“Workers are no longer quietly experimenting with AI. Companies are now tracking how much employees spend talking to machines.”
Uber has revealed that some of its employees are spending thousands of dollars every month on artificial intelligence coding tools, showing how deeply AI assistants are becoming part of daily work inside major tech companies.
The discussion emerged after Uber’s Chief Operating Officer, Andrew Macdonald, spoke about the company’s growing use of AI tools such as Claude Code during a recent conversation about workplace productivity and software development.
According to Macdonald, several Uber engineers are already using massive amounts of AI “tokens,” the units companies use to measure AI usage and pricing. Some workers reportedly spend several thousand dollars monthly while interacting with advanced AI coding assistants to help write software, solve technical problems, and speed up development tasks. The comments highlight how artificial intelligence is rapidly moving from an experimental tool into a central part of everyday operations inside global companies. AI assistants are quietly becoming digital coworkers inside the modern office.
Macdonald specifically referenced Claude Code, an AI coding assistant developed by Anthropic, one of the leading artificial intelligence firms competing against companies such as OpenAI, Google, and Meta. Claude Code is designed to help programmers write software faster by generating code, debugging errors, explaining technical problems, and assisting with development workflows.
Instead of manually searching through documentation or writing every line themselves, engineers can now hold conversations with AI systems that generate solutions almost instantly. This dramatically reduces development time for many routine tasks. The cost of talking to AI is becoming a real business expense.
The growing usage inside Uber reflects a much larger shift happening across the technology industry. Companies worldwide are increasingly paying for AI subscriptions, usage credits, and token based services as workers integrate tools like ChatGPT, Claude, Gemini, and other AI systems into their daily workflows.
In many cases, businesses now treat AI access the same way they treat cloud software subscriptions or productivity tools. The difference is scale. Some advanced AI systems consume huge amounts of computing power, especially when workers use them continuously for coding, analysis, research, and automation. Macdonald’s comments reveal how quickly those costs can grow once thousands of employees begin relying heavily on AI assistants.
“Every prompt, question, and AI generated response now carries a price tag.” AI “tokens” are small units used to measure how much text an AI system processes. The more workers interact with AI models, generate code, upload files, or request complex outputs, the more tokens are consumed. Companies are then billed based on usage levels.
For businesses with large engineering teams, those expenses can rise rapidly. Still, many executives believe the productivity gains justify the cost. Macdonald suggested that AI coding tools are already helping engineers work faster and solve problems more efficiently, potentially saving companies much more money than they spend on AI subscriptions and usage fees. Software development may never return to the old way of working.
The rise of AI coding assistants is already transforming how programmers operate across the tech industry. Developers increasingly use AI systems to generate starter code, identify bugs, explain unfamiliar programming languages, automate repetitive tasks, and speed up testing processes. Several major companies now encourage engineers to actively use AI during software development rather than avoid it. Some executives even believe coding assistants could eventually become standard tools for nearly every software engineer.
The trend is happening so quickly that many universities and training programs are already adjusting how they teach programming skills. Instead of focusing only on manual coding, educators are beginning to train students on how to work effectively alongside AI systems. The programmer of the future may spend more time guiding AI than writing code manually.
The rapid adoption of AI assistants has also sparked growing debate inside the tech world. Supporters argue that AI tools help remove repetitive work, improve productivity, and allow engineers to focus on higher level problem solving. Critics worry that over reliance on AI generated code could reduce critical thinking skills, introduce hidden security vulnerabilities, or eventually replace large numbers of software jobs.
Those fears have intensified as AI systems continue improving at extraordinary speed. Several technology leaders recently warned that artificial intelligence could dramatically reshape white collar work across industries including software engineering, customer support, finance, marketing, and administration.
The companies building AI are now using AI to build themselves faster. Uber’s comments also reveal how competition among AI companies is intensifying. Anthropic’s Claude models are competing aggressively against OpenAI’s ChatGPT, Google’s Gemini, and other AI systems for dominance inside workplaces. Winning enterprise customers has become one of the most valuable battles in the technology industry. Companies no longer see AI as a side experiment.
Many now consider it critical infrastructure. Businesses are investing billions into AI subscriptions, computing power, data centers, and custom AI integrations as they race to improve efficiency and remain competitive. Artificial intelligence is becoming a permanent expense inside modern companies.
The growing financial commitment also raises bigger questions about the future relationship between humans and AI at work. As employees depend more heavily on AI systems, companies may eventually restructure teams, workflows, and hiring strategies around machine assisted productivity.
Workers who understand how to use AI effectively could become far more valuable than those who resist adapting. At the same time, companies will likely face increasing pressure to balance efficiency with concerns around job displacement, training, and long term workforce stability.
For now, one reality is becoming increasingly clear. Inside some of the world’s biggest companies, employees are no longer casually testing artificial intelligence. They are using it constantly, relying on it heavily, and spending enormous amounts of money communicating with machines that are rapidly becoming part of everyday work itself.
