I’m feeling nostalgic this week. Perhaps it was attending a dinner with old friends in the tech/VC industry, some of whom I’ve known since the 1990s. Perhaps it was sitting down for a biographical podcast. Regardless, I was already primed for nostalgia when I read about the end of AnandTech.

For those who are younger, you might only know AnandTech as one of many different websites that covers the tech industry, especially on the hardware side. But I remember when it represented a revolution.

Anand Lal Shimpi was just 15 years old when he created Anand’s Hardware Tech Page on GeoCities. (History lesson: GeoCities was one of the first free web-page hosting companies. It was weirdly divided into virtual neighborhoods, like a real city. It was poorly managed and disorganized (I remember attending a presentation by the company’s leaders and leaving it thinking, “This company is doomed.”). But because it was the first free web page host, it grew like a weed and was fortunate enough to sell to Yahoo for nearly $3.6 billion. The lesson, as always: It’s better to be lucky than good.

Many of the folks reading this likely can’t remember the world before the internet, but there was a time when we weren’t globally connected and always on. To obtain information, people had to go to libraries to look up information in physical books. Forget mobile phones, people didn’t even want to make “long-distance” phone calls outside their area codes, because they cost so much money.

Because of this fragmentation, it was the era of gatekeepers. When I was in high school, and studying the profession of writing, the way you made a living was writing physical letters to magazines with a proposal for an article you would like to write. If your query letter was successful, you would be paid a certain amount per word when your article was accepted. One or two cents a word was standard, five cents was exorbitant. (Imagine if you had to pay ChatGPT one cent per word!)

Media was the province of professionals. If you wanted a job in media, you would have to work your way up from your local outlets, climb your way up the ladder to progressively larger markets, and if you were truly gifted and lucky, get hired by a national television network or magazine.

That’s why it was so exciting that a 15-year-old kid with a free web page could become one of the leading computer hardware reviewers, just based on the quality of his content. Anand didn’t write query letters or slowly work his way up a slippery ladder over the course of decades; he started writing great content about stuff he cared about, like graphics cards (which we now call GPUs!).

I remember when, in my Strategy class at Harvard Business School, Professor Bharat Anand (no relation) used the example of Anand Lal Shimpi to illustrate the disruptive nature of the internet. Here was a 15-year-old who had giant companies like AMD and Intel competing for his attention. It was disorienting, but also exhilarating. The “rules” of media, which had held for over a century since the rise of newspapers, had suddenly been overturned.

By July 2005, AnandTech was drawing 50 million pageviews per month, and was one of the most important publications in the technology industry. But to quote Brian Flanagan from the movie Cocktail, “Everything ends badly. Otherwise it wouldn’t end.” In 2014, Anand retired from tech journalism and handed over the reins to a new Editor-in-Chief, though he didn’t abandon his passion–he joined Apple’s hardware division. AnandTech was sold twice, first to Purch, and then to Future plc. And today, after over 27 years in business, AnandTech ceased publication (it should be noted, to its credit, that Future plc is keeping the website alive and accessible, and will continue to operate its popular forums).

27 years is a long time. When Anand first set up his GeoCities page, I was still working at my first post-Stanford job, at D. E. Shaw & Co., L.P. (even after all these years, the exact combination of ampersands, periods, and comma are burned into my memory). 27 years is long enough for an upstart to become a stalwart, and eventually, a fading veteran.

But beyond the nostalgia, I think there are some important lessons I draw from the AnandTech story.

1. Do something you love. Anand Lal Shimpi didn’t set out to build a company. He just loved hardware. But pursuing that thing he loved allowed him to build an entire career, and launched him into his second career as well.

2. Technology can disrupt entire industries, and it’s better to be the disruptor than the disrupted. AnandTech became a model for a certain kind of journalism, much like TechCrunch (which began as Mike Arrington’s personal blog) became a very different model. When technology revolutions arrive, the old rules are often swept away.

3. Every lesson has an expiration date. The AnandTech approach worked until it didn’t. The world is constantly changing around us, and we can either adapt to the changing circumstances, or stubbornly stick with the same approach until we’re forced to change. The revolutionary of today is the aging politico of tomorrow. Don’t stick with things just because they worked in the past.

4. Everything ends, so ending isn’t failure. It’s a chance to do something new. Anand Lal Shimpi left the publication that bore his name, but that didn’t make him a failure. He had a profound impact on his industry, which is more than most people get to experience. And then he got to continue pursuing his love of hardware at Apple. Endings may be bittersweet, but you should focus on the memory of the sweet beginning and middle, rather than endlessly chewing on the bitter but ultimately inevitable end.