|
|
|
|
|
|
|
|
|
|
The Morning Download: Virtual Software Developers Are Here
|
|
By Steven Rosenbush | WSJ Leadership Institute
|
|
|
|
|
|
|
Good morning. It’s one thing to highly automate tasks or parts of a job, another to replace a human worker with an AI worker. One Silicon Valley startup, JustPaid, has stood up a nearly fully autonomous software engineering team.
Vinay Pinnaka, co-founder and chief technology officer of the Mountain View, Calif.-based startup, used a combination of OpenClaw and Anthropic’s AI coding tool, Claude Code, to create a team of seven AI agents to grind out code 24/7.
Be sure to read the story by the WSJ Leadership Institute’s Belle Lin, which looks at the frontier of human-AI collaboration.
|
|
|
|
|
Content from our sponsor: Deloitte
|
|
|
Deutsche Bank Tech Exec: ‘Diligence and Determination’ Pave a Path to Innovation
|
|
Putting engineers at the heart of modernization efforts can help drive innovation in financial services, according to Deutsche Bank’s Margaret Dolson. Read More
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
JustPaid co-founders from left: Anelya Grant, chief product officer; CEO Daniel Kivatinos; and Chief Technology Officer Vinay Pinnaka. JustPaid
|
|
|
|
|
|
The company, which makes an AI-powered platform that automates financial operations, was founded in 2023 and has nine employees.
In the month since Pinnaka put his AI team to work, it has turbocharged workflow. The agents have built 10 major features, each of which would have taken Pinnaka’s human developers a month or more to build. Recently, the JustPaid team hired a new human developer who was trained almost entirely by the AI agent engineers.
Each of his seven AI agents has its own identity and tasks, from writing code to reviewing it and conducting quality assurance.
|
|
|
|
|
|
|
|
|
Many developers also say they’re certain that personal tasks—as well as their jobs—will one day become fully automated with OpenClaw, or one of its many alternatives.
When that shift happens, Pinnaka says his own human engineers will still have jobs, but they’ll no longer be hands-on with code. “As we automate every single task, there should be a manager,” he said.
|
|
|
|
|
|
|
|
There are drawbacks to relying heavily on bots to do software development work. For one, it can get expensive to use tokens—the basic units of text that AI processes. Developers say token usage adds up extremely quickly, especially when using AI agents to run multiple tasks in the background.
When Pinnaka first began experimenting with Claude Code and OpenClaw, he racked up a bill of $4,000 each week. Then, with some adjustments like using a smaller, more efficient Claude model, he decreased his budget to $10,000 to $15,000 each month.
Pinnaka isn’t alone. Kuse, a startup that develops “AI co-workers” on behalf of companies, built its own AI agent employees using OpenClaw. The company said its AI agent employees collaborate with each other and alongside human employees. The agents also have their own identity—they send Slack and Gmail messages, speak in Zoom meetings and proactively start work. “We didn’t remove anyone or stop hiring, but we have fewer reporting lines and fewer meetings,” the company said.
For large businesses, it’s still considered too risky to deploy AI agent platforms like OpenClaw. That’s why companies like Nvidia have released their own tools that aim to make them safer for enterprises. For OpenClaw to work as a true personal assistant, it must have access to all of a user’s data and systems. When agents go rogue, they can tamper with or delete valuable files.
But even enterprises are catching on to the idea that, soon, AI agent software developers are going to be taking over much of what human engineers are doing, according to Arun Chandrasekaran, an AI-focused analyst at market research and consulting firm Gartner.
Large teams of business developers focused on different parts of the software development process could also become much smaller, he said, especially as AI-based coding tools improve.
“They’re starting to wonder, ‘How does this fundamentally change the way we do software development?’” Chandrasekaran said.
Are you a bot? Let us know.
|
|
|
|
|
CoreWeave raises $8.5 billion in chip-backed debt deal. CoreWeave raised $8.5 billion in what it calls the largest chip-backed debt deal of its kind, Bloomberg reports. The investment-grade-rated loan is secured by GPUs and a Meta customer contract worth at least $19 billion. Institutions led by Mitsubishi UFJ Financial Group and Morgan Stanley provided the new loan.
|
|
|
|
Iran War Rattles AI Supply Chain
|
|
|
|
|
|
VAT Group is a key supplier to the semiconductor industry, which relies on extremely clean environments for production. Jens Schlueter/Agence France-Presse/Getty Images
|
|
|
|
|
|
VAT Group warns of weaker sales on Iran war disruption. The Swiss maker of vacuum valves critical to contamination-free chip manufacturing said semiconductor demand remained strong in the first quarter, with order growth intact, but Middle East supply-chain disruptions delayed key components and materials, cutting into sales.
|
|
|
Iran war chokes off helium supply critical for AI. The global supply of helium is being squeezed by a halt in natural-gas exports from Qatar, the source of about a third of the world’s total, the WSJ reports. Helium is critical for cooling chip-making equipment, MRI scanners, rockets and fiber-optic manufacturing.
The U.S., the world’s largest producer of the gas, is more insulated for now, though analysts say a prolonged Qatari outage would hit it hard, too.
|
|
|
|
|
Why OpenAI shut down Sora. OpenAI shocked many last week with its decision to shutter its video generation app Sora. WSJ reporter Berber Jin joins us for an exclusive look behind the scenes of the decision.
|
|
|
|
|
|
|
|
States defy White House on AI regulation. California Gov. Gavin Newsom issued on Monday an executive order requiring safety and privacy guardrails for AI contractors. The New York Times reports that the order was the latest in a skirmish pitting President Trump and his tech allies looking to block state laws on AI. Over 100 state laws are already on the books, and states keep passing more. More below.
|
|
|
Half of U.S. believes AI likely to cause harm. Americans aren't buying the AI hype. Bloomberg reports on a new Quinnipiac poll that finds that 55% of Americans believe AI will do more harm than good in their daily lives, up 11 points since last April. Seventy percent expect AI to reduce job opportunities, and nearly two-thirds think it will worsen education. Sixty-five percent oppose AI data centers in their communities.
|
|
|
|
|
|
|
Microsoft said it plans to invest more than $1 billion in Thailand over the next two years in cloud and artificial-intelligence infrastructure.
|
|
|
|
|
Amazon.com struck a deal to provide internet access on Delta Air Lines, the tech giant’s highest-profile victory in the race to supply in-flight Wi-Fi, the WSJ reports.
|
|
|
|
Everything Else You Need to Know
|
|
|
President Trump told aides he’s willing to end the U.S. military campaign against Iran even if the Strait of Hormuz remains largely closed, administration officials said, likely extending Tehran’s firm grip on the waterway and leaving a complex operation to reopen it for a later date. (WSJ)
U.S. stocks are set to deliver their worst quarter in nearly four years. The tech-heavy Nasdaq composite lurched into correction territory on March 26, meaning it had fallen 10% below its recent high. A day later, the Dow Jones Industrial Average (a benchmark for the real economy) joined it. (WSJ)
Unilever is in advanced talks to combine its food business with spice maker McCormick in a deal that would create a new food behemoth worth roughly $60 billion, including debt. (WSJ)
Washington State on Monday enacted a tax on income over $1 million, the latest Democratic state to levy a tax aimed squarely at the wealthy. The 9.9% tax will apply only to income over $1 million. Until now, Washington hasn’t had any state income tax. (WSJ)
|
|
|
|
The WSJ Technology Council
|
|
|
|