Thirty years ago, the first Mac clones rolled off an assembly line in Austin, Texas.

If you’re not of a certain age, you might not even believe that there were once Mac clones. For most of its existence, Apple has been a singular company, selling products that were a fusion of custom hardware and custom software.

But for about three wild years in the 1990s, Apple defied its own nature and allowed other companies to build computers that ran the Mac OS and compete directly with Apple. It was an era that made some long-standing contributions to the history of the Mac, but also one that Steve Jobs dramatically ended pretty much the moment he returned to power at Apple.

But…why?

The mid-’90s were a weird time for Apple. Microsoft and Intel dominated the computer industry so much that Apple’s tiny market share just kept shrinking, and Apple struggled to make money. You might think that allowing other companies to compete with Apple would just make things worse, but desperate times called for desperate measures, and Apple CEO Michael Spindler (who had taken over for the deposed longtime CEO John Sculley) decided to take those measures.

The clone strategy was designed to allow third-party hardware makers to create systems to serve markets that Apple didn’t serve very well, allowing Mac OS to penetrate into areas where Windows was winning and turn the tide. But that didn’t really happen. Instead, Apple found itself boxed in on both sides.

At the high end, Georgia-based DayStar Digital built the first multiprocessor Mac clones targeted at the professional publishing market. DayStar built expensive computers that were faster than any of Apple’s systems, and sold them to some of Apple’s most profitable customers.

Meanwhile, Texas-based Power Computing took a page from Dell and started ramping up highly configurable beige clones, including both low-end and mid-range models. They weren’t pretty to look at, but neither were Apple’s Macs during this era. Power Computing’s killer feature was the fact that you could pick your device’s specs to order and then order it on the web (or via fax!) and it would be shipped to you. This sounds completely normal today, but Power was likely the very first company to sell computers via a web-based configure-to-order system.

Most clones were based on existing Mac motherboard designs, especially in the early days. Basically, Apple did all the core hardware and software engineering–and then let the clone makers take all of that and innovate their way into taking sales away from Apple.

When Steve Jobs took control of Apple in 1997, he put a stop to the clone licensing program immediately. (Macworld’s cover story was “Why Apple Pulled the Plug.”) “Licensees does not begin to cover their share of the expenses to engineer and market the Mac OS platform,” Jobs wrote in a memo to Apple employees. “Our Board is convinced that if Apple continues this practice the company will never return to profitability, no matter how well Apple performs, and the entire Macintosh ‘ecosystem’ will continue to decline, eventually killing both Apple and the clone manufacturers. This scenario has no winners–and customers end up with no Macintosh choice.”

Jobs was pretty ruthless in killing the clones, too. Apple’s license with clonemakers covered the 7.x versions of classic Mac OS, so Apple changed the name of its next update to Mac OS 8, so that clone makers didn’t have access to the latest version of the OS or any of the new chips that were only supported by Mac OS 8. Clone makers wanted to start building laptops, but the existing license didn’t cover that, either. And Apple announced it wouldn’t renew any of those licenses, which eliminated any hope of long-term business.

Jobs didn’t find the entire clone industry as a negative, though: Apple bought the core assets of Power Computing for $10 million in cash and $100 million in Apple stock. Part of that was probably an attempt to fend off lawsuits, but Jobs specifically cited Power’s “expertise in direct marketing and sales” and “pioneering of direct marketing and sales (a direction we want to move it)” in announcing the sale. That was recognition that Power, with its web-based direct sales strategy, had hit on something that Apple needed to replicate. Which it did!

While it’s a popular perception that the clone market was one of the causes of Apple’s near-bankruptcy in the late ’90s, Jobs wrote that it wasn’t true. Clones could really only amount for a tiny percentage of overall Mac OS computer sales–and meanwhile, “during the past two years the total number of Mac OS computers sold has declined by almost 20%.” The bottom was dropping out, clones hadn’t solved the problem, and Jobs saw them as a drain on resources that Apple simply couldn’t afford.

The legacy of the clones

There are a few surprising impacts of that brief clone era, even three decades later.

Obviously, Power’s approach of custom-building computers to order became the standard across the industry, and it certainly inspired Apple to change its processes to allow custom-built Macs to be directly ordered, rather than requiring users to pick from a set of stock configurations at a local or mail-order retailer.

The name Mac OS and the familiar two-smile symbol that now represents the macOS Finder both emerged from the clone era. Originally, there was no name for the Mac’s operating system. We generally just called it “the System,” which is why the groundbreaking 7.x update was called “System 7.” With the arrival of clones–which couldn’t be called Macs–there needed to be a way to turn the Mac operating system into a product, with a recognizable label and brand name, so that clone makers could explain that their devices weren’t Macs but ran Mac software. The result was the Mac OS name and the accompanying dual-smiley logo.

And there’s the entire approach the Mac takes to multi-threading software across multiple processor cores. It all emerged from the work Apple and clonemaker DayStar Digital did to create the DayStar Genesis MP and, eventually, Apple’s own Power Mac 9500/180 MP. In those days, the idea of putting additional processor chips inside a Mac to boost performance was novel, but if you were a high-end publishing professional, you would pay a lot of money to run Adobe Photoshop as fast as possible. So DayStar shipped a clone with two or even four PowerPC chips, and Apple and DayStar collaborated on the software that would allow a Mac OS device to send work to multiple processors.

That legacy lives on today in all of our multicore Macs. And just now, I searched my own M4 MacBook Pro for the words “DayStar Digital” and found several header files, deep down in a directory installed by Xcode, that bear the notice: “Copyright 1995-2011 DayStar Digital, Inc.” Thirty years later, that little legacy of a forgotten Mac clone lingers inside my Mac.

The following is a guest post and opinion from Mitch Liu, CEO at Theta Lab.

The debate around artificial intelligence often focuses on “knowledge workers” and creative professionals, but there’s a compelling case that sports fans may be the ideal early adopters of AI technology. While industries like real estate, e-commerce, and education lead current chatbot adoption, professional sports present a unique opportunity for AI to enhance rather than simply automate the user experience.

Sports naturally generate the kind of structured information that current AI systems handle best: statistics, schedules, historical data, and clear performance metrics. Unlike more nebulous domains, sports provide concrete data points and clear outcomes that AI can analyze effectively. The predictable cadence of games and seasons also creates natural cycles for gathering data and testing new features.

But the real advantage lies in how sports fans already consume information. They’re used to engaging with multiple data streams simultaneously—watching games while checking stats, following social media commentary, placing hopeful sports bets, and engaging in real-time discussions. This behavior aligns perfectly with AI’s ability to process and synthesize information from various sources.

Following the Numbers and Creating Experiences

Recent academic research supports this potential. An October 2024 review published in AI and Tech in Behavioral and Social Sciences found that AI is already transforming sports through enhanced analytics, predictive modeling, and personalized fan experiences. The key is that these use cases don’t replace the core emotional experience of being a fan—they augment it.

Consider how the NFL has integrated AI into its operations. The league’s Digital Athlete program, launched in 2024, uses machine learning to analyze player data and predict injury risks. While primarily focused on player safety, this same tech could help fans understand the game at a deeper level and provide insights into player performance and team dynamics that weren’t previously accessible. FIFA Manager might never be the same again!

More direct applications are emerging in fan engagement. The NHL teams, the Vegas Golden Knights and New Jersey Devils, for instance, recently partnered with our team at Theta Labs to develop an AI-powered chatbot that assists fans with inquiries about upcoming games, tickets, broadcasts, statistics, merchandise, and other curated team news and information. These interactive tools, named Sir Lancelbot and Bott Stevens, are available 24/7 on the teams’ official websites, enhancing the fan experience by providing timely and accurate responses through a customized and branded AI agent.

With a strong foothold being carved out in traditional sports, this technology is proving to click with fandoms and communities looking for ways to digitally interact with their favorite teams. It is going exceptionally well with traditional sports, but esports—with top games like Valorant, League of Legends, Rocket League, and CS—offers an even more incredible opportunity, potentially with a higher ceiling. With digitally native games, leagues, teams, and fandoms, it is an incredibly tailored use case for agentic AI.

This type of real-time proactive technology is capable of adding an interactive and life-like experience seamlessly into a digitally native environment, and some teams are already creating these experiences. Remember seeing your local club’s mascot on the sidelines at games? With agentic AI agents, those mascots can be created to have the same team spirit and aesthetics with the added benefit of infinite team and league knowledge and 24/7 interactivity. Missed a play? Just ask your team’s agent and get a recap in real time.

The Personalization Opportunity

This points to a broader opportunity: using AI to create more personalized and interactive fan experiences. Traditional sports media follows a one-size-fits-all model, but AI enables customization at scale and hyper-personalization. A casual fan might receive basic strategy explanations, while a dedicated follower gets advanced tactical analysis—all generated in real time from the same underlying game data.

This tech is particularly suited to addressing common fan pain points. Want to understand why your team’s defense collapsed in the fourth quarter? AI can analyze tactical patterns and explain the breakdown. Wondering about the historical context of a player’s performance? AI can instantly surface relevant comparisons and trends. These capabilities don’t diminish the emotional core of fandom—they enhance it by adding layers of understanding and engagement.

AI With Purpose

However, sports organizations need to approach this thoughtfully. The goal isn’t to replace traditional fan experiences but to augment them. The best implementations will feel natural and intuitive, adding value without forcing technology into places where it doesn’t belong. The human elements of sports—the community, the drama, the unexpected moments—must remain central.

The infrastructure for this transformation is already emerging. Decentralized blockchain platforms like Theta EdgeCloud are making high-performance AI computing more accessible, while advances in natural language processing are enabling more natural fan interactions. The key is applying these capabilities in ways that genuinely enhance the fan experience rather than simply chasing technological novelty.

Sports fans may not fit the typical early adopter profile, but as a group, they are shaping up to be the next frontier for driving AI adoption: regular engagement with structured data, comfort with multiple information streams, a desire for deeper understanding of their passion, and a willingness to spend a lot of time on it. As AI continues to evolve, the sports industry has an opportunity to lead rather than follow—creating experiences that demonstrate how technology can enhance rather than replace human connection.

The real question isn’t whether AI will transform the sports fan experience, but whether sports and esports organizations will seize the opportunity to shape that transformation thoughtfully and purposefully. Those that do may find themselves not just improving fan engagement, but helping to define how AI technology can enhance human experiences more broadly.