Welcome back to The Prompt,

What should a world with artificial intelligence that surpasses human cognitive abilities look like — and how should the government respond? OpenAI has some ideas, starting with higher taxes on corporate profits and capital gains. In a policy paper released this week, the company proposed that this could help offset the impact of businesses replacing people with AI tools, as payroll-based tax income would shrink if people lose their jobs. The ChatGPT maker also suggested creating a public wealth fund that would allow every citizen to benefit from AI-driven growth, and experimenting with four-day work weeks.

The policy proposals come as questions surface about whether OpenAI’s chief executive Sam Altman is even the right person for the job. There’s no doubt that he’s one of the most powerful leaders of our time, shepherding one of the most consequential technologies ever built. But the 40-year-old billionaire has a reputation for being untrustworthy and deceitful, according to a lengthy New Yorker deep dive story sourced from interviews with 100 former and current OpenAI employees and people in his circle as well as documents and memos. One example that stuck out: tapping wealthy Middle Eastern countries like Saudi Arabia and the United Arab Emirates for billions of dollars to construct huge data centers— a plan that board members reportedly knew nothing about. While the story contained some new details, it largely rehashed what others have reported since Altman’s 2023 ouster and reinstatement.

But Altman is continuing to clash with OpenAI insiders, according to The Information. He has privately said that he wants OpenAI to go public as soon as the fourth quarter of the year, even though the company expects to burn $200 billion before generating profits, per the report. But CFO Sarah Friar has internally voiced concerns about the risk inherent in OpenAI’s spending plans and said that the company isn’t ready for an IPO this year, according to the report. Friar was also excluded from some conversations about OpenAI’s financial plans. In a statement OpenAI said the two executives are “fully aligned” and have both been directly involved in important financial meetings.

Now let’s get into the headlines.

On Tuesday, Anthropic announced a cybersecurity initiative called Project Glasswing and a new AI model called Claude Mythos Preview that can identify security weaknesses and flaws in software. It’s so good at finding vulnerabilities that the AI company is limiting the models’ access to 40 companies including Apple, Google and Microsoft to prevent hackers from misusing it. “The dangers of getting this wrong are obvious, but if we get it right, there is a real opportunity to create a fundamentally more secure internet and world than we had before the advent of AI-powered cyber capabilities,” CEO Dario Amodei wrote on X.

The company also announced that its revenue run rate has increased to $30 billion up from $9 billion at the end of 2025. Anthropic said it plans to work with Broadcom and Google to access computing infrastructure as it scales up its powerful LLM Claude.

Medvi, a telehealth startup that sells GLP-1 weight loss drugs, recorded $401 million in sales last year with just two employees: CEO Matthew Gallaghar and his brother, the New York Times reported. The Los Angeles-based profitable startup projects $1.8 billion in sales this year. That was made possible by a medley of AI tools that handle different parts of the business: writing code, churning out website copy, creating ads and talking to customers. But in February, the FDA issued the company a warning letter for making misleading claims about compounded drug products. There are other red flags: the company has used fake AI-generated doctors to market its drugs Business Insider reported. It also posted before-and-after weight loss images stolen from news articles and altered with AI, according to Futurism.

HED: Generalist Is Betting Its Robot-Training Gloves Will Usher In Robotics’ ChatGPT Moment

The robot, a pair of disembodied arms with crablike pincers at the end, wasn’t supposed to pick up the bag. It had been told to do a single tedious job: open plastic bags on a conveyor belt and stuff toy potted plant plushies inside them.

Then one plushie snagged halfway in. The robot paused — briefly, as if assessing its work — then did something it had not been programmed to do. It raised its other arm, grabbed the other side of the bag, gave it a quick shake so the toy slid all the way down, and then placed it back on the belt.

For a human worker that’s muscle memory. For the engineers at Generalist, a Silicon Valley startup developing robot “brains”, it was a tell: the robot wasn’t just replaying a scripted task. It was improvising.

Those kinds of “emergent” behaviors are the reason Generalist’s CEO, Pete Florence, who was a lead on PaLM-E, one of Google’s foundational robotics papers, thinks robotics is nearing its ‘ChatGPT moment.’ The startup, which raised $140 million at a $440 million valuation in 2025, and which Florence founded with Google co-worker Andy Zeng and Boston Dynamics roboticist Andy Barry, has largely been under-the-radar. Now it’s releasing a new model called Gen-1, and Florence says it can help off-the-shelf robots handle a wider range of high-dexterity tasks usually performed by humans — folding laundry and “kitting,” packing multiple different types of items into a single box—while improvising in the messy, unpredictable edge cases that have historically stumped robots.

“What’s happening now with robotics parallels when people opened GPT-3 and asked it to write a completely new limerick,” he told Forbes. “The limerick didn’t exist before. To achieve that, you need an improvisational level of intelligence. What we’re doing applies to robotics and beyond.”

After years of playing second fiddle to software, robots are back in fashion in Silicon Valley. Nvidia CEO Jensen Huang helped ignite the latest frenzy last year when he declared robots were entering the ChatGPT era. Since then, the internet has been flooded with videos of humanoid robots performing backflips, breakdancing and vaulting. Meanwhile, most real-world robots still struggle outside carefully defined tasks. ChatGPT may write code and boilerplate emails, but robots still don’t make lunch, handle DoorDash deliveries or run factories without an army of human babysitters.

Read the full story on Forbes.

Companies like Databricks, Nvidia and Meta are publicly celebrating engineers that are spending thousands of dollars on AI “tokens” —an atomic unit of data processing for an AI model to generate text, Forbes reported. That’s one way to incentivize employees to embrace AI tools, encouraging them to max out AI budgets and supercharge their productivity. One customer service startup Sendbird has a running leader board that ranks engineers based on how much they spend on tokens— from beginner to “AI god.” “If that $500,000 engineer did not consume at least $250,000 worth of tokens, I am going to be deeply alarmed,” Nvidia CEO Jensen Huang said on a recent podcast.

© Forbes