Most universities are proud when a class fills a lecture hall. In the fall of 2011, a Stanford professor watched his online class fill up with 160,000 students — and the world of education quietly broke in half. That professor was Sebastian Thrun, the German computer scientist who had led Google’s self-driving car project, and the class was “Introduction to Artificial Intelligence.” Within months, Thrun left Stanford to start a company. He called it Udacity: audacious + university.
This is the story of that company, told the way it happened.
2011 — The class that broke the internet (before that was a cliché)
Thrun and his colleague Peter Norvig offered Stanford’s AI course online, for free, to anyone in the world. No admissions office. No tuition. 160,000 people signed up from 190 countries. When the final exam came around, more online students passed than the 200 students sitting in the actual Stanford classroom. 
Thrun later said the experience changed how he thought about teaching forever — he realized a single great online course could reach more students in one semester than he could teach in a lifetime. So in early 2012, he founded Udacity with two collaborators: David Stavens, a robotics researcher, and Mike Sokolsky, a software engineer. Their goal was simple and radical: democratize higher education.
2012–2013 — The MOOC gold rush
Udacity launched as part of the first great MOOC wave, alongside Coursera and edX. The early courses were free, built with university partners, and wildly popular. For a moment, it looked like traditional universities were about to be disrupted out of existence.
Then reality arrived. Completion rates for free online courses were abysmal — often under 10%. People signed up enthusiastically and vanished by week three. Thrun himself admitted in 2013 that the company had a “lousy product” for its original mission, and that the free-course model wasn’t producing the learning outcomes he’d hoped for.
A much-hyped partnership with San Jose State University, meant to bring MOOCs to credit-seeking students, fizzled out. The press, which had crowned Thrun the man who would reinvent college, turned on him fast.
2014 — The pivot: meet the Nanodegree
Here’s the thing about Udacity’s lowest moment: it forced the company’s best idea.
Instead of trying to replace college, Udacity decided to do something college was bad at: training people for actual tech jobs, fast. In June 2014, it announced a partnership with AT&T to launch the first “Nanodegree” — a compact, job-focused credential students could earn in 6 to 12 months, for around $200 a month. The first tracks prepared students for roles like front-end web developer and data analyst, with content shaped by what AT&T actually wanted to hire for. AT&T even promised up to 100 paid internships for graduates.
It was a fundamentally different bet from the MOOC era. Not “learn for the love of learning” — learn to get hired. The curriculum was co-built with industry partners, the projects were portfolio pieces, and the promise was employment, not enlightenment.
2015–2020 — The industry-partner era
The Nanodegree model caught on. Google, Facebook, Amazon, NVIDIA, and others partnered with Udacity to design programs in data science, machine learning, cloud computing, and self-driving cars — Thrun’s old specialty coming full circle. Students paid per program instead of per video, worked through real projects with human reviewers, and graduated with something they could show recruiters.
This is also when Udacity’s audience shifted. It stopped chasing 18-year-olds looking for a free college alternative and started serving working adults: the accountant learning Python, the QA tester moving into data engineering, the career switchers who needed proof of skills, not a diploma. If you’re on that path yourself, turning a course certificate into a paycheck is exactly the playbook Udacity’s students were writing in real time — and building a portfolio while you learn is the habit that made their Nanodegrees actually pay off.
The company raised serious money along the way — a $35 million round in 2014 to push the Nanodegree concept, more rounds after that — but it also went through layoffs and leadership changes. The road from “free education for the world” to “sustainable job-training business” was bumpy. Not every pivot worked. But the core idea held: employers would pay attention to credentials they helped design.
2020–2023 — The enterprise turn
By the 2020s, Udacity had quietly repositioned again — this time toward companies rather than individuals. Instead of selling Nanodegrees to single learners, it sold talent-transformation programs to enterprises: whole teams upskilled in AI, data, and cloud. The platform that started as a free Stanford experiment was now a B2B training business serving clients in nearly 200 countries, with content in English, Arabic, Korean, and Spanish.
It was a long way from the 2011 dream of free college for everyone. But it was, arguably, the version of the dream that survived contact with economics.
2024 — The acquisition
In March 2024, Accenture announced it would acquire Udacity, folding the company — its 230-plus staff, its library of industry-built courses, and its 21-million-strong learner base — into Accenture’s new LearnVantage training business. The deal terms weren’t disclosed. The logic was clear: in the generative-AI era, every large company suddenly needs to reskill its workforce in data and AI, and Accenture wanted Udacity’s content and expertise to sell exactly that.
Thirteen years after a Stanford AI class accidentally enrolled a small country, Udacity became part of one of the world’s biggest consulting firms. The audacious university ended up as corporate infrastructure.
What the timeline really shows
Udacity’s story isn’t a straight line from genius idea to success. It’s three different companies wearing the same name: the free-MOOC idealist (2011–2013), the Nanodegree job-trainer (2014–2019), and the enterprise upskilling platform (2020–2024). Most startups don’t survive one pivot. Udacity survived two, plus a public humiliation in the middle.
The lesson for learners is practical, not romantic. Free courses are great for exploring — choosing your first online course wisely matters more than choosing the fanciest platform. But when the goal is a job, what mattered about Udacity was never the videos. It was the projects, the reviews, and the employers at the table. Skills you can demonstrate beat certificates you can print.
Thrun set out to make Stanford’s classroom available to the world. He ended up proving something narrower and more useful: that the internet is very good at training people for jobs — once you stop pretending it’s a university.
