist Greg Mank for his assessment of the economy and what the next president needs to do about it.
I think we need to at some point move to a sustainable fiscal policy.
That's next on Wall Street Week on Bloomberg.
This is Bloomberg Wall Street Week with David Weston from Bloomberg Radio.
This is Waltere Week. I'm David Weston. Jobs numbers may have come in weaker than expected, and some consumers spending may be softening, but overall the economy remains strong and inflation is moderating. We talked with Harvard economics professor Greg Mankew at the Aspen Economic Strategy Group meetings about how the FED is doing and what comes next.
I think we're there.
I give FED low marks early on because I thought they were slow off the mark, but then they once they realized the problem, they reacted vigorously. And I think we're basically very close to target now. And if you look at the measures of inflation, a big part of that is shelter, and shelter we know is measures a lag. If you look at the private sector measures of shelter inflation, rents.
Are basically flat. Now there's no inflation in that.
So I think the overall measured inflation rates we're get to CEPI say that's going to be coming down to target in the next six months.
So the economy is doing well by most measures. A lot of people want to move to our country, right, they want to invest in our country, and then a lot of people are very unhappy with the economy. How do you square those two things.
Well, the pandemic had a strange effect on people's finances because if you think back to the early pandemic, we were sending people lots of checks, whether it's just general stimulus checks or expanded unemployment insurance. So people were getting lots of income, but they couldn't spend it, so they're saving a lot. They are paying back down their credit card bills. So even though people are kind of unhappy being stuck up in their houses, their finances are pretty good. So let's look where are now. All these stimulus checks have disappeared. People are spending more and so their credit cards bollses are building up, credit card delinquencies are arising a little bit. So I think, if you're comparing where we were to twenty twenty, well better off because we're not in a pandemic.
But people's finances are actually slightly worse.
There's a lot of talk about when the federal cut how much we will cut? Put that to one side, Where do you think we will end up? Where we end up through this cycle, Because there's a debate really on whether we're going to go back down to really low interest rates or whether there are structural factors that will keep them elevated.
I don't think we know, and I think the Fed is going to have to play it by year. I think there were good reasons why we had a thirty year decline in real interest rates prior to the pandemic and the recent recent events to what I sent, all those forces still in play, and to what I sent, they're new forces, like very big budget deficits that are going to keep them straights high. I don't think we really know yet. Partly it's going to depend on future policy, and the next present is going to have to make decisions over taxes and spending, and so I don't think we really know where the FED is going to end up. It's probably lower than it is today, but probably higher than it was before the hiking cycle began.
You mentioned the debt and the deficit. How big are problem is after the economy?
Well, I think there's two problems with the debt.
I think this is one of the ordinary problems, and the extraordinary problems. The ordinary problems are basically passing a debt onto our children. You can't leave your children a negative bequest unless you're doing it through the government, and that's what we're basically doing because we're leaving our children in negative bequest by running up the debt. So they're going to face higher taxes that will encourage some crowding out of private capital, reduced activity growth. And then this is the extraordinary problems of debt. You know, could the United States turn into Greece or Argentina where there's a fiscal crisis. The markets don't think there's going to be one, that they're expecting us to be responsible. I hope they're right, But the laws of economics that played out in Greece and Argentina could happen in the United States. And so I think we need to at some point move to a sustainable fiscal policy.
Even short of Greece or Argentina. At some point, does the deficit problem curtail the fedibility actually to control the economy, because at some point the markets take over in setting rates from the FED.
Well, the FED would have to set higher rates.
I think that I'm not really worried about the FED losing control sort of a fiscal crisis. I'm not worried about the FED losing control of the economy, but it doesn't mean the FED is going to have to place higher rates, and that could be a problem for fiscal policy. These higher rates put pressure on the budget deficit, and as a result, the tension between fiscal policy makers and monetary policy makers could be exacerbated.
How we defined debt crisis? What are the prospects for the next president, whoever it is, we'll have to deal with the debt crisis. And I guess that really is bond vigilantes, right.
Well, it is exactly. It's bond bond vigilantes. And I don't think we know it's partly psychological. At what point do people become scared. I don't think it's gonna happen soon, but I think at some point in our lifetimes the current policy has to be changed, because if you look at the CBO projections on their current policy, that GDP ratio is going up to infinity, and we know that's not gonna happen.
So at some point that's gonna change.
It's probably not gonna change the next few years, but it's something we're gonna have to deal with, and how to deal with that does not be an easy political problem.
What about the election coming up? In so far as we know anything, we have I think a sense maybe if we're a second president, Trump would take us. I don't know how much we know abou Kamala Harris, but what are the alternatives? You see them?
Well?
One of the big issues for me is the role of the United States and the global economy.
I know it's a sort of a bad word these days that I'm a globalist.
I actually believe that integrating the United States with the global economy is a good thing. So I'm in favor of free trade agreements. I'm in favor of more relaxed immigration rules. This is a sort of mainstream view back in the Bush and Clinton years. Now is a sort of backlash against it. I think that backlash has been is ill informed, and I think at some point we need to go back to being a leader in the global economy, which I think is good not only for the rest of the world, but also for the United States.
What about US relationships with China, when it comes to the economy.
Well, China is the big challenge, not not so much economically, but I think politically and especially the threat in Taiwan. So I do worry a lot about sort of what's going on in that part of the world. I understand I don't really like industrial policy in general, but I understand the motivation had the Chips Act because almost all the high end chips are coming from Taiwan's semiconductor and that's a very vulnerable.
Part of the world.
So relationship with China is very important, but it's not so much an economic relationship as a geopolitical relationship, and that's the part we need to negotiate.
That was Harvard economics professor Greg mank. Whoever wins the US presidency in November, they will confront geopolitical hotspots around the world, from the Middle East to Ukraine to the South China Sea. Michelle Flornoy was the Deputy Secretary of Defense under President Obama, and when we sat with her at the Aspen Economic Strategy Meetings last week, we started with what the Defense Department needs to do to be prepared for the range of challenges it faces.
I think the United States you know, we're a global power. We have interests, very real interests in many parts of the world. So if we get engaged in say deterring a Chinese attack on Taiwan, or deterring Russian aggression beyond Ukraine and to NATO, we have to also be able to still have presence and deterrence ability elsewhere so that the Middle East doesn't also erupt. So you never want to be in a situation where your engagement in one region basically provokes or encourages another adversary elsewhere to start to take advantage of that and start something else. And so you really do need to have the ability to look across multiple theaters at a time and to be fully engaged in one while deterring in the others.
You mentioned some of the relationships with allies. Obviously we have NATO when it comes to Europe, and now there's more and more development in Asia, including these talks among South Korea, Japan, and the United States we just have happened. How much can that relieve some of the pressure on the United States itself?
First, it's extremely important politically and in terms of demonstrating resolves. So if you're trying to affect Hijipang's calculus about whether to use force in the region, knowing that he won't just have to deal with a US response, but a Japanese response, a Korean response, an Australian response, and others in the region who you know coming together in coalition to try to protect the rules based order in Asia. And then these allies increasingly, particularly Japan, Korea, Australia, they have real capability to contribute to any sort of crisis situations.
There are a lot of fiscal pressures in the United States right now. At the same time, many people think we're going to have to commit more resources actually to defense, given some of the things you've just talked about. Larry Summers said that repeatedly on this program. We're looking forward, we're going to spend more on defense. Are we spending enough? Do we need to spend it more and if so, how much more?
Well? I do think that given the way the world is evolving in our interests, we probably do need to increase our investment. But just throwing money at the problem is not going to be enough. At the same time, we really have to take a fresh look at what we're spending on and where we're investing because in a lot of these situations, you know, buying more of the same thing that has served us well, you know, in a force that was optimized for the Middle East and counter terrorism is not necessarily what we need to deal in a maritime and air theater predominantly in the Asia Pacific and elsewhere. Plus you have this profound period of technological disruption, and so so you really have to look at, you know, how do we adopt innovation, integrate it into the force to enable the legacy forces, we have to be able to operate differently and get different outcomes. So it's innovation adoption and it's new operational concepts. We have to think much more asymmetrically than we have in the past. If we're going to keep our edge in the.
Future, how should we think about the role of technological innovation because some of themselves a lot less expensive than some of the big weapons systems.
Right. I think the US is hands down the leader globally in technical innovation, and that includes in the defense domain. But we are not all that good at innovation adoption actually taking particularly if something's coming out of the commercial sector, like take for example, small cheap attritable drones. We are seeing how that is impacting the war between Russia and Ukraine. We are starting to see Iran use them against Israel, you know, in an Asia of Asian theater where China will always have a quantitative advantage because they have their whole force right there. It's their backyard where we have to project power from the United States and other regions. How do we buy back mass? How do we sort of leverage things like drones under sea, on the sea, in the air controlled by human operators. How do we leverage that to really create new problems for an adversary, and how do we use them to contribute to deterrence. So the technological piece is absolutely key to keeping our edge in the future.
Financial markets right now are really almost obsessed with artificial intelligence. How does that apply a defense area?
It is being adopted already. I think the Department of Defense gets credit and this administration gets credit for setting out a framework for responsible AI. How do we keep it safe? How do we make sure it behaves appropriately? How do we make sure, you know, it's trans parent, that we're following data rules and so forth. We're not so sure as whether our adversaries will be so responsible in how they develop it. But we see it coming into in the intelligence field, helping analysts sort through the massive amounts of information to focus on what's important for insight. We see it in the area of maintenance, preventative maintenance, so getting data that suggests you need to repair something before it breaks, which takes less time and is cheaper. And then we'll eventually seeing applying it to this human machine teaming. I talked about where you have a single human being operating very large, large numbers of unmanned systems, so you still have the human in the loop, but you've bought back a lot of mass for your forces. Ultimately, I also think it's going to be important to give us an edge and decision making. If we can process information faster, get insight faster, tee up options fast, master for the human decision maker, we will have an edge in any future crisis.
That was former Deputy Secretary Defense Michelle Flornoy, managing partner and co founder of West Exec Advisors, coming up to the rush. Degenerative AI has shifted enormous amounts of capital and driven tech shares higher. We talked about whether it's all gone too far too fast with AI pioneer Jack Heitory of Sandbox AQ.
The bigger part of the economy really needs a different kind of AI, and that AI is large quantitative bondus.
That's next on Wall Street Week on Bloomberg.
This is Bloomberg Wall Street Week with David Weston from Bloomberg Radio.
This is Wall Street Week. I'm David Weston. Invidios stock has gone through the roof the so called Hyperscalers are investing tens of billions of dollars as the dream of generave AI has seized the imagination of investors. But now some people are second guessing the land rush to take us through what's real and what's overdone. Welcome now a pioneer in the field. He's Jack Hitory, CEO of Sandbox AQ. So Jack, so good to have you with us. Thank you for being here. So for those of us who really are not that well versus this, give us your overall take. We have seen some backing off, some suggestions, for example in Vidio may have some delays in some of their chips. Maybe some questions about it. How much is real and how much of it is hype?
David Firstell, great to see you this is these are very exciting times right now for AI. Not only the kind of AI that people have seen before, like chat, GPT and AI focused on words, but also AI now applied to the biggest industries in the world to make new drugs, to make new alloys, to lightweight vehicles, to.
Create new energy opportunities for the world.
This is a great moment for AI that takes us beyond the chat GPT. When we look at Nvidia to your question about Nvidia, we see a company with a great set of high performance chips, initially the g and GPU of course for graphics, and then of course to large language models. But now there's new demand curves coming to Invidia, and so I think, well, there may be a couple of initial bumps in the road here in the near term. What I recommend people do is look at the mid and long term nature of opportunity, which now has a big demand curve for quantitative AI, which is the next wave AI that's now hitting.
So Jack, from your experience, you guys are really heavy hitter investors. Is it patient capital? I mean when they see some of these bumps along the road, do they get nervous? Did they call you up? Or are they in it for the long haul.
Most investors that call me up, and many do, are in it for the long haul. They see the value of these kind of platforms. They see the demand from the large data centers, from many companies buying these kinds of chips. But what's most important now is the next wave of AI that's coming. The initial wave, of course, was focus on downloading the Internet training language models like chat, GPT and other kinds of large language models known as llms. That's certainly a good first wave, and there are certain key applications of llms. David, for example, customer service. Right now we all agree, I think we can have better customer service from telcos and cable operators and airlines and folks like that, and I think the chat GPTs of the world will help with that. But the bigger part of the economy really needs a different kind of AI, and that AI is large quantitative models l qms instead of just l lms. And that's because we really want to see new drug drugs in the marketplace, new treatments for brain cancer, pancreatic cancer, for Alzheimer's, for dementia. I think everyone watching this show today David knows a loved one, a family friend who is impacted by these diseases. Forty years of research has not yielded much for the patients that are afflicted by these conditions. What we're seeing now is the advent of a new, bigger AI, an AI that handles quantitative relationships. It understands chemistry, it understands physics, It understands financial relationships for asset management and portfolio construction for financial services.
These are the big verticals.
Biopharma is eight trillion in size, financial services ten trillion, Automotive and aerospace fifteen trillion, and finally chemicals energy materials another fifteen trillion, part of the one hundred trillion dollar GDP economy worldwide. These are the bigger applications that are now upon us in the world of AI.
So Jack, that's very helpful. The distinction between large language model and large quantitative model big difference. That's one queue. Another queue is quantum. You're involved in quantum. What does quantum add to the mix when it comes to AI.
What's fascinating is that let's look at AI and quantum, David, as part of a bigger hole of advanced compute and so when we look at AI, I think now people get a feel of how powerful that is already, and it's going to even get more powerful over the next two three four years. And now into the world of quantum. Most people when they think about quantum, they think about quantum computers, and certainly quantum computers are going to have a wonderful impact on our society. Give us even more computer on top of and with the AI that we just discussed. But what's not often discussed is the other applications of quantum. For example, quantum sensors. These are senses that are here today, their room temperature, they're very compact their solid state, and for example, we can use them to improve medical diagnostics.
Bring us to the here and now, right now in concrete terms. I was with somebody recently from one of the hyperscalers saying, though you know, a little bit like electricity and the latter part last century, you have the power planning of the grid, and then you have the appliances, and what we need in part with AI are the appliances as it were. You had an announcement just this week with the Mayo Clinic about some cardia analysis, and wow, you can help with that. What is that and how close is it to being real?
That's a great point, David.
It's the applications of AI and quantum sensing that are the most exciting things, not just the theory. And so this week, as you pointed out, we in the Mao Clinic announced that our AI and device are inside the Mayo Clinic right now in yet another clinical trial. We already completed trials now at UCSF Hospital and the Mount Signi Hospital, New York, and now this new trial is starting at the Mayo Clinic, one of the premier hospitals of the entire world. Very very proud to be working with the Mayo Clinic on this and very excited to see some of the outcomes. What's important about that is that the ability to bring quantitative AI LQMS combined with advanced sensors in a single format of this box allow us to put that in the cardiac hospital and look at real patients in real time and how their heart is doing in a way that no other modality can do. What's happened is that quantum sensors have become room temperature, they become very portable and small format, low power draw but the challenge was how to read this huge plethora, this fountain of information coming out of these sensors. That's where AI comes in. Specifically, that's where lqms come in. You cannot add a chat GPT to this box and have it interpret those signals.
Being trained on Wikipedia.
And Reddit and Twitter is not going to help you with this kind of signal. Rather, this kind of training requires quantitative data. So this is a major step forward for cardiovasculine disease. Doctor Toby Cosgrove, who used to head up the Cleveland Clinic for the past thirteen years, called it the key to transforming cardiovascular medicine. So it's great to see this kind of advancement combining AI and quantum jack.
That was very helpful and very informative. One last question in a very different area, and that's the area of defense. We've talked to some people in the defense establishment. What's going on, but even maybe an arms race when it comes to AI and use of AI in defense contracting. Where are we that in so far as you can tell us about it.
Sure, defense is a very very critical application.
National security is of paramount importance for both United States and its allies. I think we all agree with the confabgrations going on around the world. It's even more important today than it was just a few years ago. This is both on the cyber aspect as well as the kinetic aspect of war and of conflict. And so when we look at the applications, let me give some examples. If we take navigation, when you're in a wartime scenario, even a pre wartime scenario, the first thing your adversary will do, David, is they will block and jam GPS. You want to do that as an adversary because you want to stop planes from coming at you and dropping bombs. You want to stop missiles from coming at you, you want to stop drones from coming at you.
And you see it today.
Russia right now is jamming not only would be Ukraine, but over enough of area of Europe that it has disrupted thousands of commercial flights. Thousands of commercial flights have been disrupted by Russ's jamming of GPS. And this is another application of both AI and advancedensing.
How is that the case, Jack, thank you so much as has been really terribly helpful. That is Jack Heitory of Sandbox AQ. If you want a friend in Washington, get a dog. President Truman is thought to have said this when he was running for reelection back in nineteen forty eight. There is little doubt that the companionship of an animal can be a consolation when it feels like no one else is on our side. Maybe that's why nearly every one of our presidents has had a cherished pet, going all the way back to George Washington's foxhounds, and those presidential pets haven't all been dogs. Thomas Jefferson broke from mister Washington by keeping mockingbirds. Andrew Jackson had a parent named Paul, who had to be removed from Jackson's funeral because of the parrot's foul language. You can decide where the bird may have picked that up. But despite the assorted goats, sheep, and pigs kept by presidents down through the ages, by far the most popular presidential pets have been dogs. The most famous was FDR's beloved Falla, a Scottish terrier who was President Roosevelt's constant companion. Fala even became an issue in mister Roosevelt's campaign for reelection in nineteen forty four, when Republicans claimed that millions of taxpayer dollars had been spent sending a destroyer to retrieve the dog from the Aleutian Islands after he'd been left behind. That turned out to be an urban myth, but President Roosevelt took it on directly and humorously in a national radio dress in the middle of the campaign.
The Republican leaders are not in contempt with a pack on me, are on my wife, are on my son? They now include my little dog foul.
By the way, Fala is the only presidential pet to be immortalized as a bronze statue as part of the Roosevelt Memorial on the National Mall in Washington. Since then, there's been a steady succession of dogs in the White House, from Ronald Reagan's Rex and Lucky to George Herbert Walker Bush's Millie.
To George W.
Bush's Barney, made famous by the Barney Cam showing up every Christmas from the White House. Indeed, it was only President Trump who went without a pet in recent years, but not all of the stories about presidents or presidential wannabes have been happy ones when it comes to the animals. Those of us of a certain age remember the controversy when President Johnson decided to show off his hounds they were called Him and Her, by picking them up by the ears and hearing them yelp, something the President insisted was good for them. President Biden's German shepherd commander was banished to a farm after repeatedly biting members of the Secret Service.
He put hey doing lord.
But some in national politics have gone well past making their dogs yell or sending him to a farm. South Dakota Governor Christy Noam, in the running to become Donald Trump's running mate, at one point this year, did herself no favors when she admitted in her autobiogree that she'd shot her dog when he misbehaved. And then this week, presidential candidate RFK Junior, in an interview with Roseanne Barr, admitted that he'd come across a dead bear cub in upstate New York, put it in his car, drove it to Manhattan, and then staged its death by bicycle in Central Park because, according to mister Kennedy, he didn't want the meat to go.
To waste in Central Park.
And we'll make it look like however it all really happened. It was a long way from FDR's beloved Falla. But then again, let's go back to the source of that supposed Truman quote about getting a dog if you want a friend in Washington. It turns out that there is no evidence President Truman ever said anything about making a friend of a dog. Indeed, when he was asked about a Cocker Spaniel he'd been given by a supporter, said simply that it was around someplace. The someplace turned out to be Ohio, where Feller had been sent, because, as the President put it, I didn't ask for him and I don't need him. But if my own dogs, Whiskey and Walker, are watching, don't worry. We have no plans of shipping you to Ohio. That does it for this episode of Wall Street Week. I'm David Weston. This is Bloomberg. See you next week.