TheChek News Flow

Goldman Sachs’ annual private tech conference this week was, from one angle, a study in East Coast realism versus West Coast optimism. Spreadsheets versus innovation. Bankers’ slacks versus Palmer Luckey’s shorts.

One of the most high-profile founders on stage, Perplexity’s Aravind Srinivas, cited the satirical TV show “Silicon Valley” as an inspiration during his start in the tech industry. A big Chicago-based fund manager, minutes later, openly fretted about whether artificial intelligence investments would really provide decent returns.

The tension between the two mindsets is, of course, more nuanced than this. Wall Street loves AI plays like Nvidia and big tech. And even venture capitalists are getting more skeptical of early-stage AI valuations, as we’ve written.

To me, the upshot of the conference was that investors are generally optimistic about markets but still fret about how exactly to make the right startup bets. Nearly everyone expects a steady uptick in IPOs and M&A to help the investments they made five to 10 years ago. The latest vintage of funds, however, is still a jump ball.

A few slides senior Goldman tech bankers Jane Dunlevie and Ryan Nolan presented encapsulated some of these strange realities hovering over tech markets. The volume of tech mergers of $500 million or larger is up 59% this year, about half the volume in 2021. But private equity firms are doing more than half the deals, way above historic norms.

Big tech, in part because of antitrust, has been quiet on the acquisition front. Cash spending from big tech firms on M&A was down 56% this year. Meantime, their capital expenditures, largely to build data centers, more than doubled. “For big tech, is capex the new M&A?” a slide read.

The question at least adds depth to the current understanding of the stalled M&A activity in tech. Maybe antitrust isn’t the entire problem. The AI cycle has tech firms preferring to put their money into infrastructure rather than AI startups, which are rapidly gaining traction and then just as quickly losing steam.

“It’s still a new technology, and it’s hard to pick winners,” Nolan told me.

DAX:

MDAX:

SDAX:

Since Donald Trump won reelection as president, many investors have assumed that the chief beneficiaries would include businesses run by Elon Musk, his biggest financial backer and recently his frequent companion. Investors have pushed up the share price of Musk’s one listed business—Tesla—36% since the Nov. 5 election.

But the Tesla ventures that Wall Street investors most expect to prosper from White House favor—artificial intelligence plays such as self-driving Robotaxis—are long-shot bets that rely more on technological breakthroughs than on friends in government. The Trump administration can loosen federal oversight of autonomous vehicles, but experts say Tesla’s far bigger obstacle is getting its self-driving technology ready for the road.

“Any supposed regulatory hurdles are remote and trivial compared with Tesla's technical hurdles,” said Bryant Walker Smith, a law professor at the University of South Carolina who studies autonomous vehicles. “Or, to phrase it differently, any imminent regulatory hurdles are because Tesla does not have and cannot show a reasonably safe automated driving system.”

Musk himself appears to have a range of political reasons to seek proximity to Trump, beyond advantageous government treatment of Tesla, SpaceX and other businesses he controls. As for investors, their current fervor for Tesla shares may simply reflect a desire to invest in Musk’s relationship with Trump in the only way they can, and not a firm conviction of how or even whether Trump will help Tesla.

Either way, Tesla shares rose almost 8% on Monday after Bloomberg reported that Trump seemed likely to propose straightforward federal regulations for self-driving vehicles that would supersede current state-by-state rules. At the moment, the only federal rules for self-driving vehicles are safety regulations and a 2,500-car limit on the number of vehicles without steering wheels or pedals a company can deploy each year.

If the Department of Transportation issued a single set of rules for driverless cars, that could make it easier for Tesla to catch up with Waymo, owned by Alphabet, and General Motors’ Cruise, whose driverless cars already ply roads in cities including San Francisco, Los Angeles, Phoenix, Dallas and Austin, Texas. Waymo also takes paying passengers in Los Angeles, Phoenix and San Francisco. With federal rules, Tesla would no longer have to seek permission of regulators in those same states, as Waymo and GM did.

Last month, Musk said he would deploy self-driving Teslas in two stages: Next year, he said he would release Tesla Model 3s and Ys powered by Tesla’s autonomous Full Self Driving software; those would be capable of driving without human supervision—motorists could drive hands-free and not have to watch the road—but would still come equipped with pedals and steering wheels. In 2026, he said he plans to deploy driverless Tesla Robotaxis with no steering wheels or pedals.

One big Musk gamble is that Tesla can achieve nearly risk-free driving using just cameras and AI; almost all of Tesla’s rivals in the West and China use sensors such as radar and lidar in addition to cameras.

Analysts appear reluctant to say flatly that there is a chance that Musk’s bet on cameras and AI does not pan out and that he has to add other sensors onto Tesla vehicles. Musk himself has repeatedly urged investors who don’t believe Tesla will solve self-driving to take their money elsewhere, though last month he admitted that, in terms of timing for delivering on his promises, “I tend to be a little optimistic.”

A number of analysts are giving him latitude. In a Nov. 14 note to clients, Tom Narayan, an auto industry analyst with RBC Capital Markets, said he had just returned from a visit to Tesla’s Austin gigafactory. FSD still “has a ways to go,” he wrote. Even so, he added, the tour “gave us increasing confidence in Tesla’s ability to achieve its autonomy goals.” Narayan attributes 76% of Tesla’s value to FSD and Robotaxis. “In our view, Tesla FSD is the most robust autonomy software on American roads today,” he wrote.

Other analysts aren’t so sure. In a note to clients on Monday, Morgan Stanley analyst Adam Jonas said he does not expect mass deployment of Tesla’s autonomous vehicles until the mid-2030s. Musk winning federal regulatory oversight of self-driving technology will “not magically catapult Tesla FSD efficacy to the 99.999% (reliability) threshold where Teslas can drive themselves unsupervised,” tweeted Gary Black, managing director of the Future Fund and a Tesla investor.

Even if Musk does get there on his timeline over the next two years, it won’t be like the 2010s, when Tesla—once it got through production hell—profited handsomely because it was virtually the only electric vehicle on the road. In autonomous cars, he will face stiff competition here in the U.S. and also in China.

“Right now there are real automated vehicles carrying real people on real roads. None of them are Teslas,” said Smith, the University of South Carolina professor.

Nuclear startup Kairos Power received approval from the U.S. Nuclear Regulatory Commission to start construction on two test reactors in Oak Ridge, Tennessee. The permit marks a significant milestone for Kairos, which in October inked a deal with Google to provide 500 megawatts of electricity for its data centers.

The fluoride-salt cooled, high-temperature reactors are scaled down versions of what Kairos hopes to ultimately build to supply Google with electricity starting in 2030. And while the new reactors are technically test beds, Kairos intends to connect the power plant to the grid, spokesperson Ashley Lewis told TechCrunch.

The Hermes 2 reactors will be capable of producing 35 megawatts of heat each, and they’ll be connected to a 20 megawatt turbine to turn that heat into electricity. Kairos’ commercial-scale power plant will also feature two reactors capable of generating a combined 150 megawatts of electricity.

Kairos’ design differs from existing nuclear reactors in two key ways: The fuel is made of uranium coated in carbon and ceramic shells, which are intended to be durable enough to contain fissile material in the case of an accident. And the reactor isn’t cooled by water but by molten salt.

The small modular reactor (SMR) startup, which has received a $303 million award from the U.S. Department of Energy, has been working for years to refine its molten salt-cooling system. Fluoride salts’ extremely high boiling points allow the coolant to flow under low pressure. That means in the case of an accident, there won’t be any high-pressure, radioactive material waiting to burst forth should pumping systems fail. Plus, Oak Ridge National Laboratory says that, should power to the pumps fail, molten-salt reactors can rely on passive convection to move salt through the reactor to cool it.

Altogether, those features are enough to qualify Kairos’ designs as “Generation IV” reactors, a classification system created by an international organization backed by national nuclear agencies. The classification system is both vague and broad, so it’s hard to tell exactly how Hermes 2 might score on the rubric.

Kairos has been inching toward approval for the reactor design for the last year and a half. Hermes 2 passed its safety review with the NRC in July and its environmental assessment in August. All told, it took 18 months for the NRC to issue the construction permit, a relatively swift timeline compared with previous reactor permits.

Now the pressure is on Kairos to deliver on its promises. The company says it hopes to have the first reactor for the Google deal online in 2030 and the rest completed by 2035. In the world of nuclear power, a decade isn’t much time at all.

Officials say Beijing will remain ‘steadfast’ in resisting protectionist measures

China is willing to engage in “positive dialogue” on trade with the US under a Donald Trump administration, senior trade officials said, a day after Beijing introduced a swath of measures to fortify its exporters ahead of anticipated higher tariffs imposed from Washington.

At a press briefing on Friday, officials said Beijing would remain “steadfast” in resisting protectionist measures. They also pledged to maintain a stable exchange rate despite expectations that Trump’s policies, which include imposing 60 per cent tariffs on Chinese goods, could lead to a stronger dollar.

“China and the United States share strong economic complementarities . . . China is willing to engage in positive dialogue with the US,” Wang Shouwen, international trade representative and vice-minister of commerce, said when asked about the expected Trump tariffs. “At the same time, it remains steadfast in safeguarding its sovereignty, security and development.”

His comments came as Beijing on Thursday announced policies to support its exporters ahead of the start of the Trump administration in January, whose early cabinet appointments indicate it will be particularly hawkish on trade with China.

The commerce ministry pledged to guide Chinese banks in channelling more credit to the export sector and help companies with foreign exchange hedging. In addition, it would “promote the development of cross-border ecommerce” and encourage agricultural exports, helping companies to “actively respond to unreasonable foreign trade restrictions”.

As part of the measures, China would also “attract and facilitate cross-border exchanges of business personnel” through measures such as visa-free travel.

China relies heavily on manufacturing investment and exports to boost an economy that is suffering from weak domestic demand following a prolonged property downturn.

The country’s surging exports, which in dollar terms rose 12.7 per cent year on year in October, have ratcheted up tensions with trading partners from the US and the EU to developing countries.

Brussels accuses Beijing of failing to do enough to stimulate domestic demand and of not removing barriers for foreign companies operating in China or exporting to the Chinese market. China’s imports declined 2.3 per cent year on year in October.

Wang said China’s economy had “already demonstrated remarkable resilience” and that the previous round of tariffs initiated by the US had mainly been borne by American consumers.

Some economists have speculated that China could counter Trump tariffs by allowing a depreciation of the renminbi, which would make Chinese goods more competitive in foreign exchange terms.

If Trump’s tariffs and tax cuts prove to be inflationary, driving up the prices of goods in the US, that could increase the interest rate differential with China and also drive a weakening of the renminbi, they say.

But Liu Ye, director of the international department of the People’s Bank of China, said at Friday’s briefing that the central bank would ensure “the renminbi exchange rate remains fundamentally stable at a reasonable and balanced level”.

China’s President Xi Jinping has called for a stable exchange rate as the world’s largest exporter and manufacturer seeks to portray itself as a reliable trading partner.