Skip to content
Billionaires Digest
My team

Tuesday, October 20, 2026Oct 20, 2026 · morning edition

Top story

Tesla's new Roadster flies on SpaceX thrusters

Musk's long-delayed sports car finally launched off a ramp and landed itself using SpaceX-built thrusters, feeding talk of a bigger deal between his two companies.

Signal feed

9 moves
Elon MuskProduct launchAutosTSLAAI read · — Flashy, not profitable yet

Tesla's new Roadster flies on SpaceX thrusters

The move

Tesla showed its second-generation Roadster on Oct. 15 at a SpaceX test site in McGregor, Texas. In a livestreamed demo, cold-gas thrusters built with SpaceX engineers launched the car off a ramp, flipped it in the air and landed it on its wheels; a second run had it jump over a row of vehicles. Musk said production will begin in 2028, eight years after the car's original 2020 target. Tesla gave no price, taking $50,000 deposits for a Performance version and $250,000 for a Thrust version.

Why it matters

The SpaceX-built thrusters and the timing fed analyst speculation that Musk could be building toward merging Tesla and SpaceX.

The bear case

The car won't ship for two more years, and one analyst estimated volumes of only about 2,000 a year, tiny next to Tesla's regular sales.

What to watch

Whether Tesla's board raises a Tesla-SpaceX merger at its annual meeting, expected by Dec. 6.

Source · Deccan HeraldRead original ↗Report an error
Mark ZuckerbergProduct launchAI & techMETAAI read · — Rare truce with Apple

Meta and Apple team up to run AI model on Macs

The move

Meta released its Muse Spark Open AI model on Oct. 15 and said it worked with Apple to let the model run locally on two linked Mac Studio computers using Apple's M5 Ultra chips and its MLX software. The model has 816 billion total parameters, 42 billion active at a time, and a 1-million-token context window, aimed at coding and multi-step agent tasks. Apple CEO John Ternus said it's 'amazing to run a model of Muse Spark Open's caliber and scale this efficiently, right from your desk.'

Why it matters

It's an unusual collaboration between two companies with a history of friction over privacy rules, and shows Meta pushing its open model onto more hardware.

The bear case

The 512GB Mac Studio needed to run it hadn't shipped yet, and the setup needs multiple machines that already cost over $10,000 each, not an ordinary single Mac.

What to watch

Whether Apple ships the 512GB Mac Studio this month as planned.

Warren BuffettInvestmentFinanceBRK.A · BRK.BAI read · ▲ Buffett doubles down

Berkshire adds $330 million to Bending Spoons stake

The move

Berkshire Hathaway agreed on Oct. 14 to buy 10 million new Bending Spoons shares at $33 each, a $330 million private placement that raises its stake in the Italian software company to about 4%. It follows Berkshire's $600 million investment at Bending Spoons' July 2026 IPO. Bending Spoons said the new capital would support further acquisitions.

Why it matters

It shows Buffett still betting on Bending Spoons' acquisition-driven model just three months after its IPO.

The bear case

Berkshire is concentrating more money in a young, newly public company without the long track record Buffett typically demands before investing.

What to watch

Any new acquisition Bending Spoons announces with the fresh capital.

Masayoshi SonLegalAI & techARM · QCOMAI read · — Pyrrhic win for Qualcomm

Jury finds Arm breached Qualcomm license, awards $2

The move

A Delaware federal jury ruled on Oct. 13 that Arm Holdings, majority-owned by Masayoshi Son's SoftBank, breached its architecture license agreement with Qualcomm by withholding software patches and tools, and that Arm intentionally interfered with Qualcomm's relationship with Meta. The jury awarded Qualcomm only $2 in nominal damages and ruled for Arm on five other claims. A separate, potentially larger royalty dispute between the two companies now goes to the judge.

Why it matters

It's a narrow, low-dollar win for Qualcomm in a long-running fight that could still affect how much Arm collects in chip-design royalties.

The bear case

The damages were trivial and the jury rejected most of Qualcomm's other claims, so the practical impact on Arm's business may be limited until the royalty ruling.

What to watch

Judge Maryellen Noreika's ruling on the remaining royalty dispute.

Jamie DimonPolicyFinanceJPMAI read · ▼ Drawing a hard line

JPMorgan blocks outside AI agents from accounts

The move

On JPMorgan's Oct. 13 third-quarter earnings call, CEO Jamie Dimon said the bank won't let outside AI agents, such as Meta's Muse, log into or act on customer accounts. 'We're not going to allow just any outside agent to come inside the company,' he said, adding 'you don't want rogue agents.' A bank spokesperson said JPMorgan blocks third-party agents that try to access accounts using customers' banking credentials. The quarter's results also missed estimates, and Dimon said the bank cut back share buybacks.

Why it matters

Banks are racing to let AI agents handle payments, and Dimon is drawing a line at the risk that poses to clients and the bank itself.

The bear case

The policy could leave JPMorgan behind if rivals let AI agents manage money more freely and customers migrate there.

What to watch

Whether JPMorgan sets a timeline for its own customer-facing AI agent tools.

Mukesh AmbaniIPO / MarketsOtherPrivateAI read · ▲ Big push for demand

Reliance pays staff early ahead of record Jio IPO

The move

Ahead of Jio Platforms' IPO, Reliance paid October salaries to its nearly 420,000 group employees about 15 days early and encouraged them to invest, without saying why, according to a Bloomberg report. The IPO's anchor book opens Oct. 19 and public bidding runs Oct. 21-23, targeting up to 30,200 crore rupees (about $3.2 billion) at a price band of 1,065 to 1,119 rupees a share, valuing Jio at up to 10.3 lakh crore rupees (about $108 billion). Up to 500 crore rupees of shares are set aside for eligible Jio employees at a 5% discount.

Why it matters

It's one of India's largest-ever IPOs, and the early pay and outreach show how hard Reliance is pushing to build demand beyond big institutional investors.

The bear case

Encouraging employees to put personal savings into a company-linked stock concentrates their financial risk in one bet.

What to watch

Jio Platforms' public subscription window, Oct. 21-23.

Source · Business TodayRead original ↗Report an error
Michael DellPolicyAI & techDELLAI read · — Fire needs rules

Michael Dell calls for AI 'guardrails' and oversight

The move

In an Oct. 13 CNBC interview, Dell Technologies CEO Michael Dell said AI models need 'safeguards and controls,' and that companies should limit AI systems to 'bounded permissions' so they can only do what they're explicitly authorized to do. He said government has a role in overseeing AI but declined to propose specific policies, saying that's outside his role as a business leader.

Why it matters

A major supplier of the hardware behind the AI buildout is publicly calling for guardrails even as his company profits from that spending.

The bear case

Calling for government to play 'a role' without specifics leaves the actual policy fight unresolved.

What to watch

Any specific AI safety framework Dell endorses.

Source · CryptoBriefingRead original ↗Report an error
Hock TanPolicyAI & techAVGOAI read · — Money, not chips

Broadcom's Tan says capital, not chips, limits AI

The move

Speaking on a panel in Singapore on Oct. 12, Broadcom CEO Hock Tan said capital, not chip or memory shortages, will be the biggest constraint on AI growth, even as Broadcom works through earlier wafer and memory shortfalls. He said demand for AI computing still outpaces supply and could stay that way through 2029 or into 2030, and compared many AI investors to casino gamblers who don't understand what they're buying.

Why it matters

One of the biggest AI chip suppliers is warning that funding, not hardware, could be what slows the AI buildout.

The bear case

Tan has a financial interest in talking up continued AI chip demand even while flagging funding risk elsewhere in the industry.

What to watch

Whether AI financing costs rise enough to slow data-center spending.

Jensen HuangPolicyAI & techNVDAAI read · — Talk, don't restrict

Nvidia's Huang calls AI 'potentially dangerous,' urges talks

The move

At a media briefing in London on Oct. 14, Nvidia CEO Jensen Huang called AI 'potentially dangerous' but said the U.S., U.K. and China should keep talking rather than add new rules, saying 'dialogue, dialogue.' He said he keeps in regular contact with AI researchers in China and wants the U.S. and U.K. to do the same, while telling investors to model Nvidia's China revenue as zero under current export rules.

Why it matters

Huang is pushing cooperation over restriction even as U.S. export curbs already keep Nvidia largely out of China's market, a stance that puts him at odds with rivals pushing for tougher AI rules.

The bear case

Favoring dialogue over regulation also serves Nvidia's commercial interest in keeping access to Chinese research ties open.

What to watch

Whether the U.S. eases any chip-export restrictions on China.

Source · CryptoBriefingRead original ↗Report an error

← All editions