YT YouTube2Text

YouTube2Text → Transcripts → Central Banks Just Ran the Numbers on AI. Report Warns Collapse is Coming.

Central Banks Just Ran the Numbers on AI. Report Warns Collapse is Coming. — Transcript

by Brendan Dell · 4,640 words · 717 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Before the 2008 financial crisis, the
0:02most powerful financial institution that
0:04you've likely never heard of issued a
0:06warning that the system was building
0:08toward collapse five full years before
0:12it actually happened. And now, that same
0:15institution is issuing a warning about
0:18AI. All those years ago, they warned
0:20that indicators of risk perception tend
0:23to decline during the upswing. Those
0:27warnings were ignored, and we all know
0:28how that ended in 2008. And now, that
0:31same institution is showing that markets
0:33are once again ignoring risks. They are
0:36accepting less and less payment for
0:38holding it, exactly the pattern they
0:41flagged before 2008. And they warned of
0:43four major pressure points converging at
0:46this moment with AI at the center of all
0:50of them. The report warns that, quote, a
0:52major equity market correction could
0:54have larger macroeconomic consequences
0:57today than in the past. That it would
0:59have more pronounced wealth effects, a
1:01sharper consumption pullback, and with
1:03US stocks at 64% of global equity
1:05markets, a US-led repricing could,
1:08quote, propagate globally. Financial
1:11stability, they write, could also be at
1:14risk in the event of an AI burst. In
1:17plain terms, AI, coupled with the oil
1:20shock and the ongoing war in Iran, are
1:22creating a flashpoint that we must
1:25understand.
1:26>> Finally, we also want to
1:28stress developments in in the public
1:31finances.
1:32Near record high public debt and higher
1:34interest rates are straining fiscal
1:36positions globally, leaving governments
1:39with limited room to respond to crisis
1:41as the cost of servicing debt escalates
1:43and deficits remain stubbornly high.
1:46>> What's worse about this particular
1:48confluence of risks is that while the
1:50short-term gains of this boom flowed to
1:53the 10% of our population who own 87% of
1:57the stock market, aka the rich, the bust
2:00will be distributed to everyone.
2:02>> So, I don't think it it has a problem
2:04with productivity. I do think not about
2:07productivity, it has a big wealth gap
2:10implication. A very small percentage of
2:12the population is going to do
2:14unbelievably, and a lot of people won't.
2:16So, what do we do? Can we work together
2:19politically to deal with those issues,
2:21and how do you do I do not believe I'm
2:24not optimistic on us working together.
2:26>> And how this ends may come down to one
2:28chart, which is buried deep in the
2:30report on page 23. And what it shows is
2:33that using the AI industry's own
2:35projections, the math puts the industry
2:38$2 trillion in the hole, unless
2:42AI delivers everything
2:45that's been promised. And we all know
2:47that so far, it has not come close.
2:51Goldman Sachs finds, quote, "No
2:53meaningful relationship between
2:55productivity and AI adoption at the
2:57economy-wide level." And their own chief
3:00economist calls AI's contribution to GDP
3:02growth basically zero. And if this
3:06happens,
3:08if the bust comes, governments have a
3:10fraction of the room that they had last
3:13time to help counter the impact. When
3:162008 hit, US debt was 35% of GDP. Today,
3:20it's 100%, and interest payments already
3:23eat a fifth of federal revenue, double
3:27the prior crisis. So, let me show you
3:29what this report shows, what it means
3:31for you, and what you can do to prepare.
3:34I'm Brendan Dell. This is the Leverage
3:35Class. Let's see through it.
3:39There's a tower in Basel, Switzerland,
3:41that houses one of the most powerful
3:43institutions that you've likely never
3:45heard of. It's called the Bank for
3:46International Settlements, but it's a
3:48bank in the technical sense only. You
3:51can't open an account there, neither can
3:53I, neither can Apple or Goldman Sachs or
3:56any company on earth. Its clients are
3:58only central banks, things like the
4:00Federal Reserve or the European Central
4:02Bank or the Bank of Japan, which is why
4:04it's often called the central bank of
4:05central banks. When the people who print
4:07the world's money need somewhere to
4:09meet, they go to Basel. They meet every
4:12two months behind closed doors and then
4:14once a year, they go to receive a report
4:16card on the system that they run. In
4:19fact, the rules that govern how much
4:20capital every major bank on the planet
4:22must hold are named after this city
4:24because they're written here. And in
4:27August of 2003, two economists from this
4:30institution walked into the Federal
4:32Reserve's own annual symposium in
4:34Jackson Hole, Wyoming and told the
4:36assembled central bankers of the world
4:38that their victory was making them
4:41blind.
4:42The paper was by Claudio Borio and
4:44William White, two of the BIS's most
4:47senior economist. And their argument,
4:49compressed, was that central banks had
4:51spent 20 years winning the war on
4:54inflation and the prize was a new kind
4:57of danger. With inflation conquered,
5:00nothing forced the brakes anymore.
5:02Credit could expand, asset prices could
5:04climb and every instrument on the
5:07dashboard would read normal right up
5:10until it didn't. In their words, the
5:12system's very stability was raising its,
5:16quote, elasticity, which was making it
5:19more vulnerable to boom and bust cycles.
5:22And the central bank, they wrote, can be
5:24a victim of its own success. In layman's
5:28terms, they told them, "You guys are
5:29getting cocky and you're not
5:31appropriately managing risk." And they
5:33even dedicated that paper to the memory
5:35of Charles Kindleberger, the man who
5:37wrote Manias, Panics, and Crashes and
5:40who himself was a former BIS staffer.
5:44That symposium was opened by Alan
5:47Greenspan himself, who was chairman of
5:49the Federal Reserve at the time, aka the
5:52man in charge of America's money. And he
5:54told that room that uncertainty was
5:56{quote} the defining characteristic of
5:59monetary policy. And then two of BIS's
6:02most senior economists presented a paper
6:05showing that room exactly where the
6:08uncertainty was hiding. But the Fed was
6:10not persuaded. They kept rates low, the
6:13housing market continued its tear, and
6:16by 2006 William White, the official we
6:19just met, wrote that one hopes that it
6:21will not require a disorderly unwinding
6:24of current excesses to prove
6:26convincingly that we have indeed been on
6:29a dangerous path. But unfortunately, we
6:33as human beings seem to like learning
6:35things the hard way. And 2 years later
6:38that disorderly unwinding arrived in the
6:40form of the Great Recession, which is
6:42one of the worst economic events in
6:45modern history. The BIS had been
6:48correct. They had been early, and they
6:51were ignored. And being ignored in 2008
6:55cost millions of people their jobs, and
6:57their homes, and their savings. Which
7:00brings us to 2 weeks ago when this same
7:03institution published its annual report
7:06card on the world economy. And their
7:09warning lights are flashing again. But
7:12this time they're pointing at AI, and
7:13using the AI industry's own numbers,
7:16their analysis shows that if things
7:18don't go perfectly to plan, the fallout
7:20would hit harder [snorts] than it did in
7:222008. Because this time the exposure
7:25runs through your retirement account,
7:28whether you've ever bought a share of
7:30Nvidia or not. So to understand why, we
7:32have to do the thing that always gets us
7:34closer to truth. We have to follow the
7:37money, and the money is going one place,
7:40AI.
7:41So the single biggest thing that we must
7:44understand to properly weigh the risk of
7:46the AI boom and its impact on our lives
7:49is that unlike the last 15 years, where
7:52companies were borrowing money because
7:55they were making so much of it, what
7:57this report shows is that the most
8:00profitable companies in the history of
8:02the world have completely flip-flopped
8:04and are now borrowing money because they
8:06can't keep up with their costs. And the
8:10only rational thing that they can do,
8:12even if it seems crazy, is to keep
8:14spending and spending and spending, or
8:17major players like Meta and Microsoft
8:20and Google all run the risk of ruin. The
8:24report names four pressure points
8:26converging on the world economy all at
8:29once: persistent inflation, AI
8:31investment, growing financial
8:32vulnerabilities, and weakening fiscal
8:34positions. And this video walks through
8:37all four. We start with the engine, the
8:40AI spending spree. 30 seconds of boring
8:42finance because three dull terms are
8:44about to matter a lot. Term one,
8:47corporate bonds. So, when a company
8:48wants money without selling ownership,
8:50it borrows from investors and promises
8:52to pay it back with interest. This is
8:54like if you ask your buddies to loan
8:55you, you know, like a cool billion, and
8:57then you'll pay them back with 3%
8:58interest. But the bigger you look from
9:00the outside, the more you can raise and
9:02the cheaper it gets. Term two, free cash
9:05flow. The money a company has left after
9:07paying to both run and grow itself. It's
9:10the finance world's answer to, "Yeah,
9:12but how much do you actually make?" And
9:14then term three, financial engineering.
9:16So, if you're a person or a small
9:18business, you basically have two
9:19options, which is earn money or borrow
9:21it. And how much you can borrow is
9:23capped by what your monthly income
9:24supports. But if you're a very big
9:26business, those rules loosen and you can
9:29structure your borrowing and your taxes
9:31and accounting so that spendable cash
9:33shows up where you need it. Yes,
9:36accountants, I know that is not a
9:38precise definition. It's the sentiment.
9:40Apple ran a very famous version of this.
9:42So, in 2013, it borrowed $17 billion
9:45while sitting on the biggest cash pile
9:47in corporate history because that cash
9:50was overseas, and borrowing it was
9:52cheaper than paying the taxes for
9:54bringing it home. S&P even gave this
9:56trick a name, which they called
9:58synthetic repatriation. Most people
10:01borrow money if they don't have enough
10:03money. But, big companies will often
10:05borrow when they have too much of it.
10:08The BIS data shows that AI commitments
10:11at the five biggest spenders have passed
10:13earnings and past free cash flow. And
10:17Morgan Stanley estimates that $3
10:19trillion
10:20of data centers through 2028, only half
10:24are covered by the cash that these
10:25companies can generate, and the rest has
10:29to come from debts and from cuts. For
10:32example, Microsoft just laid off 4,800
10:35people while carrying a $190 billion
10:37spending plan. What makes this scary is
10:40that despite these huge risks, it's
10:43basically a forced play. AI threatens
10:46the base business of all these major
10:48players, ads, search, enterprise
10:50software, everything that these
10:52companies own.
10:54>> Great technology changes
10:57um
10:59produce bubbles.
11:01And the reason they produce bubbles is
11:03because nobody can get get it exactly
11:06right. Okay. There um
11:08you have to either spend a ton of money
11:11to capture your market share and so on.
11:14Or um and and you might and don't worry
11:16about whether it's too much or not. Uh
11:18or you don't spend enough money and you
11:20lose your market share. And it's very
11:22imprecise with a lot of competition,
11:25okay?
11:26>> And the BIS sees this precise risk
11:29growing. They say it's a contest. Bond
11:32issuance by these AI-related firms is
11:34now a very significant share of total
11:37issuance by corporates. And what they're
11:39finding is that even if things go
11:43completely to plan,
11:45if AI delivers absolutely everything
11:47that it's promised, we are still
11:50entering a very risky scenario.
11:53Which brings us to the chart from the
11:55beginning of this video. So, first, the
11:56roughly $2 trillion already committed.
11:58This is the red line. AI delivers
12:00everything promised, and then the
12:02sector's payoff that we see here still
12:05falls with every trillion spent because
12:08they're all chasing the same prize. Now,
12:11the blue line is AI disappoints, which
12:13in their model means AI delivers half.
12:16It doesn't mean it fails outright. It
12:18doesn't mean it produces no value. It
12:20just means it does only half of what
12:21they're saying. So, at marker B, the
12:24three to four trillion dollars that
12:26Nvidia's own CEO projects by 2030,
12:31and this is one of the largest bulls in
12:32the entire AI economy, by the way, puts
12:35the entire sector $2 trillion
12:39underwater. But, the most important
12:41point is this.
12:43The money is borrowed. The shortfall
12:46lands on whoever lent it, and
12:48increasingly, that is not the banks that
12:50you'd expect. It's private credit funds,
12:52it's bond portfolios, it's pensions, and
12:54as we're about to see, no one can even
12:57fully understand where all this is going
13:00to land. So, the big question then
13:02becomes, how much is AI actually
13:04delivering? Well, this is being
13:07measured, and so far, things are not
13:09looking good. The biggest challenge of
13:12large language models is that the
13:13technology itself is being
13:15anthropomorphized and used as a synonym
13:17for all of AI,
13:19all technologies of artificial
13:21intelligence. And as a result, the
13:23companies building the models are being
13:24priced as a replacement for all of
13:27thinking when they are actually normal
13:29technology with specific applications
13:32whose limits are already being found.
13:34And we see this happening in three
13:36places in real time. The first is the
13:38frontier models are being commoditized
13:40by their own customers. Yesterday
13:42Bloomberg reported that Microsoft has
13:43started replacing OpenAI and Anthropic
13:46with its own cheaper models inside Excel
13:48and Outlook. Microsoft's AI chief, on
13:51the record, "We pay a lot of money to
13:52Anthropic, so our goal is to reduce
13:55these costs and ultimately eliminate
13:57them." Chinese open models charge a 20th
14:00of frontier prices for all of the
14:03everyday work, which is most of what the
14:05technology is used for. If AI were a
14:08thinking replacement, its biggest
14:10customers would not be swapping them out
14:12to save money on spreadsheet formulas.
14:14Second, the returns from scalar
14:16flattening, which said plainly means the
14:18models are unlikely to keep just getting
14:20better and better and better. Ilya
14:22Sutskever, who co-founded OpenAI and
14:24built the scaling era, in his own words
14:26said, "Is the belief that if you just
14:28100x the scale, everything would be
14:30transformed? I don't think that's true.
14:32The age of scaling is over," he says.
14:34"We are back in the age of research."
14:37Two computational scientists published
14:39the math of why the scaling laws own
14:41exponents make reliability gains
14:44brutally expensive. The paper is
14:46literally titled The Wall Confronting
14:48Large Language Models. Third, the
14:51economy-wide reports. Goldman Sachs
14:53reports that there is no meaningful
14:54relationship between productivity and AI
14:56adoption. Their chief economist puts
14:58AI's GDP contribution at basically zero.
15:01They do show real gains of about 30% in
15:04exactly two jobs, coding and customer
15:07support.
15:0830% in two places, zero everywhere else.
15:11That's what a tool looks like. That is
15:14not a workforce replacement, and it
15:17cannot justify these valuations. Now, in
15:21fairness, many technologies have a lag
15:24between when they deploy and when we can
15:26actually measure productivity
15:27improvements. The normal technology
15:29researchers, who I referenced earlier,
15:31explained in that document that
15:33electricity took nearly 40 years to show
15:35up in the productivity statistics. The
15:37internet took a long time also. The
15:39diffusion of technology is slow. And it
15:41is very likely that AI is in that gap.
15:45But even if it is, this won't save the
15:47industry.
15:49Let's go back to that chart. The
15:50industry spending plan only works if AI
15:53delivers absolutely everything. That's
15:55the best case per the biggest AI bull
15:58alive. And even his line barely pays.
16:01Miss by half and every measurement that
16:04we just looked at shows them missing and
16:06the sector is $2 trillion in the hole
16:10with borrowed money that has to be paid
16:12back. Which brings us to the single
16:14largest insight of this report. The
16:17plumbing that moved all this money into
16:19AI creates risks that reach far past the
16:22industry and into the financial system,
16:25into your finances and it's revealed
16:27through one phrase hidden on page 25 of
16:31the report and that phrase is pledged
16:34multiple times. That's the risk that the
16:37report calls growing financial
16:39vulnerabilities and it's pointing at a
16:41financing structure that means no one,
16:43not the BIS, not anyone can fully see
16:46where all this will land. But what we do
16:48see is the people best positioned to
16:51understand where those losses will land
16:53are already heading for the exits. So,
16:56you've likely seen these spaghetti
16:57diagrams running around the internet
16:59showing how the AI boom is being
17:01financed. What it shows is that big
17:02companies like Nvidia agree to invest in
17:04startups like Open AI in exchange for
17:06purchase orders on their chips and the
17:08whole thing then spins round and round
17:10and round fueling this economy. But the
17:12BIS flags this risk in the same
17:16understated but severe way that they
17:19flagged the mortgage risk. They explain
17:22the opacity of AI sector financing
17:24compounds these vulnerabilities.
17:26Hyperscalers, chipmakers, and AI labs
17:29are linked to a complex web of private
17:33arrangements. The most prominent is
17:35circular financing. Chipmakers and
17:37hyperscalers take equity stakes in AI
17:39labs or neo cloud providers who in turn
17:41commit to multi-year purchases of chips
17:43or computing power. They continue by
17:45saying signs of stress are already
17:48visible and that the real economy
17:50implications could be substantial.
17:53So, then, what specifically are those
17:56signs of stress?
17:58Insurance against these companies
18:00defaulting, called credit default swaps,
18:02have been growing steadily more
18:04expensive since January of last year,
18:07even as the stock prices keep climbing.
18:11What this means is that people in the
18:13know see risk. The bond market and the
18:15stock market are pricing very different
18:17futures for the same companies. And the
18:19retail credit funds that lent to this
18:21sector are already facing redemption
18:24requests and forced sales. And what we
18:26see as all this is happening is the
18:28people with the most knowledge of what's
18:30going on inside starting to run for the
18:32exits.
18:33>> The companies that go public are those
18:35whose current investors are saying, "We
18:38don't believe this company will go up in
18:39value, so we'll sell it to public market
18:42investors."
18:42>> Last month, the biggest IPO in history
18:45happened and it was priced at $135 a
18:47share.
18:49Morningstar's analysts said it was worth
18:50only $63 a share. It spiked and then has
18:54since declined. SpaceX set aside roughly
18:5730% of the shares for retail investors,
18:59which is to say ordinary people through
19:02things like Robinhood or Schwab.
19:05Normal [snorts] IPOs allocate
19:07single-digit
19:09allocations to retail investors. What we
19:12are seeing happen is after 20 years of
19:15private appreciation, they stacked
19:18companies together, then opened the
19:20doors to the public at twice what the
19:22professionals said it was worth. And
19:26it's what OpenAI and Anthropic have
19:28confidentially filed to do next. And
19:30yes, I know that OpenAI offering is
19:33delayed. We can't go into all of it in
19:35this video.
19:36And they're planning to IPO at
19:38sales-to-price ratios that far exceed
19:40what Jay Ritter, who is a man who has
19:42studied every major IPO of the last
19:44century, calls the danger zone. We'll
19:46link that video in the description if
19:48you want to learn more.
19:49To be fair, none of this is illegal, and
19:52none of it is new. Telecom vendors ran
19:55money in circles in 1999. As one
19:57example, Lucent alone extended more than
19:59$8 billion in loans to customers so they
20:01could buy their stuff. But, when the
20:03funding stopped in 2001, most of those
20:05loans were never repaid, and Nortel went
20:09from a $390 billion valuation to
20:11bankruptcy. What's new about this
20:13particular situation is the scale, and
20:17the fact that this version runs through
20:19all these private deals that no one can
20:21fully audit, which is what creates so
20:25much risk for the average person and for
20:28the economy at large. The professionals
20:31are buying insurance, credit default
20:33swaps, the lenders are all stretched,
20:35and the founders are getting out, which
20:38leaves one group still fully committed,
20:41mostly without knowing it. If you have a
20:43retirement account at all, or a job, it
20:46is you. And the risk the report flags is
20:49that unlike 2008 when the government was
20:51able to rescue our economy through
20:53bailouts, this time it would be much
20:56harder for it to do so. This is the
20:58pressure point that the report calls
21:00weakening fiscal positions. So, let's
21:04say that the The is right, the returns
21:06disappoint, financing unwinds, then
21:09those losses land on everyone. When
21:11Lehman collapsed, US government debt
21:14stood at stood at 35% of GDP. Today,
21:17it's roughly 100%. Nearly three times.
21:21And interest payments alone now eat
21:23about a fifth of all federal revenue,
21:25which is double the burden of any prior
21:27crisis. The United States now spends
21:29more servicing its debt than it spends
21:31on national defense. The BIS says, in
21:35their calm and understated way, "Near
21:38record high public debt and higher
21:39interest rates are straining fiscal
21:41positions globally, leaving governments
21:43with limited room to respond to crisis."
21:45They are calmly warning that the whole
21:48system may go up in flames with no fire
21:51extinguisher. Which brings us to the
21:53last pressure point of the report, which
21:55is inflation. Specifically, the oil
21:57shock out of the Strait of Hormuz. This
21:59week, the Iran ceasefire collapsed and
22:02oil started climbing again. Last time
22:04that shock hit, the report notes that AI
22:06spending is what held up the economy.
22:09But this time, the shock and the doubts
22:11about AI are arriving together. So, at
22:14the press briefing for the release of
22:16the report, a Reuters journalist asked
22:18the question directly, "How big could
22:20this get? Could this be financial crisis
22:22big?" And the response was,
22:24"Individually, each pressure point might
22:27not be particularly worrisome, but it's
22:29the combination of the four. We cannot
22:31exclude that they might materialize and
22:33damage the global economy in a more
22:35significant manager."
22:38The general manager of the central bank
22:40of central banks was asked on record
22:44whether this could be 2008 scale, and he
22:47didn't say no.
22:48And the distribution
22:50of the losses of all of this is what
22:53makes it worse. The boom's gains have
22:55already been distributed. The top 10% of
22:58our economy own roughly 87% of the stock
23:01market, which means that the benefit of
23:04the AI boom has flowed mainly to people
23:06who in that sector or hold those stocks.
23:10The IPOs move the risk to retail, the
23:12debt moved it to bond funds and
23:13pensions, and if the bust comes, it
23:16arrives as a recession, which will
23:18impact consumption and jobs, all with
23:20the government will be too stretched to
23:22cushion the blow. So, the gains went to
23:25the people who own the boom, but the
23:26costs will be borne by everyone. Last
23:30week, the Financial Times reported that
23:31Open AI has proposed the US government
23:33take a 5% stake in their company. And in
23:36every leading American AI lab through a
23:39sovereign wealth fund. In 2008, the
23:41government took stakes in companies as
23:42the price of rescue after they had
23:45failed. This is a company offering
23:47equity prematurely, while insisting
23:49everything is fine,
23:51and offering that equity to the entity
23:53that writes its rules and would run its
23:56bailout. But regardless of the framing,
23:58however they couch the offer, what it
24:00would unquestionably do is give the
24:03government a multi-billion dollar
24:06interest in these companies staying
24:08afloat, as well as regulatory capture,
24:12talk for another video.
24:13Now, these talks are very early and
24:15might require an act of Congress. It may
24:17never happen. But it's the implication
24:19of the proposition itself that we need
24:21to understand. The BIS closed its
24:24briefing with one sentence of advice.
24:27Policy makers must act now to label only
24:30make the necessary adjustments more
24:33costly. In 2006, the BIS warned, "One
24:37hopes that it will not require a
24:38disorderly unwinding of current excesses
24:41to prove convincingly that we have
24:42indeed been on a dangerous path." But
24:45unfortunately, it seems that it may
24:47again require that disorderly unwinding
24:51to show that again.
24:53Though we very much hope not. So, I've
24:55been asked in the comments to conclude
24:57these videos with my short perspective
24:59on what I'm doing. Here it is. Years
25:02ago, I got to meet one of my friend's
25:03godfathers. This This was a guy who had
25:05made close to a billion dollars in real
25:07estate. And this guy knew everything
25:11about real estate. He knew rental rates
25:13in every market. He knew everything
25:15about his buildings. He could tell you
25:17what the copper wire was worth in his
25:19walls. And me being young and trying to
25:21be smart, I asked him, "So, what's the
25:24difference between a good investment and
25:25a bad investment?" And I thought he was
25:27going to have a fancy answer. But
25:29without blinking, what he said was 10
25:32years. So, I always remember this quote
25:34because what I find is that all too
25:36often in life, rather than seeking
25:39compounding,
25:41looking for long-term returns, we try to
25:44run to find an exit. We run away from
25:47what we don't want instead of toward
25:49what we do. So, I would suggest asking
25:51yourself this question. If I could spend
25:54my days doing any kind of work for the
25:56next 30 years,
25:58can't be lying on a beach, it has to be
25:59doing something productive for others,
26:02but it could be anything. What would
26:04that be?
26:05And then I would make a direct plan to
26:09pursue leverage in that direction. Gain
26:12specific skills that you can sell.
26:16Specialize, don't generalize. Build an
26:18audience of people who know who you help
26:21and how you help them. And you don't
26:23have to dance on TikTok. If you have a
26:25list, an email list of a thousand CFOs,
26:30right? You're in finance. You will
26:32always have opportunity if those people
26:35look at you as a voice to listen to in
26:38your space. And this is buildable by
26:41anyone, and you will always have
26:42opportunity.
26:44Be an expert in how modern technologies
26:46can give you leverage in your specific
26:48area of expertise. Build processes so
26:51that you can detach money from time and
26:54then and only then use money as an
26:56amplifier to compound. Don't make bets
26:59on things you can't control. I built my
27:01life as a sovereign professional where I
27:04control my time and I do work I enjoy
27:06and I earn money far in excess of what I
27:08need and it was using that exact process
27:11and it's exactly what I would do again
27:13if I had to start from zero. It works in
27:15any economic environment, of course
27:17save, you know, total world code
27:19collapse and if that happens then I'm
27:21all bets are off and it compounds in a
27:23way that will fuel your life in the
27:26perpetuity. Start by building those
27:28skills and then learn how to sell them
27:30in a way that gives you income far in
27:33excess of your needs through consulting,
27:35through productized services, through
27:37content. There's a variety of ways to do
27:39this and do it in a way where you
27:41control your hours so that you work in a
27:44way that suits you and not in excess and
27:47that will provide far more life returns
27:50than some windfall bet that you keep
27:52waiting for. So then I would ask you,
27:54what work would you do if you knew that
27:58you couldn't fail? What's like the one
28:00part of your job that you wish was the
28:01whole part of your job? What are the
28:03skills you have that people talk to you
28:05about that friends ask you about that
28:07are your unique moat? That is where both
28:09your long-term protection and your big
28:11upside lies. If you want to check out
28:13additional resources to help with the
28:15above, I'll leave links in the
28:16description. To understand more about
28:18these IPOs, I'll link that video next.
28:20If you want to learn more about the
28:21leverage stack, see the I'm 43 video.
28:24See you in the next one.

About this transcript

This page contains the full transcript of Central Banks Just Ran the Numbers on AI. Report Warns Collapse is Coming. by Brendan Dell , generated from the public captions YouTube serves with the video. The transcript has 4,640 words across 717 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.