Tesla competitor Slate Auto said it closed a $650 million Series C funding round led by TWG Global, giving it the “operating capital to reach the next stage of development.” Slate’s new CEO, Peter Faricy, says it has more than 160,000 reservations, up from 150,000 in December, and is “on time and on budget” to deliver its first mid-$20,000 electric trucks to customers by the end of 2026.
Microsoft is feeling the heat from all corners of the tech world as it tries to infuse its productivity apps with useful AI tools.
OpenAI, Anthropic, and now open-source OpenClaw are enabling powerful agentic AI that can do work on your computer for you — including productivity functions like managing emails, spreadsheets, and slide decks.
This is obviously an area where Microsoft needs to compete, or it will be left in the dust by AI startups.
The Information reports that Microsoft is indeed realizing this, and is now trying to reboot its many Copilot tools to act more like the extremely popular DIY agentic AI tool OpenClaw.
OpenClaw is usually set up running on a dedicated personal computer, and given access to all of a user’s permissions and logins. The user issues orders to OpenClaw through messaging apps like Telegram or WhatsApp, and the agent goes off and completes tasks in the background, notifying you when they’re done. But many users have had security disasters with the setup, so Microsoft is looking to borrow the popular concept but implement the strict security controls needed for use in enterprise environments.
According to the report, Microsoft CEO Satya Nadella has made revamping 365 Copilot a top priority.
This is obviously an area where Microsoft needs to compete, or it will be left in the dust by AI startups.
The Information reports that Microsoft is indeed realizing this, and is now trying to reboot its many Copilot tools to act more like the extremely popular DIY agentic AI tool OpenClaw.
OpenClaw is usually set up running on a dedicated personal computer, and given access to all of a user’s permissions and logins. The user issues orders to OpenClaw through messaging apps like Telegram or WhatsApp, and the agent goes off and completes tasks in the background, notifying you when they’re done. But many users have had security disasters with the setup, so Microsoft is looking to borrow the popular concept but implement the strict security controls needed for use in enterprise environments.
According to the report, Microsoft CEO Satya Nadella has made revamping 365 Copilot a top priority.
All eyes on are SpaceX as it prepares for a blockbuster IPO as soon as this summer, and everyone is eager to get a look at the company’s official numbers for the first time.
The Information is reporting that last year, SpaceX posted $18.5 billion in revenue with a $5 billion loss.
According to the report, the numbers reflect the combined finances of SpaceX and xAI, which it acquired in February.
After acquiring xAI, SpaceX’s successful space launch and satellite business may have been dragged down by xAI’s massive data center spending. Earlier this year, Bloomberg reported that xAI had burned through $8 billion in the first nine months of 2025.
According to the report, the numbers reflect the combined finances of SpaceX and xAI, which it acquired in February.
After acquiring xAI, SpaceX’s successful space launch and satellite business may have been dragged down by xAI’s massive data center spending. Earlier this year, Bloomberg reported that xAI had burned through $8 billion in the first nine months of 2025.
Amazon is looking for a magic trick that can help it get past data center construction bottlenecks so it can work through the $244 billion worth of cloud computing backlogs it wants to deliver.
It may have just pulled a rabbit out of its hat. (I know, groan.)
Business Insider is reporting that Amazon’s Project Houdini seeks to slash labor costs and installation time by building modular “data halls” — the rows of racks of servers that make up the heart of data centers — in factories, and then shipping them fully assembled on trailers to data center sites.
According to the report, the modular plan would save weeks of construction time and tens of thousands of hours of labor costs.
This week in Amazon’s letter to shareholders, CEO Andy Jassy wrote that the company is planning $200 billion in capital expenditure this year, and that it is embracing its tradition of taking big bets on experiments like Project Houdini:
“You need to invent and experiment like crazy. Many of these experiments will fail, and it might feel like you’re getting nowhere. But, your culture must possess the tenacity to keep at it.”
Business Insider is reporting that Amazon’s Project Houdini seeks to slash labor costs and installation time by building modular “data halls” — the rows of racks of servers that make up the heart of data centers — in factories, and then shipping them fully assembled on trailers to data center sites.
According to the report, the modular plan would save weeks of construction time and tens of thousands of hours of labor costs.
This week in Amazon’s letter to shareholders, CEO Andy Jassy wrote that the company is planning $200 billion in capital expenditure this year, and that it is embracing its tradition of taking big bets on experiments like Project Houdini:
“You need to invent and experiment like crazy. Many of these experiments will fail, and it might feel like you’re getting nowhere. But, your culture must possess the tenacity to keep at it.”
AI benchmark leaderboards are often where mysterious new models make their debut, stoking speculation about the unnamed companies behind them.
That was the case with an impressive new text-to-video model named HappyHorse-1.0 that shot to the top of public leaderboards. CNBC reports that Chinese tech giant Alibaba has confirmed that it is the owner of the new model.
HappyHorse beat out the popular Seedance model from rival ByteDance in blind human evaluations to claim the top spot on the Artificial Analysis text-to-video leaderboard.
While OpenAI has announced it is shuttering its text-to-video Sora app, the category continues to see intense competition as a flurry of video models improve with more realistic physics and cinematic effects.
HappyHorse beat out the popular Seedance model from rival ByteDance in blind human evaluations to claim the top spot on the Artificial Analysis text-to-video leaderboard.
While OpenAI has announced it is shuttering its text-to-video Sora app, the category continues to see intense competition as a flurry of video models improve with more realistic physics and cinematic effects.
