On Tuesday, Mitchell Hashimoto — the co-founder of HashiCorp and the mind behind Terraform, Vault, and Vagrant — announced his new company, Superlogical. Its first product: a terminal multiplexer. The Hacker News thread hit 668 points and 405 comments within hours, a mix of reverence and polite bewilderment. A terminal multiplexer? In 2026? We have tmux. We have screen. What exactly is there left to build?

Plenty, as it turns out. But the real story isn’t the multiplexer. It’s what the multiplexer says about where work is heading — and who’s betting real money on it.

The Multiplexer Is a Trojan Horse

Read past the headline and Superlogical’s pitch sharpens. This isn’t a tmux clone with nicer keybindings. Hashimoto is building a system for durable sessions — terminal environments that persist across machines, applications, and users. Close your laptop at the office, open it at home, and your shell sessions, running processes, and context are exactly where you left them. Share a session with a colleague the way you’d share a document. The terminal stops being a local scratchpad and becomes a collaborative workspace.

That’s a fundamentally different product category. It’s also a direct answer to a problem anyone who’s worked remotely for more than a week has felt in their bones: the developer toolchain was built for a world where you sat at one desk, in one building, on one machine. SSH and tmux were patches, not solutions. Six years after the pandemic scattered knowledge workers across time zones, the tools for distributed software work remain a patchwork of hacks held together by hope and muscle memory.

One senior engineer at a Fortune 500 company that just announced a five-day return-to-office mandate put it to me this way, in a Slack DM sent from a cubicle: “I’m back to commuting 45 minutes each way so I can sit on Zoom calls. The terminal is the only part of my workflow that still feels like mine. If someone makes that portable across machines, I’ll pay for it myself.”

The Office Isn’t Coming Back — and the Smart Money Knows It

It’s tempting to read the RTO headlines and conclude the remote-work experiment is winding down. JPMorgan, Goldman, and a parade of tech firms have spent the last two years tightening in-office requirements. The narrative is that distributed work was a pandemic expedient, not a permanent shift.

Superlogical is a $6.4 billion counterargument. That’s what IBM paid for HashiCorp in February 2025 — a company Hashimoto co-founded and built into the infrastructure backbone of the modern cloud. He left in December 2023, before the ink dried, and spent nearly three years watching from the sidelines. His next move wasn’t to build an office-first collaboration tool. It was to build the terminal multiplexer for a workforce that will never fully return to a single desk.

Founders with Hashimoto’s track record don’t build for yesterday’s workflow. They build for the workflow they expect to dominate five years from now. If the office were truly roaring back, the smartest infrastructure mind of his generation would be building something else. He’s not. That’s a signal worth taking seriously.

The Free Ride Is Ending — and That’s Fine

There’s a second, quieter bet embedded in Superlogical: that developers will pay for tools they’ve historically gotten for free. Tmux is open source. Screen is open source. The terminal itself is a commodity. Charging money for a terminal multiplexer sounds, on its face, like selling ice to penguins.

But the durable-sessions layer Hashimoto is building isn’t a utility — it’s a productivity platform. And the last decade of developer tools has demonstrated, repeatedly, that engineers and their employers will pay for software that saves them even modest amounts of friction. GitHub charged for private repos until Microsoft made them free. Linear charges for issue tracking. Figma charges for design collaboration. The era of “free because it’s developer tools” is fading, replaced by a willingness to pay for polish, reliability, and integration.

That’s a healthy development. Free software gave us incredible infrastructure, but it also created a world where maintainers burned out and critical tools went unmaintained. A commercial terminal multiplexer with a clear business model is more likely to still be around in a decade than yet another neglected GitHub repo with 14,000 stars and a single overworked maintainer.

The Bet Nobody’s Talking About

The Hacker News thread will debate keybindings and whether this needed to exist. That’s fine. But the real bet Hashimoto is making isn’t technical — it’s demographic. He’s betting that the most valuable software engineers in the world will, over the coming decade, demand tools that treat distributed work as the default, not the exception. He’s betting that the companies still fighting that reality with RTO mandates will lose the talent war to the ones that don’t. And he’s betting that developers will open their wallets for tools that make remote work feel less like a compromise and more like an upgrade.

If he’s right, Superlogical’s terminal multiplexer won’t look quaint in five years. It’ll look like the first brick in the foundation of how software gets built.

Sources