YOUR DAILY EDGE: 16 December 2024

The Fed’s Game Plan on Interest-Rate Cuts Keeps Shifting Investors widely expect a third-in-a-row rate cut this week. Officials are ready to slow—or even stop—lowering rates after that.

Nick Timiraos is setting the table for the Fed so nobody is surprised Wednesday:

(…) Investors widely expect a third consecutive rate cut this week.

Powell is trying to find the right gear amid signs the labor market is less wobbly and inflation is a touch firmer than they appeared in September. He faces misgivings from some colleagues over continuing to cut and less conviction from others who strongly backed those first two moves. One option this week would be to cut by a quarter point, then use new economic projections to strongly hint that the central bank is ready to go more slowly on the reductions. (…)

Even if officials still think price growth will gradually slow to their target, some could be less confident in that forecast due to promises by President-elect Donald Trump to deport workers and impose tariffs when he takes office next month. Those steps could reverse two developments that have underpinned officials’ sanguine inflation forecasts: falling prices of goods and a slowdown in wage growth. (…)

They are also uneasy that euphoric conditions in the stock market and for speculative assets such as bitcoin could provide grist for spending that keeps inflation entrenched. Given recent economic activity, “it’s hard to think that the level of interest rates is restrictive at this point,” said Fed governor Michelle Bowman in a speech this month.

Dallas Fed President Lorie Logan warned against cutting too far on what she views as a mistaken belief that a more “normal” interest rate for the economy is much lower. She compared the situation to a ship captain whose depth finder might mistake mud for water.

Another group of officials, including Powell, have suggested they share that concern but don’t think the Fed is at risk of cutting too much—yet—given how high they lifted rates over the past two years.

“We’re mindful of the risk that we go too far, too fast, but also of the risk that we don’t go far enough,” Powell said last month. “It seems like we’re right where we need to be. (…)

The FT’s Colby Smith says that the FOMC will chart a more gradual pace of cuts next year on concerns that progress on driving down inflation is stalling.

She says that many officials still view policy as restrictive and is “weighing on demand and, in turn, inflation”, necessitating “that monetary policy should continue to be dialed back to a more “neutral” setting that is less stifling for growth. (…) Officials also appear to be taking comfort in a recent “mini boom” in US productivity which some argue has raised the prospects that higher wages and a strong economy are not incompatible with inflation continuing to decline.”

• Goldman Sachs Economists See Fed Skipping a Rate Cut in January
• Economists trim Fed rate cut estimates on fear of Trump inflation surge

Apollo’s Torsten Slok asks “Is inflation rising again?” and offers these charts:

  

  

The Bank of Canada’s Tiff Macklem to the Globe and Mail last Friday:

(…) For the last two years, we’ve been very much focused on getting inflation down to 2 per cent. We’ve got it down to 2 per cent. It’s been down around 2 per cent since the summer. (…)

The recent experience with high inflation has been a reminder just how much people hate inflation. And even as inflation has come back down to 2 per cent, prices for many goods, and many of the things that people buy regularly, like food, are considerably higher, and they’re reminded of that every week. It feels unfair. It feels like they’re ripped off. More broadly, it undermines trust in the central bank to control inflation. It undermines trust that our market-based economy works, and that it’s working for everybody.

We saw our own credibility take a hit when inflation went way up. What we’ve seen, though, more recently, is by being clear that we were going get inflation back to 2 per cent, and then delivering on that, that is restoring trust. We survey households and businesses on where they expect inflation to be, and in the last months it has come down quite a bit. And we do our own surveys of Canadians, and what we see is that trust in the Bank of Canada has come back. We’re not seeing permanent scarring. All I can say is, we do not take the public’s trust for granted. (…)

If you look back historically, there have been periods with large divergences between Canada’s interest rate and the U.S. policy rate. We’re currently a little over one percentage point. It’s certainly been wider than that in the past.

We are starting to see a bit more of a depreciation of the Canadian dollar. How much of that is related to the interest-rate spread I think is an open question, because much of the depreciation of the Canadian dollar vis-à-vis the U.S. dollar really is more reflective of U.S. dollar strength. If you look at Canada’s exchange rate relative to other major currencies, it actually has not changed very much. (…)

Well, the rate spread has “certainly been wider than that in the past”, but not very often Mr. Macklem:

TARIFFS WATCH

U.S. Manufacturers Are Stocking Up on Imports Ahead of Tariffs A gauge of procurement activity in manufacturing in November hit its highest level in more than a year

Buying activity among North American manufacturers, measured in a survey of 27,000 businesses worldwide by GEP and S&P Market Intelligence, in November hit its highest level in more than a year. 

U.S. industrial manufacturers and consumer-packaged-goods companies are buying up critical parts and raw materials, such as emulsifiers and flavor enhancers, driving up demand, according to GEP, a supply-chain software company that gathers the data. 

“We are seeing a massive pull forward,” said Inna Kuznetsova, chief executive of ToolsGroup, a supply-chain planning and optimization company. (…)

U.S. businesses are giving priority to the purchase of their most critical components before tariffs hit, said John Piatek, a vice president of consulting at GEP who works with global manufacturing, consumer-goods and biotechnology companies. Piatek said manufacturers’ main concern is that if they wait, competitors will secure the items at lower cost and the companies will be forced to raise prices before competitors do. (…)

Retailers and manufacturers say China is such a reliable and efficient supplier of parts and finished goods that it is hard to pull away completely. Instead, companies are seeking out new suppliers in one or more additional countries to mitigate the risk of additional tariffs or other geopolitical shocks.

The shifts in supply chains are throwing up new challenges for importers as they navigate new markets that sometimes lack good road, rail or port infrastructure. Importers must also vet suppliers and sub-suppliers to make sure they don’t violate increasingly stringent U.S. rules on issues such as forced labor.

Some retailers and manufacturers looking to mitigate the risk of a fresh round of Trump administration tariffs are placing orders as quickly as possible because it can take months for products to arrive from factories in Asia. The National Retail Federation recently raised its forecasts for U.S. imports, predicting a surge in container shipping activity through the spring.

“We know that it’s going to happen,” said Jason Miller, a professor of supply-chain management at Michigan State University. “Because we saw it happen the last two times.”

Miller pointed to the two most recent quarters that saw the highest volume of container imports from China—the final quarter of 2018 and the third quarter of 2024—noting that both periods coincided with the implementation of China tariffs under the Trump and Biden administrations.

Miller said the manufacturing components most likely to see a pull forward in the coming months include auto parts, electrical components and fabricated metal products.

From GEP’s website:

The GEP Global Supply Chain Volatility Index — a leading indicator tracking demand conditions, shortages, transportation costs, inventories and backlogs based on a monthly survey of 27,000 businesses — signaled the smallest level of spare capacity in global supply chains since June in November, as the index rose to -0.20, from -0.39 previously.

Driving this increase was Asia, as suppliers to the region reported stretched capacity for the first time since July. This was caused by a surge in procurement activity by manufacturers in the continent, and especially China, as new orders rebounded sharply. This could reflect greater production requirements stemming from domestic stimulus measures, as well as from international clients, who may be stockpiling to mitigate the risk of higher import costs under the Trump administration. Only India reported a greater rise in raw material purchases than China in November. Preparations to ramp up production further were evidenced by our data showing factory procurement activity across Asia rising at its fastest pace for three-and-a-half years.

Indeed, in North America, reports of safety stockpiling were at their most pronounced since July, highlighting how procurement managers have already implemented changes to their inventory strategies as a result of the incoming U.S. administration’s public commitment to impose significant tariffs. Subsequently, a pickup in activity across North American supply chains resulted in fewer vendors with idle capacity. In fact, our index tracking the region’s supply chain activity hit a four-month high in November.

“In November, U.S. manufacturers, particularly in the consumer goods sector, increased their safety stocks to help blunt any immediate tariff increases,” said John Piatek, vice president, GEP. “In contrast, Chinese manufacturers are getting busier as a result of government stimulus and growth in exports, led by automotives and technology products. Strategically, many global companies have a wait-and-hope approach, while simultaneously planning to remake their global supply chains to respond to a tariff and trade war in 2025 and beyond.”

But let’s not get carried away. Beyond words, there’s the actual rather young indicator which is not yet above the zero demarcation between underutilized and stretched capacity:

Here are the November Key findings per GEP:

• DEMAND: Demand for raw materials, commodities and components is rising after a sustained period of weakness. Although our tracker remains slightly below its long-term average, it picked up again in November. This was principally driven by Asia, as procurement activity surged due to companies, particularly in China, preparing to ramp up production to meet new orders from clients.
• INVENTORIES: The stockpiling indicator, which measures to what extent companies are building safety buffers into their inventories to mitigate against risks such as shortages or price rises, ticked higher in November. Most notable was a rise in safety stockpiling from manufacturers in both North America and Asia.
• MATERIAL SHORTAGES: The item shortages indicator continued to show robust global supply levels in November, with the frequency at which businesses reported poor availability remaining historically low.
• LABOR SHORTAGES: Reports of manufacturers’ backlogs rising due to staff shortages were at historically typical levels during November. Therefore, the data does not suggest that labor capacity is a limiting factor for goods producers. 
• TRANSPORTATION: The transportation cost indicator remained anchored at its long-term average value in November.

REGIONAL SUPPLY CHAIN VOLATILITY

• NORTH AMERICA: Index went up to -0.36, from -0.72, its highest level since July, signaling the smallest amount of slack in the region’s supply chains in four months. Stockpiling activity ticked higher in North America in November.
• EUROPE: Index fell to -0.72, from -0.52, close to its lowest level year-to-date, signaling a worsening of the continent’s industrial recession.
• U.K.: Index ticked up to -0.12, from -0.40. However, input demand at U.K. factories worsened in November, indicating spillover effects from weakness in mainland Europe.
• ASIA: Index rose to a four-month high of 0.15, from -0.20. Crucially, the index signaled stretched capacity for the first time since the summer as a surge in procurement activity, particularly in China, squeezed vendors.

Tightening, but not tight just yet.

Americans Are Stockpiling to Get Ahead of Tariffs Some consumers are snapping up computer parts, vacuum cleaners, coffee and olive oil before levies take effect

A quarter of Americans surveyed said it was a good time for major purchases as they expect prices to go up next year, up from 10% a month prior and a record high, according to the University of Michigan’s monthly survey of consumers. And a third of the 2,000 people surveyed recently by CreditCards.com said they were buying more now because they feared tariffs. (…)

The last time the U.S. levied tariffs on all other countries was in 2018, when Trump imposed a 20% to 50% tariff on all imported washing machines in response to a petition by Whirlpool, said Felix Tintelnot, an economist at Duke University.

The action resulted in a 10% across-the-board price increase on washers and dryers, according to a study by Tintelnot and others. (…)

Christopher Foote said the possibility of tariffs have driven him to “go all out” for items he has had his eye on. The 35-year-old software consultant said he has spent more than $12,000 on new devices since Election Day, including a $8,087 Samsung heat pump, a $3,214 LG television, a $1,081 Denon audio receiver and a $509 Miele vacuum cleaner. (…)

Canada to Consider Export Tax on Commodities as Part of Trump Tariff Response Commodities that could be affected include crude oil, potash and uranium

(…) Over 20% of Canada’s gross domestic product is tied to trade with the U.S. About three-quarters of Canadian exports are intended for the U.S., led by shipments of crude oil, automobiles and auto parts. (…)

Danielle Smith, the premier of the commodity-rich province of Alberta, said Thursday that an export tax “would be a terrible idea.” She added Alberta has no intention of cutting off or thwarting energy exports to the U.S. as part of a broad Canadian retaliation.

Ontario Premier Doug Ford said this week he would be willing to cut off electricity exports to northeastern U.S. as part of a trade row. (…)

Time magazine just proclaimed Donald Trump “person of the year”. One year from now, it will likely proclaim Tariffs “word of the year”.

How badly widespread tariffs will impact prices and economies is anybody’s guess.

But strange things are already happening on the price front:

• Chinese producer prices have been declining for 2.5 years and are still dropping 2.5% YoY.

• U.S. nonfuel import prices were also deflating YoY between early 2023 and March 2024. But they have been reflating since and were up 2.3% YoY in November, the highest inflation rate since 2011 ex-pandemic.

• Import prices of consumer goods ex-autos were also deflating all of 2023 but are now up 1.2% YoY, also the highest since 2011 ex-pandemic.
• Imports of automotive vehicles, parts and engines never really deflated due to extreme scarcity. They are still up 2.4% in November, also the highest since 2011 ex-pandemic.
• Prices of imports from China have also deflated since 2023, reaching –3.2% in November 2023. They have declined only 0.2% sequentially since July and are now down 0.9% YoY, meaningfully less than China’s PPI.
• Some Chinese exports are now diverted through ASEAN countries. Import prices from these countries are down 0.3% YoY but they have increased 1.5% annualized since July.

So, even before the tariffs brawl gets underway, and even with Chinese PPI still down 2.5%, U.S. nonfuel import prices are already broadly reflating at rates above the Fed’s target.

Maybe the 30% jump in Americans’ volume purchases of durable goods since 2020 has absorbed much of the world’s excess production capacity.

Or maybe, importers are boosting their margins on this strong demand, amplified by the current pre-tariffs stockpiling.

Whatever the reason(s), the disinflationary effect from goods may be about to end, tariffs or not. In November, core CPI is up 3.3% YoY. It is +3.6% ex-durable goods and would have been 3.8% if durables inflation had been 2.0% YoY.

China Consumer Slowdown Shows Urgent Need to Encourage Spending Retail sales rise 3% in November, slowest pace in three months

(…) Retail sales rose 3% from a year ago, the slowest pace in three months and undershooting even the most bearish of forecasts. Industrial output increased 5.4%, keeping momentum as the manufacturing side of the economy continues to outperform consumer spending.

“The data show that the recovery in domestic demand has remained sluggish, while the stabilization in industrial production was likely due to some order front-loading ahead of US tariffs and is not sustainable,” said Michelle Lam, Greater China economist at Societe Generale SA. (…)

The weakening in retail sales showed the limits of a government initiative to stimulate consumption by subsidizing purchases of home appliances and cars. While sales for those two categories remained strong in November, a number of discretionary goods recorded a slump. Cosmetics led the decline with a 26% plunge in sales from a year ago, while those of clothing, jewelry, beverages and tobacco and alcohol also decreased.

NBS spokesperson Fu Linghui attributed some of the slowdown to the fact that the Singles’ Day online shopping festival — traditionally held on Nov. 11 each year — started earlier than usual in October this year, which squeezed sales last month. (…)

Home sales grew in year-on-year terms in November for the first time since the economy re-opened in the first quarter of 2023 from Covid lockdowns, based on Bloomberg Economics’ calculation of official data. Floor space sold in November rose 2.7% from a year ago following a 1.8% decline in October.

New-home prices in 70 cities, excluding state-subsidized housing, dropped 0.2% from October, the smallest decrease in 17 months, National Bureau of Statistics figures showed Monday. Values of used homes fell 0.35%, the least since May 2023.

Also;

• Fixed-asset investment increased 3.3% in the first 11 months of the year from the same period in 2023, slowing from the pace in January-October
• Jobless rate remained unchanged at 5% from a month ago
• Property investment fell 10.4% in the period, slightly worsening from a slump of 10.3% in the first 10 months

Home prices are declining at slower rates, but still declining. YoY, new-home prices are down 6.1% (October -6.2%). Existing-home prices are down 8.5% (-8.9%).

China is experiencing the exact opposite of the U.S. wealth effect and as the Fed is slowly finding out, it is powerful and resilient.

A few weeks ago, I met an 80 year-old lady on a rather expensive cruise. Not really wealthy but truly ecstatic about her Broadcom and Nvidia investments, boasting about her $400 facial treatment…

She might need a stock market correction to slow her spending (her husband kept asking me what I expected next year ).

Meanwhile, Chinese are now avoiding displaying any signs of being well-off.

They need a recovering housing market to start spending again.

The ruling Communist Party faces a “long, long battle” to reflate the economy, said Robin Xing, chief China economist at Morgan Stanley, adding that 2025 will “be the year of trying.”

“They will try a lot of things — see it’s not enough and keep trying,” Xing told Bloomberg Television. “Maybe by 2026 finally they will find the right dose of policies — a combination of consumption-centric stimulus plus social safety net reform.” (Bloomberg)

  AI Wants More Data. More Chips. More Real Estate. More Power. More Water. More Everything Businesses, investors and society brace for a demand shock from artificial intelligence.

It looks easy enough: Ask ChatGPT something, and it responds. But pull back the curtain, and you’ll find that every ChatGPT prompt and Microsoft Copilot task consumes vast resources. Millions of human beings engineering, correcting and training models. Enough terawatt-hours of electricity to power countries. Data center mega­campuses around the world. Power line networks and internet cables. Water, land, metals and minerals. Artificial intelligence needs it all, and it will need more.

Researchers have estimated that a single ChatGPT query requires almost 10 times as much electricity to process as a traditional Google search. Your typical search engine crawls the web for content that’s filed away in a massive index. But the latest AI products rely on what are known as large language models, or LLMs, which are fed billions of words of text—from the collected works of William Shakespeare to the latest forecasts of the Federal Reserve.

The models detect patterns and associations and develop billions and billions of so-called parameters that help them mimic human behavior. Using these models, ChatGPT and the like create new content—hence the term generative AI.

The resource-intensive nature of AI will create winners and losers. Those with the most resources will have the most advanced AI systems. It’s leading to clashes over increasingly scarce commodities, as well as access to chips. It’s motivating tech com­panies to seek more efficient means of developing AI. They’re throwing ­billions of dollars into alternative energy solutions such as nuclear fusion that have spent years if not decades sputtering along without heavy spending and technological breakthroughs. At the same time, AI’s demands are adding to the ­pressure to keep burning fossil fuels to feed the power grid, even as the world is on track to blow past crucial emissions targets in the fight against climate change. (…)

Here’s a closer look at everything the AI industry needs to keep its models running.

Another 1,000 Terawatt-Hours of Power

AI largely lives and runs in data centers humming with ­motherboards, chips and storage devices. The electricity demand from these centers is now outstripping the available supply in many parts of the world. In the US, data centers are projected to use 8% of total power by 2030, almost triple the share in 2022 when the AI frenzy took off, according to Goldman Sachs Group Inc., which has described it as “the kind of electricity growth that hasn’t been seen in a generation.”

Similar surges in demand have been forecast for Sweden and the UK. By 2034 annual global energy consumption by data centers is expected to top 1,580 ­terawatt-hours—about as much as is used by all of India—from about 500 today. (…)

The world’s largest tech companies have grown acutely aware that power could be the most disruptive kink in the AI supply chain, and they’re racing to lock in long-term supplies. In May, Microsoft and Brookfield Asset Management Ltd.’s green energy arm signed the biggest corporate clean energy purchase agreement.

In October, the world’s largest solar and wind power generator, NextEra Energy Inc., said it had struck deals for the potential development of a remarkable 10.5 gigawatts of renewable energy and storage by 2030 for only two Fortune 50 companies. In a sign of the clashes to come, they aren’t even tech companies. The boom has created “even more of a premium on other industries outside of data centers to try to lock up low-cost renewable generation,” NextEra Chief Executive Officer John Ketchum told investors. “All ships are rising with the tide here.” (…)

A technique that Google pioneered has emerged as a solution: Use software to hunt for clean electricity in parts of the world with excess sun and wind on the grid, then ramp up data center operations there. Otherwise, arguably the only dependable, around-the-clock source of zero-emissions power at the moment is nuclear. This explains why Microsoft struck a deal in September that will reopen a reactor at the Three Mile Island nuclear power plant in Pennsylvania, site of a notorious partial meltdown in 1979.

About a month later, Amazon.com Inc. signed three agreements to develop small-scale nuclear reactors, and Google invested in and committed to buying power from a company similarly developing modular reactors. (…)

Power lines and substations are the most underrecognized links in the AI chain. All the new data centers will need to be connected by a grid that’s already old, under stress and vulnerable when weather goes bad. (…)

OpenAI co-founder and CEO Sam Altman is talking about data centers that could need 5,000 megawatts. Building a power system that can support that much load in a single place from scratch in short order is “functionally ­impossible,” Constellation Energy Corp. CEO Joe Dominguez says. Constellation is the owner of the Three Mile Island nuclear plant that’s reviving a reactor to feed power to Microsoft. Dominguez says data center builders need to be thinking about co-locating around giant, already existing power resources—such as his nuclear power plants. Build a mega­campus next to a couple of nuclear reactors, surround them with renewable energy resources and batteries, connect them all together with new wires and load-shifting controls, and you can create a self-contained grid.

Billions of Liters of Water a Day

Every watt of electricity that’s fed into a server generates heat. Temperatures too high can destroy equipment and slow systems. Right now, some of the most energy- and cost-efficient ways to chill the air in centers rely on water. Bluefield Research has estimated data centers use more than a billion liters of water per day, including water used in energy generation. That’s enough to supply 3.3 million people for a day.

One 2023 study estimated that a conversation with ChatGPT consisting of roughly 10 to 50 questions and answers, requires a standard 16.9-ounce bottle of water. Training only one earlier AI model behind ChatGPT was estimated to have consumed almost 200,000 gallons of water. Making matters worse: Much of the water is of drinking quality, to avoid environmental problems and equipment failure. (…)

Twice as Much Internet Bandwidth

The large language models that ­underpin generative AI learn by digesting huge amounts of data over the internet, and users of AI tools in turn will only add to the demand. AT&T Inc. CEO John Stankey said in May that the network’s wireless demand was already up 30% a year and won’t slow with AI raising usage. “If you’re going to continue to see usage go up 30% to 35% a year, you’ve got to build bigger highways to take that,” he said.

Over the past five years, the network traffic growth at Verizon Communications Inc. has more than doubled thanks to people watching and streaming videos, Verizon Consumer Group CEO Sowmyanarayan Sampath said in an interview at about the same time. In the next five years, he predicted, growth will double yet again because of prompts and data fed into AI models.

AI, he said, “is the next growth machine for us.” Tech com­panies are so hungry to lock in fiber ­networks right away that the telecommunications company Lumen Technologies Inc. in August announced that it had secured $5 billion (and was in discussions to land another $7 billion) in new business tied to AI-driven demand for connectivity.

Real Estate for Thousands of Data Centers

Globally, there are more than 7,000 data centers built or in various stages of development, up from 3,600 in 2015. And that still probably won’t be enough. The demand for data center services had been growing dramatically even before ChatGPT, mostly because companies have been increasingly moving their data processing off-premises and turning to remote cloud services. And every major country wants its own homegrown AI hubs, touching off a global infrastructure race.

Data centers require land. For reference: The data-­center-focused real estate investment trust Equinix Inc. bought 200 acres for a multihundred-­megawatt campus. Another company recently signed a lease development agreement on 2,000 acres for a gigawatt-size one.

Finding land that works just right for the power requirements of a data center is tough, leading to bidding wars. These complexes also need construction materials and crews to install all of it. Material is on backorder, and there’s a shortage of workers. Meanwhile, Venturo of cloud services provider CoreWeave says some of his clients want him to devote entire campuses just to their business. “The market is moving a lot faster than supply chains that have historically supported a very physical business have been set up to do,” Venturo says.

Chips, Chips, Chips

Graphics processing units, or GPUs, are the workhorses for training AI models. They’re designed to handle thousands of tasks simultaneously, a concept known as parallelism. A data center may use hundreds or even thousands of these processors, each one costing more than a family car. Virtually every major tech company came up short on this type of chip when the generative AI boom first took hold. Microsoft and Google were among those that cited low GPU inventory in past earnings calls as a challenge.

Silicon, Steel, Quartz and Copper

Many of the items above require metals and minerals. Consider silicon, the foundation for chips, circuits and processors. China is the largest producer of raw silicon and refined silicon materials, which has raised concern as tensions between the Asian nation and the US and its allies rise.

The most recent supply chain scare surfaced in North Carolina. In October, Hurricane Helene, in addition to killing dozens of people and stranding others across the eastern US, disrupted operations at two mines in the state that together produce about four-fifths of the highest-quality quartz. It’s used to create crucibles where silicon is heated, melted and reformed into the single-­crystal ­structure that makes an ideal base to manufacture semiconductors.

Semiconductors contain gold, silver, aluminum and tin. There’s enough of these metals to keep the factories humming. But two obscure chip metals have emerged as potential bottle­necks: gallium and germanium.

In December, China announced a ban on exports of the metals to the US—part of an escalating tech war. Copper is in everything including chips, data centers, electrical equipment and cooling units, potentially setting the stage for a clash between the demands of AI, renewable energy and electric transportation. And then there’s steel, which is critical to building data centers and for infrastructure such as cables.

More People Than You Think

Much has been said about the jobs that AI may eliminate. In February, the Swedish fintech company Klarna Bank AB made waves after saying its AI assistant was doing the equivalent work of 700 full-time customer service agents. Global research and analysis companies have warned that employment in sectors such as finance, law and customer service will be hard-hit. The International Monetary Fund has estimated that AI could replace, or augment, almost 40% of employment globally.

But AI companies themselves directly employ millions of people today. Among AI workers are computer scientists, data architects, researchers, mathematicians, software engineers, chip designers, product and program managers, and compliance attorneys. That’s not to mention the armies of in-house analysts, marketers and salespeople. In early November, Salesforce Inc. announced plans to hire more than 1,000 workers to sell its new generative AI product.

Talent bottlenecks have emerged across much of these professions amid the rush to recruit for AI. Tech investors and AI startups have lamented the lack of properly educated and experienced candidates. The phrase “AI-vies”—a play on the Ivy League—has emerged in Silicon Valley to refer to a few companies (among them, Alphabet, Microsoft and OpenAI) that have trained the talent everyone else wants to poach. Even more have been recruited abroad, in countries such as India, to build and clean up the high-quality datasets necessary to train AI systems.

More (Good) Data Than the World May Have

Generative AI models need ­high-­quality data the way human beings need food. Large language models are “trained” by ingesting text that’s broken down into small units called tokens. From this text, LLMs identify patterns that help predict—in a process repeated over and over again—the text that should follow another set of text. The world’s foremost LLMs were trained off more than a trillion tokens each. To put that in context, consider that 2,048 tokens is roughly equivalent to 1,500 words. Estimates for exactly how many tokens of cumulative text data exist in the world are all over the place, ranging from a few trillion to thousands of trillions.

Amazingly, this abundance of data might not be enough to keep AI development moving forward as quickly as some hope. Some of the world’s most power­ful AI model developers such as OpenAI are already finding it increasingly difficult to locate new, untapped sources of high-quality, human-made training data to advance their models.

There’s limited data in non-­English languages and even less that isn’t focused on Western or White communities. This lack of diversity threatens to result in AI products that show bias against minorities, women and other underrepresented populations. A Bloomberg analysis this year, for example, found that the underlying AI model behind ChatGPT shows bias against certain racial groups based on names alone when ranking resumes. OpenAI says that the results may not reflect how its customers use its models and that it works to identify potential harms.

Producers of data and content, from media organizations to financial institutions, are waking up to the fact that their information is increasingly valuable to AI developers. Hollywood actors and writers went on strike in 2023 to protect their craft from the technology. The New York Times as well as major record labels are suing AI companies for training their data on copyrighted work. AI companies say that training on publicly available materials is a legally permitted fair use.

In a recent call with investors, S&P Global Inc. CEO Martina Cheung summed it up: “A large language model is only as good as the quality and quantity of data that it’s trained on, and we have lots of high-quality data.” Just in the past year, OpenAI has struck deals to use content from News Corp., Condé Nast, Hearst, Reddit, People magazine publisher Dotdash Meredith and Axel Springer.

Tech companies are experimenting with training models on “synthetic” datasets, content created by AI itself. In theory, this helps AI companies meet their bottom­less need for data while avoiding the legal, ethical and privacy-­related concerns surrounding scraping information from the web.

But some researchers have warned that AI models may “collapse” if they’re trained on content generated by AI rather than humans. One 2023 paper on so-called model collapse showed that AI images of humans became increasingly distorted after the model retrained on “even small amounts of their own creation.”

Or Maybe Less of Everything Than Some Fear. Or Hope

Investors, data center operators, energy companies and other businesses are pouring hundreds of billions of dollars collectively into different parts of the supply chain that feeds AI. Every major bank and private financier is positioning itself for a piece of an estimated $1 trillion in spending on AI ­infrastructure. The capital ­expenditures of Alphabet, Amazon, Meta and Microsoft are set to collectively exceed $200 billion in 2024.

An S&P 500 utility-­sector index has gained 22% over the past year, and data center-­focused REIT Equinix has almost doubled its market cap since late 2022. Nvidia shares have surged by nearly 700% over the past two years, turning the company into one of the most valuable ones on Earth.

And yet, in the end, nobody knows whether AI will keep booming. Some Wall Street analysts are starting to predict an end to the frenzy. Investors have begun questioning whether Big Tech’s heavy spending will ever result in the AI profit machine they’d envisioned. Arguably the biggest threat to the hundreds of billions of dollars being invested in AI is that the world’s most advanced model developers and their suppliers have grown obsessed with efficiency.

Gary Dickerson, CEO of chip equipment maker Applied Materials Inc., told investors in November that some AI companies are aiming for “100x improvements” in efficient computing within five years. Others are shooting for 10,000-fold gains in 15 years, he said. Efficiency, Dickerson said, “is emerging as a unifying driving force for the industry.”

See also Power Play.

Of course, the race to the best LLMs and apps will result in overspending and eventual write-downs or write-offs. But “AI will keep booming” until usage becomes near ubiquitous. Far from there yet. Not a reason to invest blindly, but a reason to invest smartly.

If you missed it, listen to this Jensen Huang interview.

Huang was also at Goldman’s Communacopia + Technology conference in San Francisco.

Huang explained how computer graphics, for example, rely heavily on AI infrastructure. “We compute one pixel, and we infer the other 32,” he says in an edition of Goldman Sachs Talks. “Computing one pixel takes a lot of energy. Inferring the other 32 takes very little energy, and you can do it very fast. And the image quality is incredible.”

Given this speed and flexibility, this infrastructure more than pays for itself, Huang says, responding to a question from Solomon about returns on investment for customers. By spending on such equipment, “the computing cost goes up a little bit — maybe it doubles,” Huang says. “But you reduce the computing time by a factor of about 20. You get 10x savings.”

Goldman Sachs:

AI adoption by firms remains modest, with only 6.1% of US firms currently nusing AI to produce goods or services (vs. 5.9% in Q3). Within industries, finance and insurance firms reported the largest increases in adoption rates while information, education, and manufacturing firms reported a decline. Publishing and filmmaking-related firms report the largest expected increase in AI adoption over the next six months.

Large firms with 250+ employees have the highest adoption rate (10%) and report the largest expected increase in adoption over the next 6 months (+6pp to 16%), while adoption rates are also above the national average for small firms with 1-4 employees (7%). Recent industry surveys suggest adoption has doubled in small to midsized businesses (SMBs) over the past year, but many firms are still concerned about cybersecurity and identifying profitable use cases.

AI’s impact on the labor market has increased slightly, and AI-related job openings now account for 20% of all IT job openings and 1% of all job postings. AI-related corporate layoffs remain limited overall, but the unemployment rate for AI-exposed positions is still slightly higher than the broader unemployment rate.

We continue to see large impacts on labor productivity in the limited areas where generative AI has been deployed. Academic studies imply a 23% average uplift to productivity, while company anecdotes imply slightly larger gains of around 30%.

Productivity Gains from Generative AI Adoption

Goldman Sachs Global Investment Research

Unfortunately, no data above of the biggest elephant in the AI room, China.

Here’s some:

China’s generative AI users reach 230 million as start-ups, Big Tech roll out LLM services

The number of generative artificial intelligence (GenAI) users in China reached 230 million at the end of June, as a crop of start-ups and Big Tech firms rushed to offer their large language model (LLM) services, according to government data.

That means around one in every six users in the world’s biggest internet market are using a GenAI product, according to a report released on Saturday by the China Internet Network Information Centre (CNNIC), a state-run agency.

Nearly two-thirds of GenAI users on the mainland employ LLM services to answer questions, while one-third rely on them as working assistants to generate meeting transcripts and slides, the report showed.

As of November, more than 309 GenAI products had registered their LLMs with the Chinese internet regulator, with Beijing accounting for 96 and Shanghai 84, according to the report.

All major Big Tech firms in China have launched their own LLMs for consumers as well as enterprise uses, while the rise of start-ups, including the four so-called AI ‘tigers’ Baichuan, Zhipu AI, Moonshot AI and MiniMax, is drawing increasing attention from users and investors.

The nearly 200 commercially available LLMs in China have attracted over 600 million registered users, according to statistics cited by an official of the Ministry of Industry and Information Technology in October.

GenAI refers to algorithms that can be used to create new content, including audio, code, images, text, simulations and videos. LLMs are the technology behind GenAI services like ChatGPT.

David’s research indicates that there are currently 4 reasoning models in the world

• OpenAI o1
  DeepSeek
  QwQ
  Marco 01

“Last 3 are Chinese, all roughly equivalent or superior.”

According to Goldman Sachs Research, if Nvidia represents the first phase of the AI trade, Phase 2 will be about other companies that are helping to build AI-related infrastructure. Phase 3 deals with companies incorporating AI into their products to boost revenue, while Phase 4 is about the AI-related productivity gains that should be possible across many businesses.

  Porsche holding company warns of writedown in Volkswagen stake of up to €20bn

The FT writes:

• Porsche SE expects to write down its stake in Volkswagen by up to 40 per cent, as the uncertainty over potential plant closures and strikes forced Europe’s largest carmaker to withhold its annual financial plan.
• VW’s flagship brand now sells roughly 500,000 fewer cars annually than it did before the pandemic.
• Since then, the brand’s share of sales in China — its most profitable market — has nearly halved, amid a consumer shift towards electric vehicles and hybrids and growing competition from local competitors there such as BYD.
• IG Metall has warned that if VW does not abandon its plan to close factories, strikes will become more intense.