OpenAI is on a new roll: Its latest image generator is dazzling millions of users, it's teasing new models and hardware, and it just landed another $40 billion to pay for all that and more.

Why it matters: The renewed mojo comes after a bumpy few months that saw handwringing over the release of advanced models from China's DeepSeek in January and disappointment at delays in the next big upgrade of the GPT models that drive the company's success.

Driving the news: OpenAI announced yesterday it has completed a $40 billion funding round — the largest venture capital round ever, led by Japan's SoftBank — giving the company a whopping $300 billion valuation.

• The funding news capped a flurry of other upbeat messages from the company.

A million new customers signed up for OpenAI in one hour, Altman said in an X post — presumably, so they can use ChatGPT's popular and powerful new image-making tool, which the company has now rolled out to ChatGPT's free users.

• ChatGPT's images can now more reliably include legible text and imitate a wide range of styles, including those of well-known artists and brands — like that of Japanese filmmaker Hayao Miyazaki, whose Studio Ghibli cartoons are inspiring a flood of AI-driven photo remixes.

What they're saying: "The ChatGPT launch 26 months ago was one of the craziest viral moments I'd ever seen, and we added one million users in five days," Altman said on X — making the million-in-an-hour number seem even crazier.

• David Sacks, the venture capitalist advising President Trump on AI policy and crypto, replied to Altman's post with a link to a blog post that reads: "The next big thing will start out looking like a toy."
• Altman responded, "yeah, i just didn't think it would be this toy :)," apparently referring to the image generator.

Altman also announced yesterday that OpenAI, for the first time since GPT-2 in 2019, plans to publicly release both a model and the weights behind it — the key numbers that determine its output and enable anyone to run it. (That's what most people mean when they talk about an "open source" model and is similar to what Meta has done with its Llama models.)

Between the lines: Over the weekend, Altman also offered a tantalizing tease of the company's long-brewing hardware plans.

• Responding to an X engineer who said that computers should be cute, Altman wrote, "We are gonna make a rly cute one," without offering further details.
• The company has recently expanded its hardware team, hiring former Meta mixed reality executive Caitlin Kalinowski to lead its robotics and consumer hardware efforts.

Yes, but: The concerns that prompted OpenAI's winter of discontent haven't evaporated.

• The company has yet to deliver the next major release of GPT. A more modest update, GPT-4.5, was released in February. OpenAI said that it would likely be the last one without the "chain-of-thought" reasoning that OpenAI includes in its current "o" series of models.
• OpenAI scrapped a full release of its o3 model, saying earlier this year it would instead be integrated with GPT-5. But the company hasn't named a release date.

The intrigue: Still hanging over OpenAI is the fate of its plan to effectively convert itself from a nonprofit to a for-profit entity.

• This latest investment round, like other recent funding, is partly contingent on OpenAI successfully completing that shift. Elon Musk has sued to block the change, and Meta has called on California's attorney general to investigate it.
• If OpenAI doesn't complete the conversion by the end of 2025, SoftBank could reduce its investment to $20 billion from $30 billion.
• During the company's last funding round — a $6.6 billion round announced in October — investors got the right to demand their money back if the corporate structure didn't change within two years.

Apple's latest iOS 18.4 update, released yesterday, brings a few new Apple Intelligence features, an updated photo app, and a fresh batch of emoji.

Why it matters: While the update doesn't bring any huge advances, it improves some current AI tools and makes others more broadly available, thanks to new language support.

The big picture: Priority notifications will appear at the top of the home screen, using Apple Intelligence to highlight "important notifications that may require your immediate attention," according to the release notes.

• Apple Intelligence now supports more languages, including French, Japanese, and Brazilian Portuguese, according to an Apple press release.
• A new "Sketch" style for images is available in Image Playground. It allows users to create AI-generated images inspired by hand-drawn sketches.

Between the lines: The new update also brings Apple Intelligence to the Vision Pro as well as a new Vision Pro app to help headset owners find new content without donning the device.

• Amazon announced Nova Act, an AI agent capable of performing simple computer actions, like controlling a web browser. (TechCrunch)
• Amazon is also starting to roll out Alexa+, though not with all the features it showed in a February demo. (The Verge)
• Tech firms are telling employees who are in the U.S. on visas not to leave the country, for fear that new Trump administration policies will bar them from returning. (Washington Post)
• ChatGPT's new image generator can create deepfake receipts, complete with folds in the paper and food stains. (TechCrunch)
• The 28-year-old founder of Brainbase, a company promising to create "AI workers," is now a DOGE staffer recently appointed to lead the U.S. Institute of Peace. (Rolling Stone)

While the notion of "bugs" in computers has been around longer, the term was popularized after a 1947 incident in which Grace Hopper and others at IBM found a literal bug — a moth — messing up one of their Mark II computers.