The Sizewell A and B nuclear power stations, operated by Electricite de France SA (EDF), in Sizewell, UK, on Friday, Jan. 26, 2024. Photographer: Chris Ratcliffe/Bloomberg via Getty Images
Bloomberg | Bloomberg | Getty Images
LONDON — Surging power demand has reignited interest in nuclear energy, but vast capital requirements and an uncertain political and regulatory climate raise questions about the sector’s fiscal capacity.
AI and data centers are the “canary in the coal mine,” World Nuclear Association Director General Sama Bilbao y León told CNBC ahead of the conference. “We are finally recognizing that the demand of electricity and energy in general is only going to increase. But the reality is that all sectors of the economy are going to need more electricity.”
In addition to AI, applications range from nuclear energy for the metallurgical industry, which is looking to electrify as fast as possible, to the chemical, maritime and shipping sectors, León said.
The question of how to meet the world’s growing power needs took center stage as chief executives of the world’s biggest uranium and nuclear energy firms, experts and investors gathered for the annual World Nuclear Association (WNA) symposium at the Royal Lancaster London hotel last week.
Opening remarks from Dr Sama Bilbao y León, director general of the World Nuclear Association, at the 2025 conference.
World Nuclear Association
Kicking off discussions at the conference, Leon told attendees in her welcoming speech that the event is a “working summit” looking to move past mere conversation.
Investments in the nuclear value chain through 2025 are projected to increase to $2.2 trillion, according to Morgan Stanley estimates, up from a 2024 forecast of $1.5 trillion. That level of investment raises questions over the role of government, banks and other financial players in providing sufficient fiscal capacity.
Investment challenges
Nuclear energy is said to provide a more reliable, 24/7 energy source compared to renewables, which can be more intermittent. The development of small modular reactors (SMRs) provides a more scalable power solution due to their size. According to the IEA, the payback period of a SMR investment is half the usual 20 to 30-year period for larger scale projects.
But SMRs have yet to reach the commercial stage, and most planned projects won’t come online until 2030. While a significant amount of money is being pledged, there have been no new large-scale nuclear projects in the U.S. in the last 15 years.
“The first positive story with respect to the financial sector with regards to nuclear, is that they are open to financing nuclear,” Mahesh Goenka, founder of market and commercial advisory firm Old Economy, told CNBC on the sidelines of WNA. “That was not the story a few years ago when a lot of banks didn’t want to touch nuclear projects. That has changed. The question now remains, do they have the risk appetite to finance nuclear projects?”
Challenges include over-running budgets, the late delivery of projects due to long construction lead times, the technical complexity of initiatives and difficulties obtaining licenses.
Goenka compared the West to China, where financial institutions are happy to finance nuclear projects because they can be delivered on time and on budget — leading to better margins than on other infrastructure projects. Meanwhile, the West has not built many new reactors in a very long time, so the learning rate is not quite there yet, he said.
Nearly all of the nuclear generating capacity in the U.S. comes from reactors built between 1967 and 1990, with no new constructions until 2013 when work started on the Vogtle units in Georgia. Meanwhile, the last plant to be built in the U.K. was Sizewell B, which started operating in 1995.
Nuclear investments are “inherently political projects,” said Mark Muldowney, managing director of energy, resources and infrastructure at BNP Paribas. He noted that, while clients are much more receptive to the investments, uncertainty over cost and build time remains.
“We are many years away from the situation in which techniques like project finance can be used by themselves to finance large nuclear [projects],” he said during a panel discussion.
“It’s not going to be the contractors, even if they were willing to, and by and large they aren’t, they will be bankrupted by some of the risks that sit with these projects. So it’s either going to be a government, or it’s going to be the electricity consumers of that country, and in some places that could be intermediated by utilities.”
Government backstop still required
Nuclear power plants are among the most capital intensive assets. The U.K., for example, has greenlit the construction of a massive two-reactor nuclear power station on the Suffolk coast that will generate 3.2 gigawatts of electricity — enough, the government says, to provide power for the equivalent of 6 million homes. But costs of the majority government-owned project have jumped to £38 billion, exceeding an initial target of £20 billion.
Trevor Myburgh, senior executive in corporate finance advisory at Eskom, stressed that the private sector cannot be a “silver bullet” and solve the problem of financing nuclear energy.
Public private partnerships are going to be “crucial” in the development of nuclear, particularly in any emerging economy, Myburgh said during a panel discussion on Wednesday.
While some European countries such as Switzerland — which currently has a ban on the construction of any new nuclear plants but has drafted legislation to lift this motion — and Germany remain adverse to nuclear energy, other governments such as those of the U.K., France, and the U.S. have leaned into the energy source.
Earlier this year, U.S. President Donald Trump signed a number of executive orders designed to fast track the development of nuclear reactors and quadruple nuclear generating capacity by 2025.
Such actions from Trump’s administration have put positive nuclear energy policies “on steroids,” said Uranium Royalty Corp CEO Scott Melbye.
“What we’re seeing are really concrete measures being taken by this administration to spur not only the building of small modular reactors, advanced reactors and large reactors, but [also] in the fuel cycle,” Melbye told WNA attendees.
Investor Arfa Karani noted the growing interest from the investor community to find opportunities with startups, particularly those that supply nuclear-adjacent tech.
The U.K. government, in particular has adopted a more “hands-on” approach in helping founders understand how to invest in clean tech, she said.
“The regulation has to figure itself out. It’s no longer a question of, where do we get the capital from? ….because now suddenly it’s become a matter of national security and global power and global dominance,” she told CNBC, adding that commitment Stateside to funding AI and nuclear has meant that “all the insolvable problems suddenly becomes solvable which is very exciting for nuclear.”
Tesla’s Full Self-Driving (FSD) v14, its first major update in a year, disappoints as data points to a lower increase in miles between disengagements than expected.
The system also features new hallucinations, brake stabbing, and excessive speeding.
The update has been highly anticipated for several reasons.
Advertisement – scroll for more content
First off, it has been a year since Tesla released any significant FSD update to customers, as it focused on its internal robotaxi fleet in Austin. The update is believed to feature improvements developed through Tesla’s robotaxi fleet, which requires supervising like its consumer FSD.
Secondly, CEO Elon Musk has claimed that Tesla still plans for “Supervised Full Self-Driving” to become unsupervised by the end of the year in consumer vehicles. For that to happen, we needed to see a massive improvement from v13 to v14.
As I previously reported, I anticipated an improvement in miles between critical disengagements from ~400 miles in v13 to ~800 to 1,200 miles in v14. It would be a significant improvement, but still way short of what’s needed to make FSD unsupervised.
Tesla notoriously doesn’t release any data about its FSD program. Musk has literally told people to rely on anecdotal experiences posted on social media to gauge progress.
Fortunately, there’s a crowdsourced dataset that gives us some data to track progress with miles between critical disengagement. It’s far from perfect, but it is literally the best data available, and Musk himself has shared the dataset in the past – albeit while misrepresenting it.
In the last week, Tesla started pushing the FSD v14 update (now v14.1.4) to more owners – resulting in more crowdsourced data and anecdotal evidence.
With now over 4,000 miles of FSD v14 data, miles between critical disengagement sits about 732 miles – below the lower end of our expectations:
Tesla would need to be closer to 10,000 miles between critical disengagements to allow unsupervised operation, and even then, it would likely be in geo-fenced areas with speed limitations.
This is unlikely to happen by the end of the year, as Musk predicted, as FSD v14 appears to have some significant issues still.
First off, many FSD v14 drivers are reporting that the update is having problems with hallucinations where the car decides to stop on the side of the road seemingly randomly:
I had a great first day with FSD v14.1.4! Multiple hands-free, point-to-point rides with no disengagements. A few “brake stabbing” events were slightly annoying but not dangerous.
It does seem like FSD v14 sometimes misinterprets other vehicles’ turn signals as emergency vehicle lights and pulls over.
In some cases, FSD v14 has been known to completely disable FSD features inside vehicles:
V14 is getting really annoying. Summon continued to fail with an error condition. Now in the car and FSD just isn’t functional. I don’t get it. Come on, Tesla. My brother got the same issue yesterday where his facia camera was nonfunctional and went black, obviously disabling FSD… pic.twitter.com/D8H8DY83sC
Many FSD v14 drivers have also reported an increase in “brake stabbing”, where the vehicle seems to hesitate and frantically applies the brakes and releases them – resulting in a stabbing motion.
Now, I don’t want to hear anything about my use of anecdotal evidence and crowdsourced data. That’s literally the best data available for FSD.
Unlike virtually all other companies developing self-driving technology, Tesla refuses to release any.
If it were to release some data, I’d be happy to use it.
One thing is clear from v14 so far: unsupervised FSD in consumer vehicles is not happening in any meaningful way this year.
I expect significant improvements in upcoming FSD v14 point updates. Maybe enough to get it to my previous expectations of ~800 to 1,200 miles between disengagements, but that’s about it.
Finally, while I generally don’t count on NHTSA to enforce any rule in any significant way when it comes to Tesla’s “Full Self-Driving” effort, I think they might actually do something about “Mad Max.”
This video on Instagram has 4.5 million views, and it shows extremely dangerous driving behavior at up to 90 mph (145 km/h)
I think the authorities will have to intervene here, because it makes no sense for an unproven autonomous driving system to be able to operate under those parameters.
FTC: We use income earning auto affiliate links.More.
Toyota’s best-selling car is finally going electric. The Corolla EV looks more like a Porsche or BMW than the Toyota vehicles on the road today, but that’s just the start.
The Toyota Corolla is evolving into a rad-looking EV
After revealing the Corolla Concept for the first time at the Japan Mobility Show on Tuesday, Toyota’s CEO, Koji Sato, said the compact car has always been “a car for everyone.”
Since it hit the market over 50 years ago, Toyota has sold well over 50 million Corollas. The Corolla even surpassed the VW Beetle in the 90s to become the world’s best-selling vehicle. Like the Prius, Toyota’s compact car lured in buyers with an affordable price and a reputation as a reliable daily driver.
Although it’s still a top-seller, the Corolla has lost some of its charm as more advanced, stylish, and efficient electric cars hit the market.
Advertisement – scroll for more content
Toyota looks to change that with a drastic overhaul that takes the Corolla to the next level. To stay relevant, Sato asked the crowd at the event, “How should the Corolla evolve?”
Toyota CEO Koji Sato reveals the Corolla Concept at the Japan Mobility Show (Source: Toyota)
We all want to drive a car that looks cool, but there’s much more to it nowadays. Buyers are increasingly seeking more efficient vehicles with the latest software, connectivity technology, and other features.
“Whether it’s a battery EV, plug-in hybrid, hybrid, or internal combustion engine vehicle―whatever the power source―let’s make good-looking cars that everyone will want to drive!” Toyota’s CEO said, adding the car is “packed with inventions aimed at making that a reality.”
Although Toyota didn’t confirm the concept was headed for production, the next-gen Corolla is expected to arrive with a similar style.
The concept still features Toyota’s newest design elements, like the “hammerhead” front end, but with a bit more of a futuristic feel.
You can barely tell the concept is a Corolla, aside from the massive COROLLA badging on the rear. Toyota didn’t reveal any powertrain details, but the charge port and closed-off grille suggest it’s an EV.
The next-gen Toyota Corolla is expected to be offered as an EV, a plug-in hybrid, a hybrid, and, likely, still an ICE variant.
FTC: We use income earning auto affiliate links.More.
Three months after Uber, Lucid Motors, and Nuro announced a partnership that would enable Gravity SUV robotaxis, the rideshare network has shared where the public will first be able to hail one. Spoiler alert, it’s easy to guess if you give it half a thought.
As we reported in July, Uber Technologies committed to a $300 million investment in Lucid Group (parent company of American EV automaker Lucid Motors), to deploy at least 20,000 Lucid vehicles as robotaxis over the next six years.
Those Lucid vehicles, which will consist of the automaker’s flagship Gravity SUV to begin, will hit public roads equipped with a Level 4 autonomous system called Nuro Driver. Nuro, the third partner in this equation, is a robotics company specializing in zero-occupant delivery vehicles, which garnered an existing partnership with Uber Eats as well as a “hefty” (yet undisclosed) investment from Uber Technologies.
Last month, Lucid delivered its first Gravity SUV to Nuro to begin the retrofitting process of the Nuro Driver system to support Uber’s hopes for a luxe robotaxi fleet. While the partners continue to work toward building an exciting new fleet of Lucid Gravity Robotaxis, Uber has shared the location where they will first go into service… Casper, Wyoming.
Advertisement – scroll for more content
Just kidding!
It’s the San Francisco Bay Area, of course.
Lucid Gravity SUV fitted with Nuro’s self-driving tech (Source: Lucid)
Uber to deploy Lucid Gravity EVs in Bay Area in 2026
Today’s update from Uber expands upon the ongoing partnership with Lucid Group and Nuro. According to the companies, the San Francisco Bay Area will be the first market where riders will see this next-generation autonomous robotaxi program in operation. That milestone is expected sometime in 2026.
Uber has shared that it has been updating policymakers and regulators at every level on the progress of its exclusive Lucid Robotaxis and continues to meet the operational requirements. Notably, Uber has shared that on-road development with the Lucid Gravity robotaxi engineering fleet is already underway in the Bay Area.
Furthermore, Nuro and Lucid intend to be operating over 100 Gravity robotaxis as part of the test fleet “in the coming months.” Lucid interim CEO, Marc Winterhoff, spoke about today’s announcement:
Lucid has always celebrated its California roots, and we’re thrilled to make the San Francisco Bay Area the first market for our new robotaxi on the Uber platform, powered by the Nuro Driver. Beginning next year, riders will experience a level of convenience, safety, and comfort unlike anything else on the road. We can’t wait to bring this service to life and expand it to communities across the country.
To build this fleet of Uber-exclusive robotaxis, the required hardware will be integrated into Lucid Gravity SUVS while they are still on Lucid’s assembly line in Arizona. Those builds will then be integrated with Nuro’s proprietary software when Uber officially commissions them.
All eyes on 2026 as we now know that residents around the Bay Area will be able to hail a driverless Lucid Gravity through the Uber platform. I’m very much looking forward to seeing this fleet in action.
FTC: We use income earning auto affiliate links.More.