Bill Gates is warning that artificial intelligence could take away millions of jobs. Yet the technology is also beginning to remove many of the costly barriers that corporations have placed between ordinary people and the ability to earn a living.
Bill Gates has been warning people about the jobs artificial intelligence could take away. In August 2026, the Microsoft co-founder published an essay predicting serious disruption to employment, with workers in administration, software development, customer service and other professions facing an uncertain future. He even suggested that he would support a credible international effort to slow AI development, although he doubted such an arrangement could be achieved.
Gates has made an enormous fortune from the computer industry. Microsoft, the company he co-founded with Paul Allen in 1975, helped determine how people would use computers for much of the following half-century. Its operating systems, office applications and business software became so widely used that employers began expecting workers to understand them, and an entire training and certification industry developed around that expectation.
Anyone who has worked in IT will be familiar with the pressure to collect qualifications from companies such as Microsoft, Cisco and other technology suppliers. People spend months studying computer networking, Windows Server, Active Directory, cybersecurity and system administration. They pay for courses, examinations and study materials, often because an employer has decided that a particular certificate is necessary to qualify for a job.
Once they have the qualification, they must keep it current. Microsoft’s associate and expert certifications generally expire after one year, although eligible holders can renew them without charge before expiry. Cisco’s professional certifications generally operate on a three-year cycle. New technologies arrive, examinations change and old qualifications are eventually retired.
The companies explain that these arrangements ensure people maintain current knowledge. There is some truth in that. Technology changes and nobody wants an administrator working on critical systems without understanding what they are doing. But the amount of time people spend learning the particular procedures of commercial software has become an expensive burden in its own right.
Someone who has spent years dealing with failed network connections, DNS problems, faulty drivers, Windows updates and servers that suddenly stop responding may understand the work perfectly well. Yet an employer can still overlook that experience because the applicant does not possess a particular certificate.
There are technicians who have collected numerous qualifications and are very good at their jobs. There are also people who have learned the material well enough to pass an examination but have difficulty resolving problems that do not resemble the exercises they studied. Years of practical experience cannot always be measured by an examination written by the company that sells the software.
Microsoft has benefited from this dependence for decades. Businesses adopted Windows, built their networks around Microsoft technologies and trained their employees to use its applications. Changing systems could mean retraining staff, replacing software and dealing with compatibility problems. Organisations already struggling with their daily operations were hardly in a position to keep changing their entire computing environment.
Microsoft’s willingness to protect its position was established in court. In June 2001, a United States federal appeals court upheld findings that the company had unlawfully maintained its monopoly in the personal computer operating system market. The case involved restrictions on computer manufacturers, Microsoft’s treatment of competing software and measures intended to protect Windows from technologies that could weaken its dominance.
Gates was chairman and chief software architect at the time. He helped build a company that became wealthy partly because its products were so deeply embedded in the working lives of businesses and their employees. Microsoft was hardly the only corporation to benefit from such dependence, but its influence over the computer industry was difficult to escape.
We have reached a point where people are expected to spend large amounts of money preparing themselves for employment before they have even earned a living. A university degree is required for many positions, followed in some industries by professional certifications, additional training and years of experience. A young person may complete three or four years of study, accumulate student debt and still be told that another qualification is necessary.
University education has its place, and some professions require years of serious academic preparation. Nobody should want an unqualified person designing a bridge, prescribing medication or taking responsibility for the safety of others. But the expectation that people must continually purchase qualifications to demonstrate their usefulness has spread into work that can often be learned through experience, practical training and access to good information.
Employers have helped sustain this arrangement. Qualifications provide a convenient way to filter applicants, sometimes with little consideration for what the job actually requires. A capable person can be rejected before anyone has an opportunity to assess their ability.
Artificial intelligence is beginning to make some of these barriers less necessary.
A technician investigating a failed network connection can describe the symptoms to an AI assistant, provide the relevant error messages and work through possible causes. AI can explain the meaning of Windows event logs, help examine DNS and DHCP configurations, identify possible Active Directory problems and suggest commands for investigating a server that has stopped communicating with other machines.
The technician must still understand the environment and check whether the advice is correct. An incorrect configuration change can cause serious problems, particularly in a business network. But having access to technical knowledge while doing the work reduces the need to memorise every procedure or spend hours searching through manuals.
Someone without a formal IT qualification may also be able to learn these skills through practical work, with AI explaining unfamiliar concepts as problems arise. Knowledge that previously required expensive training is becoming available to people who could not afford it.
The same applies beyond computing. People can use AI to understand accounting procedures, interpret complicated documents, learn technical skills and complete administrative work that previously required specialised assistance. It does not make every person an accountant, lawyer or engineer, but it allows them to do more for themselves and to understand when professional help is genuinely needed.
For families and small businesses struggling with the cost of living, that is a considerable benefit. They may no longer need to pay someone every time a computer develops a problem, a document needs preparing or an unfamiliar procedure must be understood.
Gates himself acknowledged the change in his August essay. He explained that previous computing revolutions took years to become widely adopted because people had to learn how to use the technology. Artificial intelligence, he observed, works differently because it understands natural language.
“We don’t have to adapt to it because it can adapt to us,” he wrote.
For decades, the computer industry expected people to adapt to its products. Workers learned complicated commands, employers paid for training, and students devoted years to obtaining the qualifications that businesses demanded. Gates now recognises that AI can perform much of the work involved in understanding those systems.
His warnings about job losses deserve serious consideration. Research is already finding evidence of employment pressure among younger workers in occupations exposed to AI, and there is no reason to assume that every person displaced by the technology will easily find another job.
But workers have also been living with the financial and personal costs of technological change for decades. They have had to adjust to whatever systems corporations introduced, learn whatever skills employers demanded and find the money to qualify for jobs that were supposed to provide them with a living.
It is difficult to listen to warnings about AI destroying the established employment system without also examining how much unnecessary difficulty has been built into that system.
A person should not have to spend years obtaining expensive credentials merely to demonstrate an ability to accomplish ordinary work. Nor should someone who has learned a trade through experience automatically be considered less capable than a person holding a university degree.
Gates continues to exercise influence in technology and public policy through his foundation, investments and relationships with businesses and governments. In his August essay, he acknowledged that he still had financial ties to the technology industry and worked with Microsoft and other AI companies.
He also said he intended to raise AI policy with lawmakers and international leaders and advocate for stronger national and international arrangements to manage the technology.
His stated reasons include protecting workers, reducing inequality and preventing harmful uses of AI. Those are legitimate concerns. His record at Microsoft nevertheless gives the public good reason to scrutinise how much influence established technology interests should exercise over the rules governing AI.
After all the wealth and power accumulated through the computer industry, Gates remains closely involved in discussions about how its next generation of technology should develop. Microsoft itself is investing heavily in AI and selling subscriptions and services around it.
The public has already experienced what happens when essential technology becomes concentrated in the hands of a small number of companies. Software licences, recurring subscriptions, technical requirements and professional certifications can leave individuals and businesses with little practical choice but to keep paying.
If AI becomes another system controlled by the same corporate interests, people could find themselves dependent on expensive subscriptions and new qualifications simply to access the capabilities that were supposed to make their lives easier.
Open-source technology and AI models that can operate on privately controlled computers offer some alternatives. They require investment and technical knowledge, but they also allow individuals and businesses to retain more control over the tools they use.
I would rather see artificial intelligence help ordinary people become less dependent on these corporations. We have already spent enough years paying to learn their software, adapting to their systems and meeting qualifications they persuaded employers to recognise.
The discussion about AI taking jobs also overlooks the amount of work that should never have been necessary in the first place. People spend hours resolving software problems, navigating complicated procedures, completing repetitive administration and learning systems that could have been designed to be much simpler.
Much of that work has become normal because businesses and institutions have organised themselves around it. People accept the burden because they need employment, income and some security for their families.
If AI can remove those difficulties, I would welcome it.
I would welcome a future where someone can work competently without spending thousands of dollars collecting certificates, where practical experience receives the recognition it deserves and where small businesses can operate without continually paying specialists to manage unnecessarily complicated systems.
The benefits should also extend beyond the workplace. People ought to have more time for their families, their communities, their faith, their interests and the ordinary things that give life meaning. Technology was supposed to make that possible long ago, yet every improvement seems to have arrived with another demand upon our time.
The people warning us that AI will take away work should also consider how much of that work consists of burdens that technology companies and modern institutions have created.
There will always be work that requires human knowledge, responsibility and judgment. People will still need to earn a living, and governments will have to address the consequences when jobs genuinely disappear. But we should not preserve unnecessary complexity simply because an industry has learned to make money from teaching people how to deal with it.
Gates has had decades to influence the way people work with computers. Ordinary people should now have an opportunity to benefit from technology that requires less of their money, less of their time and fewer permissions from the companies that supply it.
ABOUT THE AUTHOR
John Fuiava operates IOANET NZ Limited , an Auckland-based IT and internet services company. His work involves computer networking, server administration, web hosting and technical support for businesses. He writes about technology, online security and safety, and the importance of making technology work for people rather than people working for the technology .
