The $650 Billion Bet: Why Big Tech Is All-In on Agentic AI

Salutations, Olio aficionados! 👋

Welcome to the 234th edition of Weekly Olio. We’re thrilled to introduce a fresh new twist to your Sundays: Publisher Parmesan, our hand-picked, thoughtfully crafted edition designed to spark inspiration and insights for the week ahead.

It’s the perfect way to unwind, recharge, and prepare for the week with something truly worth savoring.

If you’re new here and you’re looking for more long-form, crispy writing, click the link to subscribe under this GIF 👇

ChatGPT is a superpower if you know how to use it correctly.

Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.

Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.

Download the free guide.

Imagine this: “Book dinner for four on Saturday somewhere close.”

And your phone does the rest. It checks everyone’s availability, filters restaurants, makes the reservation, updates your calendar, messages your friends, and sends you the confirmation. You never touch the screen.

That is the future Silicon Valley is spending $650 billion chasing in 2026 alone. Not across a decade. This year.

According to Fortune, the combined AI infrastructure spend by Amazon, Google, Meta, and Microsoft now rivals the entire GDP of Sweden. To put that in perspective: India’s entire IT services industry — TCS, Infosys, Wipro, HCL and the rest — generates roughly $280 billion in annual revenue. Big Tech is spending more than twice that on AI infrastructure in a single year.

So what are they seeing that the rest of us aren’t? It’s not ChatGPT.

We’ve all used generative AI. It writes emails. Summarizes documents. Generates images. It’s impressive. But it hasn’t transformed the economy.

McKinsey’s latest research shows that 78% of companies are already using some form of generative AI. Yet only 39% can trace measurable financial impact from it. In many cases, it’s helpful — not essential.

The “wow” phase is fading. Chatbots are becoming commodities. And when the magic wears off, Wall Street asks a simple question: where’s the revenue? The answer the industry has landed on is something called agentic AI.

Most AI today is reactive. You type a prompt. It responds. Close the tab, and it forgets you. Agentic AI is different. Instead of answering a question, it receives a goal.

Not: “What restaurants are nearby?”

But: “Book dinner for four this Saturday.”

An agent doesn’t just generate text. It breaks the task into steps. It searches. It compares. It checks your calendar. It logs into services. It executes. It verifies the outcome. Minimal handholding. Task in, outcome out.

That shift — from assistant to actor — is why leaders like Sam Altman, Jensen Huang, and Satya Nadella are pivoting hard.

Huang has been explicit: there are roughly a billion knowledge workers in the world. That’s the market. Analysts, coders, customer support agents, sales reps — anyone paid to process information and make decisions.

Automate even part of that, and you’re staring at a multi-trillion-dollar opportunity. That’s what $650 billion is buying.

But here’s where things get uncomfortable. To book your dinner seamlessly, the AI agent would need:

It doesn’t just read them. It acts inside them. It marks your calendar. It sends messages. It charges your card. Meredith Whitaker, president of Signal and a longtime AI critic, described what this really implies: root permission.

Root permission is the deepest level of system access — typically reserved for the operating system itself. In plain English, you hand the AI the keys to everything.

And it likely doesn’t run entirely on your device. Your data moves to the cloud, gets processed, and comes back. Messages, financial data, personal history — all in transit across infrastructure you do not own.

Whitaker calls this “putting your brain in a jar.” It’s not a metaphor that leaves easily.

He's measuring in gold, not dollars. And that's the point.

The dollar dropped 10% in 2025. So, yeah, your portfolio went up in dollars, but, Dalio says your real return isn’t so exciting.

And the decline is reportedly advancing as macro conditions don’t improve.

So, what investments offer protection against that currency risk?

Well.. billionaires have an answer. And now 70,819 everyday investors have joined in.

This unexpected asset class outpaced the S&P 500 overall with low correlation since 1995.*

Not real estate or PE. Post war and contemporary art. Seriously.

Plus– Art trades globally ;)

And now, you don’t need to be a billionaire–

Masterworks makes it easy to FRACTIONALLY invest in blue-chip art, with a track record of 26 net annualized returns like 14.6%, 17.6%, and 17.8% on works held over a year.

See why investors moved $1.3 billion into 500+ offerings:

Skip waitlist

_{*According to Masterworks data.  Investing involves risk. Past performance not indicative of future returns. See important disclosures at masterworks.com/cd.}

So why deploy this at breakneck speed? Because the financial pressure is enormous.

When you commit $650 billion to infrastructure, investors expect returns. Not someday. Quarterly. Agentic AI is the story that justifies that capital expenditure.

The prior wave — chatbots, image generators — didn’t move revenue enough. Agentic AI promises something different: replacing labor.

That’s not speculative. Salesforce has already said AI agents are handling half of its customer interactions, allowing the company to eliminate thousands of service roles.

The economic logic is clear. If an agent can handle structured, repetitive, decision-heavy work, the ROI becomes obvious. But that’s in controlled environments. The real world is messier.

Gartner predicts that over 40% of agentic AI projects will be canceled by 2027. Why?

Agents perform best when:

They struggle when:

Booking concert tickets in a demo is easy. Making a complex regulatory compliance decision isn’t. Yet marketing materials blur that distinction.

Underneath all of this lies the question no one has resolved: Who is responsible when an agent makes a catastrophic mistake?

If it sends sensitive financial data to the wrong recipient, who’s liable? If it executes a flawed transaction, who pays? If it misinterprets instructions, who absorbs the damage?

These aren’t engineering problems. They’re governance problems. And governance moves slower than venture capital.

Whitaker’s core warning isn’t that agentic AI won’t work. It’s that financial incentives are structured to push deployment faster than safeguards mature. The infrastructure bill is due. The product must ship.

It’s important to be honest. Agentic AI isn’t vaporware. This isn’t the metaverse cycle. Real productivity gains are happening in specific domains. Certain workflows are already faster and cheaper.

But the leap from “automated customer support” to “AI runs your life” is enormous. The promise is genuine. So are the risks.

Opting out entirely may not be realistic. These systems will likely become embedded in operating systems, business software, and enterprise tools whether individuals embrace them or not. The more meaningful choice is awareness.

Understand what access you’re granting. Understand where your data flows. Understand what recourse exists when something goes wrong.

Tech CEOs describe this as the future of work. Critics describe it as the future of surveillance. The uncomfortable truth is that it may be both.

And right now, the industry is building it at a speed that suggests we may only decide what we’re comfortable with after it’s already here.

Interested in learning more about AI? Check out our previous coverage here:

High-stakes communications need precision. Wispr Flow turns speech into polished, publishable writing you can paste into investor updates, earnings notes, board recaps, and executive summaries. Speak constraints, numbers, and context and Flow will remove filler, fix punctuation, format lists, and preserve tone so your messages are clear and confident. Use saved templates for recurring financial formats and create consistent reports with less editing. Works across Mac, Windows, and iPhone. Try Wispr Flow for finance.

Try Wispr Flow

That’s all for this week. If you enjoyed this edition, we’d really appreciate if you shared it with a friend, family member or colleague.

We’ll be back in your inbox 2 PM IST next Sunday. Till then, have a productive week!

Peeyush and Mohit

Disclaimer: The views, thoughts, and opinions expressed in the text belong solely to the author, and not necessarily to the author's employer, organization, committee or other group or individual.