What’s happened to the small guy or gal in computing? I remember going to meet a guy in Hillbrow to buy software from him. This was in the early days of personal computers. He lived in a flat, a high-rise flat, and in those days it was a sought-after place to have a flat. The block of flats was situated to the left of Empire Road, near to the Hillbrow (originally the JG Strijdom Tower in Banket Street). I bought some specialised software from him, and took it home and read the instructions and got it operating, and it was very useful. Alas, that specific program, which was excellent, no longer has anything that resembles it. This got me thinking about what has happened to the small guy or small gal in computing these days.
Over the decades, we’ve seen that, right at the beginning of personal computers, there were many personal computer manufacturers. There were two operating systems, and then Microsoft grabbed one operating system and is still making billions out of it, and has got extremely restrictive about its use. It does updates and offers no support for the old versions. You know the story. The same happened with the internet. There were a lot of browsers and search engines in the early days, and then Google, and I’m not knocking Google, but Google got in and took over. Another giant company formed. Then came cell phones and smartphones, and various app makers were able to sell their apps on platforms like Apple and Android. Now, what’s happening to those smaller business people who come up with very handy and useful products? How is AI faring? You can see that the whole AI game now is around huge investment and big companies getting involved. Will it place the small business person out of reach of the market? Well, we don’t know for sure because things are evolving. But perhaps with AI, the smaller business person will be able to come up with artificial intelligence apps that would be stronger and more niched than what is happening in the big AI world. You also get some guys that are just philanthropic, and that means, like when I came across a transcription app for the desktop or laptop that he made while his hand was hurt, and he couldn’t type. So he came up with this app, which he continues to develop with help from friends, and it’s absolutely free. It’s not as fast as, say, one of the new voice-to-text transcription apps, but it’s pretty accurate, and it’ll do the job for most types of dictation. This is a fantastic story, and there are a lot of other people out there who are making software for free, and all they require is a donation. Where this will all go, we don’t know. But the main thing – it’s always a tragedy when the smaller business person, the guy or gal who wants to get in on the act, and has a love and passion for computing, will be shoved out by the giant corporations.
The Early Days: A Market Full of Tinkerers
That software seller in Hillbrow was not unusual for his time. The personal computer industry began with hobbyists and tiny firms, not giants. In 1975, a small electronics-kit company called MITS launched the Altair 8800, a roughly four-hundred-dollar kit that generated thousands of orders in its first few months. It spawned an entire ecosystem of add-on companies like Cromemco. By 1977, the so-called Trinity of pre-assembled machines arrived – the Apple II, the Tandy TRS-80, and the Commodore PET. They were joined by dozens of others: Atari, Sinclair, Texas Instruments, Acorn, and Japanese players like Hitachi and Sharp. Each used incompatible hardware and software. Operating systems were optional extras. CP/M from Digital Research emerged as an early standard, with more than two hundred and fifty thousand licences sold by 1981 and a large library of compatible software. It was a golden age for the small guy. Anyone with a good idea could build a computer in a garage or write a program in a bedroom and sell it from a flat in Hillbrow.
The Great Squeeze: How Microsoft Locked Down the PC
The fragmentation ended with a single strategic deal. In 1981, IBM needed an operating system for its new PC. Microsoft, then a twenty-five-person language-tools company, acquired the rights to a system called 86-DOS from Seattle Computer Products for a total of seventy-five thousand dollars, adapted it into MS-DOS, and licensed it to computer manufacturers rather than selling it outright. This was the masterstroke. Because Microsoft retained the rights, it could license MS-DOS to the flood of IBM clone manufacturers that followed. By the late nineteen-eighties, MS-DOS ran on over eighty percent of personal computers. Then came Windows 3.0 in 1990, which cemented Microsoft’s dominance. Lotus and WordPerfect failed to transition from DOS quickly enough. Microsoft released Excel and Word for Windows with tight integration into the operating system, including early access to new technologies that competitors did not get. Soon Microsoft dominated applications too. By the nineteen-nineties, it held an eighty-five percent share of the office-suite market. The small hardware and software makers of the seventies were largely gone or marginalised. That excellent program you bought in Hillbrow probably died somewhere in this consolidation, unable to compete with a giant that could afford to give away its own software until the competition vanished.
From Netscape to Google: The Internet Gets Swallowed
The internet opened a new frontier, and for a brief moment the small player was back on top. Netscape Navigator, created by Marc Andreessen’s team, controlled around eighty to ninety percent of the browser market by early 1996. It introduced innovations we still use today, including JavaScript, cookies, frames, and plug-ins. Then Microsoft woke up. Bill Gates issued his famous Internet Tidal Wave memo, and Microsoft began bundling Internet Explorer for free with Windows, pressuring PC makers to hide Netscape. By 2002, Internet Explorer held around ninety-five percent market share. Netscape was effectively dead.
But the browser was only the gateway. The real prize became search. Google, founded in 1998, eventually came to dominate so thoroughly that today it holds roughly ninety-one percent of global search-engine referrals. Research shows most users never actively choose Google. They simply stick with the default. Once again, a layer of technology had consolidated under a single giant.
The Smartphone Reprieve: When the Little Guy Got a Second Chance
Smartphones could have become another walled garden dominated solely by Apple and Google. Instead, they created platform ecosystems that let the small guy back in. The App Store in 2008 and Google Play gave individual developers global distribution overnight. The numbers are staggering. The App Store ecosystem generated $1.1 trillion in total billings and sales in 2022. Small developers in particular thrived. Their earnings grew seventy-one percent between 2020 and 2022, outpacing larger developers. Research found that forty-five percent of developers earning over one million dollars on the App Store in 2021 had either not been on the store or had earned less than ten thousand dollars just five years earlier. Solo founders built apps for panic attacks, Hispanic job seekers, and service professionals, reaching millions of users. The platform layer was concentrated, but the application layer democratised. For a while, the small guy had a real shot again.
AI: The New Battleground
We are now watching the cycle repeat at hyperspeed. At the foundation-model layer, power is concentrating among a handful of American and Chinese giants. In the United States, OpenAI, Google, Microsoft, Amazon, and Meta dominate the headlines. In China, Tencent, Alibaba, ByteDance, Baidu, and SenseTime are racing to build large language models. Training these models costs hundreds of millions in compute. The small player cannot build a competitor to GPT-5 in a garage. The infrastructure layer – GPUs, cloud compute, base models – is consolidating just as operating systems and search did before it.
Finding the Cracks in the Wall
But history gives a clear answer. You cannot compete at the infrastructure layer, but you can dominate at the application layer. The most saturated, dead-end markets in AI today are generic ChatGPT wrappers – writing assistants, support chatbots, meeting summarizers, and resume builders that the big players already ship natively. That is not where the small guy wins.
The wide-open categories read like a playbook for the focused entrepreneur. Vertical software for so-called boring industries like HVAC, roofing, pest control, and auto repair. These are hundred-billion-dollar industries that still mostly run on spreadsheets and phone calls. Big AI players ignore them. AI compliance and regulatory technology is exploding because the EU AI Act’s new enforcement regime, which began in August 2026, creates a deadline-driven market that generic tools cannot satisfy. Creator economy micro-tools targeted at podcasters, Twitch streamers, or newsletter writers so specific that Canva and Notion will never build for them. Domain-specific AI trained on proprietary data in law, medicine, and manufacturing, where accuracy and compliance matter most. The moat is the data and the niche, not the model.
Perhaps most importantly, AI-assisted development tools have given the solo developer superpowers. Modern coding assistants and browser-based integrated development environments let one person do the work of what used to be a small team. The fastest path from idea to deployed product is now open to individuals again.
The Main Thing
So where does this leave our small guy or gal? It leaves them exactly where they have always been in computing – scrappy, niche-focused, and forced to outmanoeuvre rather than outmuscle the giants. The tragedy is real. That excellent program from Hillbrow is gone, swallowed by a monopoly that made compatibility impossible. Netscape is a memory. Google’s dominance means most new search ideas die in obscurity. The foundation models of AI are being built by corporations with budgets larger than the GDP of some nations.
And yet. The free transcription app I found, built by a guy with a hurt hand and maintained by friends, is the same spirit that built the Altair, that coded CP/M in a bedroom, that launched a million apps from kitchen tables. The pie is enormous. The trick, as always, is to carve out a slice so specific that the giants cannot reach their fork that far. The small guy or gal with a love and passion for computing has been declared dead a dozen times before. They are still here. They will still be here. They just have to be smarter, faster, and more stubborn than the corporations trying to shove them out.
