Boeing’s turnaround plan may be experiencing some unexpected turbulence, but it’s still at cruising altitude.

Mere days after the US Federal Aviation Administration delayed the certification of Boeing’s 737 Max 10 due to a software issue, the aerospace giant secured a big green light from another corner of the US federal government: the Navy, which awarded the company a $20 billion contract to build a next-generation fighter jet. It’s a sign that even in times of trouble, Boeing can and will still find plenty of buyers.

Flight Delay

On Saturday, The Wall Street Journal reported that the FAA was investigating a software glitch on the MAX 10 that prevented some pilots from accessing automated flight guidance tools. Investors balked at the latest delay in what had already been a yearslong and troubled certification process for the latest version of the single-aisle jetliner; shares of the company have fallen about 6% so far this week.

The good news for Boeing? FAA Administrator Bryan Bedford has said the glitch isn’t nearly as troubling as the issues in back-to-back crashes of earlier, smaller MAX jets in 2018 and 2019; pilots, in this case, still maintain full control of the aircraft. The bad news? Bedford has said the delay will continue until the agency is satisfied that the issue is resolved, and Boeing may not be able to roll out a permanent software fix until 2028. (US Transportation Secretary Sean Duffy said Wednesday that he sees no reason for a prolonged delay.)

In the meantime, the company’s order book continues to fill up, across both its commercial and defense units:

• The Navy contract is for a jet plane currently designated the F/A-XX Strike Fighter, which could be worth hundreds of billions of dollars over its lifetime, analysts told Reuters. It’s Boeing’s second big contract to build fighter jets in two years, after scoring a $20 billion contract to produce F-47 jets for the US Air Force in March 2025.
• The company’s entire order backlog stood at a staggering $715 billion at the end of June, roughly $600 billion of which is for commercial craft. It has 1,500 orders for the 737 MAX 10 already.

To the Moon: Boeing will face at least one more big test before the end of the year. On Monday, NASA said it will launch the latest test flight of the company’s Starliner space capsule to the International Space Station as soon as December. The Starliner’s previous test flight in 2024 left two US astronauts stranded on the International Space Station for nine months; the trip had been scheduled to last only a week. This time, the mission will be unmanned.

One of the biggest tech themes of this year is vibe coding. People build applications by telling artificial intelligence tools in natural language to write code for them, allowing non-technical developers to prototype and build projects in the time it normally takes us to realize we’re too dumb for the job. Startups pioneering the practice have raised billions. SpaceX paid $60 billion for one of them, Cursor, in August.

Could those same vibes soon be applied to investing? Brokerage Robinhood announced a suite of new offerings this week and, on Wednesday, CEO Vlad Tenev explained how AI-powered trading tools could let users develop their own sophisticated, around-the-clock trading strategies that once relied on Wall Street’s brainiest software engineers.

Vibe Investing

Two of the new services demonstrate how competition among online financial services firms is increasingly entangled across traditional and emerging sectors. First, Robinhood will allow customers to trade select stocks and ETFs 24/7 starting next year. That marks a broadening of the limited trading Robinhood already allows for 24 hours on five days of the week and puts it in the thick of a race for sleepless customers that includes Charles Schwab, Nasdaq and the London Stock Exchange.

Second, Robinhood will allow users to trade crypto perpetual futures contracts, which enable speculation on Bitcoin, Ethereum and other digital assets without holding the underlying assets.

And then there’s the big news: the AI-powered automated trading tools that let users build their own agents to automate trading and implement around-the-clock trading strategies. “A lot of people say, well, you’ll just be able to tell your Robinhood agent, go out and make me money,” CEO Vlad Tenev explained in a Wednesday interview with Yahoo Finance. “But, if everyone does that, that’s unlikely to be the best strategy.” Instead, he said a wholesale, vibe coding-like shift may be at hand:

• “It’ll create a new market, a new dynamic where traders can compete with one another over who can create the best strategy,” Tenev said. “And that strategy will be in English rather than in Python or R or C++, which is what it’s traditionally been in.”
• “AI democratized software engineering and made it so that anyone, even people without a formal computer science degree, can write extremely complicated code and anyone can make a website and deploy it,” he added.

Cold Reception: Analysts were encouraged. Morgan Stanley and Deutsche Bank reiterated buy ratings on Robinhood stock, as well as their respective $150 and $134 price targets (compared with a $112 closing price on Wednesday). But, investors’ initial reaction was less enthusiastic: Robinhood shares fell 3.20% on Wednesday.

In the AI world, claiming to improve upon Nvidia’s CUDA programming interface is like saying you can beat Usain Bolt in the 100-meter. DeepSeek seems to be feeling fast, saying yesterday its new software TileLang is simpler to program than CUDA while still utilizing chips to their upper limits.

DeepSeek said the new software, which it built alongside Huawei for use on the tech giant’s Ascend chips, is now available to the public and free to download. As Chinese companies try to reduce their reliance on Nvidia, Huawei’s advancing its hardware while DeepSeek is tackling software. Huawei said this month it’s planning to release its next-gen Ascend chips early next year, and DeepSeek’s TileLang will be used to program them.

Chips alone can’t compete with Nvidia, and companies have struggled to create CUDA alternatives.

Woulda, CUDA, Shoulda

CUDA (Compute Unified Device Architecture) was a major breakthrough in parallel computing when Nvidia unleashed it in 2006, and it has become the AI industry’s go-to for building neural networks on GPU chips. Without it, Nvidia chips might still be known more for running World of Warcraft than ChatGPT. In some ways, CUDA is to Nvidia what iOS is to Apple. The software helps lock customers into using the hardware, because CUDA plays better with Nvidia’s chips than those of competitors.

DeepSeek and Huawei are far from the first tech companies to attempt a CUDA copycat:

• CUDA alternatives have been around for a decade and change. AMD’s ROCm has failed to gain CUDA’s traction among developers, as has Intel’s oneAPI. OpenCL, backed by companies including Apple, AMD and Qualcomm, also failed to break through.
• Going up against CUDA today is even more challenging, after some 20 years of building and optimizing its developer libraries and entrenching itself as the common language of AI.

Limited Impact: CEO Jensen Huang said in May that Nvidia had largely conceded China’s AI-chip market to Huawei, despite huge demand from the country. Though some less advanced Nvidia chips can be sold to Chinese companies, China has been pushing for self-reliance amid US bans. TileLang could help China further separate itself from the US and Nvidia specifically. But it doesn’t seem likely to turn many heads among the millions of developers that depend on CUDA.