From Mechanical Calculators to AI: Where Bitcoin Fits

The calculator is among the most taken-for-granted inventions in human history. Available in almost every pocket, few stop to consider how affordable calculation transformed business, science and everyday life.

When ChatGPT arrived in 2022, a simple comparison came to mind: an English calculator. Much like a traditional calculator brings order to numbers, AI language models can help bring order to text. Except AI's capabilities extend far beyond language and arithmetic.

When capabilities once reserved for specialists become affordable to millions, entirely new possibilities emerge for individuals and businesses.

The history of calculators offers a fascinating perspective on AI's future. Bitcoin adds another dimension: greater independence over the money people earn, save and transfer.

The $475 Burroughs Adding Machine

In 1892, the American Arithmometer Company introduced an improved Burroughs Registering Accountant. According to the Early Office Museum, the machine sold for approximately $475, equivalent to roughly $17,600 in 2026 purchasing power, based on the Federal Reserve Bank of Minneapolis historical price series.

Mechanical calculators had existed for centuries, but Burroughs helped make automated arithmetic commercially practical for American offices.

Banks, insurers and merchants relied on clerks to add figures, calculate balances and maintain financial records. Early Burroughs machines were bulky and sometimes produced incorrect answers, but mechanical improvements gradually made them more reliable.

By 1909, Burroughs advertising claimed a clerk could add columns of figures in 90 seconds using a machine, compared with nine minutes by hand. Economist William Nordhaus reproduced the sixfold improvement in Two Centuries of Productivity Growth in Computing.

The figures came from company advertising rather than an independent experiment. Nevertheless, the commercial attraction was clear: more accounts could be processed with less repetitive arithmetic.

Mechanical Calculator Adoption: 1890–1920

Burroughs was hardly alone. Inventors including Dorr Felt helped develop practical calculating machines as demand for faster bookkeeping increased.

Nordhaus estimated the cumulative number of computational devices manufactured across the industry, excluding abacuses and counting boards.

Source: Nordhaus (2007), Table 1. Rounded estimates covering multiple manufacturers and device types.

Production grew dramatically, especially after 1900. Yet widespread household ownership was still decades away.

The lesson: Even a groundbreaking invention can take generations to become reliable, understood, trusted, affordable and part of everyday life.

Did Calculators Replace Office Clerks?

Faster arithmetic raised an obvious concern: would businesses need fewer people?

Historical US Census occupational figures reveal a surprising outcome. Employment among bookkeepers, accountants and cashiers rose from approximately 159,000 in 1890 to 931,000 in 1930, nearly sixfold.

Calculators alone cannot explain the expansion. America's population and economy grew, businesses maintained more financial records, and office employment became increasingly accessible to women.

Economic historian Elyce Rotella documented how mechanization changed clerical work and contributed to changing hiring patterns, particularly among women.

A broader explanation comes from economist James Bessen. Research covering textiles, steel and automobiles shows how falling production costs can expand demand enough to support additional employment. Eventually, market saturation can reverse the effect.

Automating a task does not necessarily eliminate the job. Greater efficiency can also create room for more business activity.

From Pocket Calculators to Smartphones

The next transformation involved size and price.

Electronic components gradually replaced bulky mechanical systems, bringing powerful calculations into smaller and more affordable devices. When Hewlett-Packard introduced the HP-35 scientific calculator in 1972, the $395 price tag represented approximately $3,180 in 2026 purchasing power. Barely half a century later, comparable mathematical functions became standard features of smartphones.

The Smithsonian's calculator collection records several milestones:

• 1972: Hewlett-Packard HP-35, $395.

• 1977: Teal LC811, $24.95.

• 1985: Sharp EL-345 solar calculator, $5.95.

The models offered different capabilities, so the prices are not directly comparable. The broader direction, however, is unmistakable. Calculators entered classrooms, homes and small businesses. Personal computers brought electronic spreadsheets into offices with the introduction of VisiCalc in 1979, while smartphones, including Apple's first iPhone in 2007, made calculators a standard feature of everyday mobile devices.

Once important enough to justify teams of clerks and expensive machinery, calculation became an everyday convenience requiring almost no thought.

AI as a Calculator for Language, Coding and Research

A similar possibility is emerging beyond arithmetic.

AI can assist with languages, computer programming, image and video creation, scientific research, engineering, medical decision-making and countless other fields. Capabilities once requiring years of training or expensive professional services are becoming more accessible, although human expertise and independent verification remain important.

An unexpected discovery emerged while researching history for this article.

The original 1955 Dartmouth AI proposal, a founding document of artificial intelligence, described programmable computers as “automatic calculators” capable of simulating other machines. The proposal also explored how computers might use language.

The connection between calculators and AI was present long before conversational AI reached the public.

Early Burroughs machines made mistakes, cost a fortune and took years to earn the trust of businesses. AI faces similar challenges, but the potential is enormous. A single person could build a company, develop software, serve customers and manage operations once requiring an entire team, at a fraction of the cost.

The pace of AI adoption helps explain why Coinjuice closely follows developments across the sector. Our research comparing Bitcoin and AI-related stocks explores their market performance, with further analysis covered during Coinjuice LIVE broadcasts.

Anthropic AI Adoption: $100,000 Customers by Year

The speed of commercial adoption becomes clearer through Anthropic.

In its October 2026 SEC filing, Anthropic reported the number of customers generating at least $100,000 in revenue for the company over the preceding twelve months.

Source: Anthropic Form S-1, filed October 26, 2026. Customer counts are rounded. The September 2026 observation arrives nine months after December 2025.

Anthropic's high-spending customer base grew roughly tenfold during 2024, then more than sevenfold during 2025. By September 2026, the count had expanded almost eightfold again in just nine months.

Why AI Is Spreading Faster Than Mechanical Calculators

Unlike mechanical machines, AI services can reach businesses through existing computers and internet connections. Behind the software, enormous investment in chips, electricity and data centers remains necessary.

Although both technologies experienced rapid growth, the figures measure different stages of adoption:

• Mechanical calculators: Nordhaus estimated production grew from just 1,500 machines in 1890 to 910,000 by 1920. Early Burroughs machines cost approximately $17,600 in 2026 purchasing power.

• Artificial intelligence: Anthropic's customers generating at least $100,000 annually grew from approximately 20 in 2023 to more than 11,900 by September 2026.

The figures cannot be compared directly. However, both reveal a familiar pattern: adoption begins with a relatively small customer base and then accelerates. 

The major difference is speed. AI can reach thousands of businesses through existing digital infrastructure, without manufacturing and delivering a machine to every office.

Beyond Adoption: AI Productivity in Practice

Anthropic's customer growth tells only part of the story. The company's October 2026 SEC filing also reveals how AI is changing the amount of work employees can produce and how much businesses pay for each task.

Three findings from Anthropic's October 2026 SEC filing stand out:

• Over 90% of code contributions: Claude wrote more than 90% of contributions to Anthropic's internal codebase during September 2026.

• 45 times more code: Average daily code contributions per engineer during July–September 2026 exceeded the 2024 daily average by more than 45 times. Greater code volume, however, does not necessarily mean an equivalent increase in completed projects.

• 25% to 40% lower costs: New Claude models reduced customer costs for typical Fable workloads by approximately 25% and Opus tasks by approximately 40%, compared with earlier versions.

Although company-reported figures cannot establish productivity gains across the wider economy, Anthropic provides a remarkable example of AI's commercial potential.

Just as mechanical calculators reduced the time needed for arithmetic, AI is reducing the time and cost of increasingly complex work.

How AI Could Change Small Businesses and Employment

Consider an independent plumber.

Answering calls, arranging appointments, preparing quotes and managing invoices can consume hours better spent on paid work or rest. AI voice agents can answer incoming calls, collect job details and send summaries to the plumber. Rather than interrupting a repair, the business owner can review each enquiry, decide whether to accept the work and prepare a quote afterward.

The benefit of AI extends beyond saving time: a small business can offer professional customer service without employing a full-time receptionist.

Larger companies can pursue similar efficiencies across hundreds of employees. AI providers and infrastructure businesses can also earn revenue from growing demand.

However, benefits will not be distributed equally. Automation can expand business opportunities while displacing workers or putting pressure on wages, a risk explored in Coinjuice's research on AI, interest rates and Bitcoin.

Cheaper work can expand economic possibilities. Ownership of the resulting income introduces a different consideration.

Bitcoin and 900 Years of Financial Recordkeeping

Long before computers, England's royal treasury faced the challenge of maintaining trustworthy financial records.

From at least the early twelfth century until 1826, the Exchequer used wooden tally sticks to record payments and obligations. Officials carved notches into wood, then split each stick into matching pieces. Comparing the pieces later helped detect alterations.

Historian Tony Moore documented more than 700 years of official use. CoinJuice explored the wider history in What 900 Years of Financial History Says About Bitcoin Lasting.

The Exchequer system still depended on a central treasury.

Bitcoin and AI: Greater Independence for Individuals

More than nine centuries after the earliest English tallies, Satoshi Nakamoto published the Bitcoin white paper in 2008, proposing a different method of recording monetary transfers without a central financial institution.

For individuals, the practical possibilities extend beyond recordkeeping.

• Self-custody: Bitcoin can be held directly without a bank controlling the account.

• Cross-border payments: Value can move internationally through the Bitcoin network without requiring a traditional bank transfer.

• Independent verification: Anyone running compatible full-node software can check transaction history and monetary rules, including Bitcoin's 21-million-coin supply limit.

Bitcoin also introduces challenges. Price volatility, transaction costs, cybersecurity, custody mistakes and legal restrictions can affect real-world use. Independent verification cannot guarantee the accuracy of unrelated information or the future purchasing power of bitcoin.

Still, the economic connection with AI deserves attention.

AI could enable an individual to build and operate a business once requiring considerably more staff and capital. Bitcoin offers the ability to hold and transfer the resulting wealth without depending on a bank as custodian.

The calculator made arithmetic accessible. AI is widening access to productive capabilities. Bitcoin offers a separate form of independence over monetary value.

AI could help more people build successful businesses, while Bitcoin gives them greater control over the money they earn.