WeWork

Magic AI

WeWork

WeWork's model was fundamentally a real estate arbitrage play dressed up as a tech company. They signed long-term leases on buildings (often 10-15 years), spent millions renovating them, then rented desks and offices to members on month-to-month or annual contracts.

The spread between what they paid landlords and what members paid was supposed to be the profit.

The problem was the mismatch. Long-term obligations on the lease side, short-term flexibility on the revenue side.

In good times, buildings are full and the spread is healthy. In bad times — say, a global pandemic that empties offices — you're locked into paying rent on empty buildings while members cancel month-to-month.

Neumann tried to juice margins with ancillary services: WeWork Labs for startups, Powered by We for enterprise buildouts, and WeWork's own internal ventures. The company also launched WeLive (apartment living) and WeGrow (a private elementary school run by Neumann's wife).

These distractions drained cash without generating meaningful revenue.

Magic AI

Magic is building an AI that works like a software engineer, not a chatbot. The goal is a model you can hand a whole code base and a task, and it writes, reviews, and debugs like a teammate.

Their edge is memory. In 2024 they showed off LTM-2-mini, a model with a 100 million token context window.

To put that in perspective, that is about 10 million lines of code the model can look at in one shot, by far the biggest of any model at the time. The plan is to sell this as an AI coworker to companies that write a lot of software.

As of the 2024 rounds, the product was still not for sale.

WeWork

Adam Neumann was a 6'5" Israeli former naval officer with a talent for fundraising that bordered on hypnosis. Miguel McKelvey was an architect from Oregon with hippie parents who raised him in a commune.

Together in 2010, they launched WeWork from a single building in SoHo, New York — though they actually started with a predecessor called Green Desk in 2008, which was a sustainable coworking space in Brooklyn that they sold to their landlord.

The original concept was dead simple: lease entire floors of commercial buildings at bulk rates, renovate them with trendy design — exposed brick, beer on tap, inspirational quotes on the walls — then sublease individual desks and offices at a premium. The "community" angle was the differentiator.

WeWork wasn't just selling desks. It was selling belonging, networking, the feeling of being a startup founder even if you were a freelance graphic designer working alone.

The timing was perfect. After the 2008 recession, commercial real estate was cheap and available.

The gig economy was exploding. Millennials were entering the workforce with different expectations about work environments.

And startups that couldn't afford traditional office leases needed flexible space. WeWork grew from 1 location to 5 within two years, and the waitlists were long.

Magic AI

Eric Steinberger and Sebastian De Ro started Magic in 2022. Steinberger is an AI researcher who became convinced that the missing piece for a true AI software engineer was memory.

Chatbots forget what you told them a few thousand words ago. He wanted a model that could read your entire code base and remember all of it.

They raised a small seed in 2022, then a $23 million Series A in early 2023 led by Alphabet's growth fund CapitalG. From day one the pitch was not a chatbot.

It was an AI colleague that could handle real engineering work.

WeWork

WeWork grew through sheer aggression funded by seemingly unlimited capital. They would enter a city, sign leases on multiple buildings simultaneously, renovate at premium cost, and absorb losses until locations filled.

The playbook was Uber-style: spend aggressively, dominate the market, worry about profitability later.

The "community" brand was powerful marketing. WeWork events, networking mixers, and the overall vibe attracted a loyal member base that became free ambassadors.

Instagram photos of beautiful WeWork interiors drove organic demand.

Enterprise was the real growth engine. By 2019, over 40% of members were from companies with 500+ employees.

Microsoft, Amazon, and Salesforce all had teams in WeWork. Enterprise clients signed longer contracts and were more predictable than freelancers, but WeWork still spent far more acquiring and building out space than it earned from these relationships.

Magic AI

Magic's strategy is research first, product later, which is both unusual and expensive. Instead of shipping an app and iterating in public, they raised huge sums to chase a technical breakthrough in model memory.

The headline move was the 100 million token context claim, which got the whole AI world talking. They also cut a deal with Google Cloud in 2024 to build two supercomputers, called Magic-G4 and Magic-G5, using NVIDIA chips, to train bigger models.

Star backers like Eric Schmidt served as validation. The bet is that if the tech is far enough ahead, the product and the customers follow.

WeWork

Where to begin? The S-1 filing in August 2019 was a masterclass in red flags.

It revealed that WeWork lost $1.9 billion in 2018 on $1.8 billion in revenue — spending more than a dollar for every dollar earned. Neumann had taken $700 million off the table through stock sales and loans before the IPO.

He owned the "We" trademark personally and charged the company $5.9 million to license it. He had family members on payroll.

He flew on a private jet funded by the company.

The planned $47 billion IPO was pulled in September 2019 after investors revolted. The valuation was slashed.

Neumann was forced out and given a $1.7 billion exit package — for nearly destroying the company. SoftBank took control, cut thousands of jobs, and spent years trying to restructure.

WeWork finally went public via SPAC in 2021 at a $9 billion valuation.

Then COVID hit the commercial real estate market like a meteor. Remote work became permanent for many companies.

WeWork's occupancy plummeted. They filed for Chapter 11 bankruptcy in November 2023, listing $18.7 billion in debt, wiped out the equity, shed hundreds of leases and came back out in June 2024 as a far smaller private company.

The cautionary tale was complete.

Magic AI

Magic has raised close to half a billion dollars and, years in, still has no product the public can buy. That is the whole problem in one sentence.

A giant context window is a great demo, but rivals like Cursor, GitHub Copilot, and Cognition's Devin already have real users and revenue. Magic has to prove that more memory actually makes a smarter engineer, not just a bigger number on a slide.

And it has to do that while burning enormous sums on compute against competitors who are already shipping. The clock and the cash both keep running.

WeWork

WeWork All Access — a membership that gives access to any WeWork location worldwide, targeting remote workers and traveling professionals. Dedicated Desks — assigned workstations in shared open-plan spaces for individuals and freelancers.

Private Offices — enclosed offices for teams, the bread-and-butter product generating most revenue. WeWork Workplace — enterprise software for managing hybrid work, office scheduling, and space utilization analytics.

On Demand — pay-by-the-day access to meeting rooms and workspace without a monthly commitment.

Magic AI

The centerpiece is Magic's LTM, or Long-Term Memory, model family. In 2024 they unveiled LTM-2-mini, a model with a 100 million token context window, pitched as the key to an AI that can understand an entire code base at once.

The broader product is an AI software engineer, or coworker, that can write and change code across a whole project. The company also announced Magic-G4 and Magic-G5, two supercomputers built with Google Cloud and NVIDIA to train these models.

Much of it is still research rather than a shipping product.

WeWork

SoftBank Vision Fund was the largest and most consequential investor, pouring over $10 billion into WeWork across multiple rounds. Masayoshi Son reportedly agreed to invest after a 12-minute meeting with Neumann.

Benchmark was an early investor. JPMorgan Chase provided debt financing.

T. Rowe Price, Fidelity, and Goldman Sachs participated in later rounds.

The company raised more money than most companies ever generate in revenue.

Magic AI

Magic's backers are a murderers' row of AI insiders. It started with a $5 million seed round in 2022 and a $23 million Series A in early 2023 led by Alphabet's CapitalG.

Then in February 2024 came a $117 million round led by Nat Friedman and Daniel Gross, two of the most connected investors in AI. The blowout was August 2024.

Magic raised $320 million with checks from former Google CEO Eric Schmidt, CapitalG again, Atlassian, Jane Street, Sequoia, and Elad Gil. That pushed total funding to around $465 million, all for a company still building toward its first sellable product.