Tesla’s top battery cell supplier, CATL, is throwing some cold water on Tesla’s battery plans and the CEO even said that Elon Musk “doesn’t know how to make battery cells.”

Contemporary Amperex Technology Co., Limited (CATL) had an incredible rise and became the world’s largest producer of battery cells for electric vehicles in the last few years – and by a significant margin.

It even supplies Tesla with many battery cells for its EV production at Gigafactory Shanghai.

CATL’s success has made Robin Zeng, its founder and chairman, one of the foremost authorities on battery cell production, which makes his new comments on Tesla’s battery cell production effort interesting.

Tesla buys most of its batttery cells from suppliers, inlcuding CATL and Panasonic, but it has also launched its own effort to produce its own cylindical 4680 battery cells, which are currently only used in the Cybertruck.

The company has had issues ramping up production of its own battery cells, but it reported making progress lately – going as far as claiming that its own cell will achieve better cost than anyone in the industry by the end of the year.

But CATL is not impressed.

Zeng spoke to Reuters recently and reported that he had a discussion with Musk earlier this year when he visited China. The CATL founder warned Musk that he thinks Tesla’s 4680 effort will fail:

Zeng said he had told Musk directly that his bet on a cylindrical battery, known as the 4680, “is going to fail and never be successful.”

The CATL founder, who has a PhD in physics, was also unimpressed with Musk’s electrochemistry knowledge when debating Tesla’s 4680 batteyr cell effort:

“We had a very big debate, and I showed him,” Zeng said. “He was silent. He doesn’t know how to make a battery. It’s about electrochemistry. He’s good for the chips, the software, the hardware, the mechanical things.”

In this interview, Zeng was very candid about his chat with Musk. He even touched on Musk’s notirous issues with timelines.

He commented:

“His problem is overpromising. I talked to him,” Zeng said. “Maybe something needs five years. But he says two years. I definitely asked him why. He told me he wanted to push people.”

This has been the excuse that many Musk fans have been using to justify his missed timelines: he is trying to motivate his troups.

Electrek’s Take

I am surprised that the head of a supplier would talk about the CEO of one of his main customers like that.

Obviously, he is biased since Tesla’s battery effort could cut into his business, but at the same time, Tesla has always made it clear that they would always need to keep buying from battery suppliers.

I think what is most interesting here is that CATL’s expertise is in LFP cells and we know that Tesla is looking to make its own LFP cells at one point. That might be what Zeng is talking about here.

With that said, Tesla did claim that it is on the verge of accomplishing its cost target with the 4680 cell. It might be true, but I have issues believing some of the things Tesla claims these days.

Tesla (TSLA) is soaring in anticipation that Trump’s administration will make an easier path for Tesla’s self-driving tech, which still doesn’t work, to be approved federally.

Currently, self-driving technology is addressed at the state level, with each state having its own regulations for approving self-driving systems on its roads.

During a conference call following Tesla’s last earnings results, CEO Elon Musk, who has been financially backing the reelection of Donald Trump and “fully endorsed” him, hinted that he could work with the new federal government to get a federal self-driving approval process going.

Now, Bloomberg reports that Trump’s transition team is discussing making it a priority:

Members of President-elect Donald Trump’s transition team have told advisers they plan to make a federal framework for fully self-driving vehicles one of the Transportation Department’s priorities, according to people familiar with the matter.

This news sent Tesla’s stock up 7%, or an increase of 470 billion in value.

That’s surprising because before now, the regulatory aspect of Tesla’s self-driving effort didn’t seem like the biggest hurdle – making the technology work still seems to be the biggest hurdle.

Tesla has been wrong about its self-driving timeline too many times to count, but the latest one is to release unsupervised self-driving in California and Texas in Q2 2025.

Ashok Elluswamy, the head of FSD at Tesla, stated that Tesla’s goal is to achieve over 600,000 miles between critical disengagement, which is based on NHTSA’s data of accidents between human-driven miles.

Tesla has not released any data about its self-driving effort, and therefore, the best data available is crowdsourced. That data currently shows about 241 miles between critical disengagement:

Tesla would need a 2,500x improvement in miles between disengagement to reach a safer-than-human level, which has been the goal before getting regulatory approval.

Electrek’s Take

That sounds like a much bigger hurdle than getting regulatory approval.

I actually agree with the Trump administration that it makes more sense to have a federal framework for approving self-driving systems than at the state level.

But I don’t see how it will help Tesla since there’s no clear path to Tesla achieving a level safer than human with their current approach any time soon.

At the current pace, the 2,500x improvement would take 10 years and we have yet to see a significant acceleration to the pace of improvement.