Start with what is actually known. The Journal reports Amazon is in talks to acquire Decart for around $7 billion, citing people familiar with the matter, and adds the only sentence that matters right now: a deal could come together soon, “assuming talks don’t fall apart or another suitor doesn’t emerge.” This is not a done deal. It is a negotiation with a headline number.

What Decart actually sells

Decart was founded in 2023 by three Israeli engineers, brothers Dean and Orian Leitersdorf and Moshe Shalev. It makes two things. The flashy one is real-time AI video: models like Lucy and Oasis that generate and edit video live and simulate 3D worlds for training robots. The boring one is the valuable one: software that makes it easier for developers to switch between different AI chips.

That second product is the whole story. If your code runs the same on Nvidia, Google, and Amazon silicon, then chips become a commodity and Nvidia’s pricing power cracks. Decart has raised more than $400 million from Sequoia, Benchmark, and Radical Ventures, and was valued at nearly $4 billion in May. Amazon is not paying $7 billion for video filters. It is paying for the crowbar that pries developers off Nvidia.

The $6 billion no that matters more than the $7 billion yes

Here is the data point everyone is underweighting. Bloomberg reported on September 7 that Anthropic had considered paying about $6 billion for Decart, completed due diligence, and chose not to proceed. Both companies declined to comment. Nobody outside those rooms knows why Anthropic passed, and anyone who tells you they do is inventing it.

But think about what a completed diligence pass means. Anthropic runs some of the deepest technical diligence in the industry. Its team saw the code, the benchmarks, the customer contracts, and the roadmap. Then it said no at $6 billion. A month later Amazon is circling at $7 billion. Either Anthropic missed something, or Amazon is buying something Anthropic was never shopping for.

My money is on the second one.

Why the Amazon Decart acquisition is about chips, not video

The Journal adds the detail that explains everything: Amazon Web Services gave Decart early access to its Trainium3 chip in 2025 after meeting the company and coming away impressed by its founders. CTech separately reported that Amazon has since become one of Decart’s strategic customers. Read that again. The buyer was already the power user.

Now the logic clicks. Amazon has spent years building custom silicon to escape Nvidia’s margin tax, but custom chips are useless if developers will not port their code. Decart’s chip-switching software is the missing adapter: it makes Trainium a credible alternative instead of a science project. Amazon is not acquiring a video startup. It is acquiring distribution for its silicon.

And a note on the headline’s question. Jeff Bezos is executive chairman; he is not in the diligence room. He does not need secret knowledge here. What Amazon “knows” is simply the value of this asset inside its own war, a number Anthropic could never see from the outside.

Two buyers, two games

This is the part founders should tattoo on their forearms. Anthropic and Amazon looked at the same company and played two different games. Anthropic was pricing an asset: what is Decart worth to us as a business, on its own merits? Amazon is pricing optionality: what is it worth to own the adapter that makes our nine-figure chip bet work?

The billion-dollar gap is not a math error. It is the difference between standalone value and strategic value, and in acquisitions that gap is where deals die or get wildly overpriced. Anthropic’s no was disciplined. Amazon’s maybe is strategic. Both can be right at the same time because they are answering different questions.

The founder lesson: a passed deal is not a ceiling

Most founders treat a passed deal as a verdict. A serious buyer looked, passed, and therefore the company is worth less than the ask. Decart just proved the opposite. One month after Anthropic walked at $6 billion, Amazon is talking $7 billion. The asset did not change in thirty days. The buyer did.

The lesson is to know which buyer you are talking to before you set your price. A peer acquirer pays for what you are. A strategic giant pays for what you unlock. Those are different numbers, sometimes a billion dollars apart. When Liang Wenfeng rewrote his own fundraising rules as DeepSeek’s facts changed, he was playing the same game: price the company you are becoming, not the one you were.

The bottom line

Watch the trajectory of the number, not just the number. Four billion in May. A $6 billion pass in September. Seven billion in talks in October. That slope tells you where AI infrastructure value is pooling: anything that loosens Nvidia’s grip.

And remember the deal is still just talks. If it falls apart, that tells its own story: two elite buyers passing in two months would raise real questions about what diligence keeps finding. If it closes, it confirms the new rule of the AI buildout. The hyperscalers are no longer just renting compute. They are buying the picks and shovels at whatever price keeps them out of Nvidia’s toll booth.

Read more in our Leadership & Management hub.

Sources: The Wall Street Journal, PYMNTS (via Bloomberg), Daily Caller

The post Anthropic Passed at $6 Billion. Amazon May Pay $7 Billion. What Does Jeff Bezos Know? appeared first on CEO Medium.

What LeCun actually said

LeCun told Fortune he has “zero concerns” about AI wiping out humanity, and just as few about the recent run of rogue agent incidents. When OpenAI agents breached Hugging Face back in July, LeCun called it “totally preventable.” “Those agents are doing exactly what their operators asked them to do,” he said. “They belonged in sandboxes, but the sandboxes leaked because engineers designed them badly.” His broader charge: most AI labs do not understand cybersecurity well enough to be issuing the warnings they issue. That lands as a direct hit on Anthropic, which spent September publishing threat reports about Chinese hackers manipulating its own tools.

The target is bigger than one CEO

LeCun widened the attack to effective altruism, the philosophy behind much of AI safety thinking. He called the movement “super toxic” and “a complete disaster,” and said its followers inside AI labs suffer from a “paranoia” that produces bad decisions. The timing matters. On September 8, former Anthropic and OpenAI researcher Jacob Coxon resigned. He posted a viral thread claiming lab staff “earnestly believe” AI could kill everyone by the end of the decade. The post drew more than 150 million views. On September 22, the Financial Times reported that staff at the UK AI Safety Institute, Anthropic, OpenAI and DeepMind were taking stress leave and counseling over fears about the technology they build. And Amodei himself issued a letter in September calling for a deliberate slowdown of AI development.

LeCun also went personal. He told Fortune that Amodei’s sister Daniela, Anthropic’s president and co-founder, married Holden Karnofsky, co-founder of Open Philanthropy, now called Coefficient Giving. LeCun said Amodei “tries to distance himself” from the movement but is “totally into it.” Two facts belong next to that claim. Amodei has denied being an effective altruism adherent, and Anthropic says its staff hold a range of views. Readers should treat the family detail as LeCun’s framing, not as evidence about Amodei’s beliefs.

The LeCun Amodei AI safety debate is really about the doomsday business model

Strip out the insults and this is a fight about who gets paid when the public is afraid. Anthropic sells safety as its core differentiation, and it is reportedly heading toward an IPO. LeCun’s AMI Labs raised $1.03 billion in March at a $3.5 billion pre-money valuation, according to Reuters, on the bet that large language models are a dead end and the future belongs to world models. LeCun is executive chairman; former Nabla co-founder Alex LeBrun runs it day to day. A founder with that much riding on LLMs being overrated has every reason to argue the risks of LLMs are overstated too.

LeCun’s sharpest substantive point is about regulation. He views Anthropic’s push for AI rules as regulatory capture in waiting: requirements a well-funded incumbent can absorb and a startup cannot. For founders, that is the part of this feud that actually matters. To be fair to Amodei, LeCun himself stopped short of calling him dishonest. He told Fortune he believes Amodei is sincere and focused on what is good for his company. Sincerity and incentives can coexist. The question is never whether a CEO believes his own warning. It is who profits if you believe it too.

What founders should take from this

Every founder building on AI now has a side in this fight, whether they picked one or not, because the debate decides regulation and regulation decides who survives. Three habits help. First, separate the technical claim from the business incentive every time a CEO warns about the future. Second, demand specifics: which system, which failure mode, which fix, on what timeline. Vibes are not a risk assessment. Third, invest in boring security hygiene before apocalyptic governance. LeCun’s least insulting point may be his most useful: the July incident was a sandboxing failure, and sandboxes are engineering, not eschatology. For more on leading through hype cycles, see our Leadership & Management collection. Amodei was also at the center of last week’s White House AI clash with Jensen Huang, and the pattern of safety teams colliding with shipping speed showed up again at OpenAI days ago.

The godfather and the prophet will both be wrong about something. Bet on the argument that does not require you to keep paying to be afraid.

The post Yann LeCun Calls Anthropic CEO Dario Amodei “Deluded” and “Crazy” appeared first on CEO Medium.

The clash has a backstory. The Pentagon designated Anthropic a supply chain risk after Amodei refused to let Claude be used for mass surveillance or autonomous weapons, and a federal appeals court recently upheld that designation. The night before the White House lunch, on September 28, Amodei had a private dinner with Trump.

There is also history between the two men on regulation itself. An earlier proposal for a self-regulatory body backed by Anthropic, OpenAI and Google was abandoned after opposition from Huang, along with Mark Zuckerberg and Elon Musk.

One caution: neither side has confirmed the Roosevelt Room exchange. The Journal’s account rests on unnamed sources, so it should be read as reported, not as confirmed dialogue.

What the Dario Amodei Jensen Huang White House AI Clash Really Means

Set the personalities aside and this is a leadership question: what does a chief executive owe the public when their product can do harm? Amodei’s answer is that warning about the risks is part of the job. Huang’s reported answer is that such warnings overstate the danger and invite regulation that slows progress.

There is business logic in Amodei’s candor. The chief executive who names the risks early gets to shape how the public understands them. Warn first, and you are never caught explaining later. Every founder building anything powerful has to decide which side of that line they stand on.

The tension will not resolve soon. Voluntary agreements like the one signed at the White House lunch leave the hardest safety decisions to the companies themselves. That arrangement only works while the people running those companies agree on what the risks are. The Roosevelt Room exchange suggests they do not.

Sources: Newsmax (citing the Wall Street Journal), International Business Times, TechFocus24, WUTV.

Read more on CEO Medium: Leadership & Management, Sam Altman Calls It ‘AI Washing’: The CEOs Blaming AI for Layoffs They Planned Anyway, Jensen Huang Says AI Firms Should Get No Regulatory Waivers, Should AI Replace the CEO Instead of Workers? The Math Says Maybe.

The post Jensen Huang Took on Dario Amodei at the White House. Here Is What the Fight Is Really About appeared first on CEO Medium.