You've probably noticed that every AI tool you use still needs you sitting there, typing prompts, reviewing outputs. Google just changed that. Yesterday at I/O, they launched an AI agent that runs in the cloud 24/7, works through your email and docs while you're asleep, and doesn't need your laptop to be open. Meanwhile, 8,000 Meta employees found out this morning whether they still have a job, and the Musk-OpenAI trial ended in under two hours.

Let's get into it.

Meta's 8,000 layoffs hit today. Notifications started going out at 4 a.m. PT. North American staff were told to work from home. Zuckerberg held a company town hall where he said "AI is not driving layoffs" while raising AI infrastructure spending to $145 billion. The company posted $56 billion in quarterly revenue. More cuts planned for H2 2026.

Musk lost the OpenAI trial in under two hours. A nine-member jury unanimously dismissed all of Musk's claims against Sam Altman and OpenAI on statute-of-limitations grounds. The court never reached the substance of whether OpenAI violated its nonprofit mission. Musk called it a "calendar technicality" and plans to appeal to the 9th Circuit.

NPR is restructuring for a world without federal funding. The network is offering buyouts to 300 employees, facing an $8 million budget gap from lost government subsidies and declining sponsorship. If not enough people volunteer by May 26, targeted layoffs follow. The network received $113 million in private gifts recently, but most of that money is earmarked for technology, not operations.

AI agents are in production at 72% of enterprises, but nobody's watching them. A new report found a 60% governance gap between deployment and oversight. Separately, 67% of executives believe their company has already had a data leak from unapproved AI tools. Production is outrunning policy everywhere.

2026 layoff tracker: 138,988 workers across 327 events. That's 993 people per day, and we're not even halfway through the year. The pace is approaching last year's full-year total. AI remains the most-cited driver, particularly in software, fintech, and content roles.

Hiring

Funding

Contracts

Mark Spitznagel, who made $1B in a single day during the 2015 flash crash, warned markets are mimicking 1929. Seems extreme but we did just see the worst quarter for the S&P since 2022.

So it’s not so surprising that Vanguard and Goldman Sachs forecasted 5% and 3% annual S&P returns respectively for 2024-2034.

Late last year, Apollo’s chief economist Torsten Slok put it this way: "expect zero in return in the S&P 500 over the coming decade."

So, what’s something investors can actually do to diversify this week?

Almost no one knows this, but postwar and contemporary art appreciated 10.2% annually with near-zero correlation to equities from 1995–2025 overall.*

And sure… billionaires like Bezos can make headlines at auction, but what about the rest of us?

Masterworks makes it possible to invest in legendary artworks by Banksy, Basquiat, Picasso, and more – without spending millions.

29 exits. Net annualized returns like 16.5%, 17.6%, and 17.8% on works held over 1 year+. $1.3 billion invested. 500+ offerings.*

Shares in new offerings can sell quickly but…

👉 My subscribers skip the waitlist.*

^{*According to Masterworks data. Past performance is not indicative of future returns. Investing involves risk. Important Reg A disclosures: masterworks.com/cd.}

Google announced Gemini Spark at I/O yesterday, and it's different from the usual assistant upgrades. This one runs on dedicated cloud servers, operates 24/7, and keeps working even when your phone is locked and your laptop is closed.

The practical version: you can email Spark a task, and it pulls context from your Gmail, Docs, and calendar to execute it. Set it up to scan your credit card bills for hidden fees every month, or have it summarize every long email thread before you wake up. It already integrates with Canva, OpenTable, and Instacart, with more coming.

This matters because it crosses a line previous AI tools didn't. ChatGPT, Claude, Copilot: they all require you to be there, typing prompts, reviewing outputs. Spark is designed to work autonomously on long-horizon tasks with minimal oversight. Sundar Pichai described it as "an active partner, not a passive tool."

The enterprise version is where it gets interesting for your career. Google also launched Gemini 3.5 Flash, an agentic model that can independently execute coding pipelines, manage research projects, and (in internal tests) build an operating system from scratch. It's 4x faster than competing frontier models at a fraction of the price.

The subtext is worth reading carefully. Google isn't betting on chatbots answering questions anymore. They're betting on agents doing work. If your company deploys these tools, the question isn't whether you'll use AI. It's whether you'll manage it or compete with it.

Why it matters: The AI assistant era is shifting to the AI agent era. The professionals who figure out how to delegate to these tools, and who learn to manage AI workflows instead of just prompting chatbots, will have a real edge over those who wait for IT to roll it out.

Companies deploying AI agents are discovering they need people to manage them. Not build them, manage them. New roles like AI Integration Specialist, AI Operations Engineer, and Automation Architect are showing up in job postings at companies that were cutting traditional tech roles six months ago.

The data backs this up. 76% of companies now have a Chief AI Officer, up from 26% in 2025, according to IBM's latest CEO survey. Organizations are standing up "AI workforce management" teams to handle task orchestration between human employees and AI agents. Think of it as project management, but half your team is software.

This is showing up at every level. Entry-level: data annotation, AI testing, prompt engineering. Mid-level: AI integration, workflow design, compliance. Senior: AI strategy, ethics governance, change management. The World Economic Forum projects 170 million new roles created by 2030, with the fastest growth in exactly these categories.

Why it matters: If you're looking for where the jobs are going, follow the AI agent deployments. Every company putting agents into production needs people who understand both the technology and the business process it's automating.

→ Break into a six-figure job at top tech firms without needing to learn how to code: Join our Inner Circle

→ Have consulting or engineering background? Position yourself for Product Management, Strategy, and Operations roles: See where you stand

→ Learn how I built an award winning agency/consulting business using AI so you can start your own: Start the free course

Open Claude or ChatGPT and paste this:

"I want to set up a recurring AI workflow for my work. Here's what I do: [describe your role and 3-5 repetitive tasks you do weekly]. For each task, tell me: (1) Can an AI agent handle this autonomously today? (2) What tool would I use? (3) What's the specific setup, step by step? (4) What should I still review manually? Format it as a weekly automation playbook I can start using tomorrow."

This gives you a personalized map of what you can delegate to AI right now, not in theory, but with specific tools and steps. Most people find 3-5 hours of weekly work they can hand off immediately.

Last issue's prediction: tech layoffs by June 30 (under 175K / 175-200K / over 200K). Current pace: 138,988 through May 19, tracking toward the high end.

Anti-aging protein puts the brakes on chronic inflammation in older mice, making them stronger, more energetic, and giving them healthier bones (ScienceDaily, May 19)

Ordinary sunlight can create quantum-linked photon pairs, a phenomenon scientists previously thought required precise laboratory lasers (ScienceDaily, May 19)

Extinct plant rediscovered in the Australian outback from a random photo, after being missing for nearly 60 years (ScienceDaily, May 18)