MIT Study Projects AI to Automate Most Text Tasks by 2029, Suggesting a 'Rising Tide' of Change

An MIT study suggests that AI's impact on work will be a gradual 'rising tide,' projecting that by 2029, AI will handle most text-based tasks with 80-95% success.

This gradual diffusion is supported by major tech companies embedding AI into workplace software, a steady supply of advanced computing hardware, and a regulatory environment that encourages controlled, incremental deployment.

Consequently, the near-term risk to the labor market is not sudden mass displacement in specific jobs, but a broad, gradual augmentation of tasks across many different roles.

A recent MIT study suggests that artificial intelligence will transform our work lives not with a sudden shock, but like a gradually rising tide.

This study, from MIT's FutureTech initiative, reframes the conversation about AI and jobs. Instead of a "crashing wave" of automation that suddenly displaces entire professions, the researchers found evidence of a slower, broader integration. They project that by 2029, large language models will be able to complete most text-related office tasks with about 80-95% success. This conclusion is based on extensive data, including over 3,000 job tasks and 17,000 worker evaluations, which showed AI's success rate climbing steadily from 50% in mid-2024 to around 65% by late 2025.

So, what's driving this gradual, tide-like change? Several key factors are working together.

First, major technology companies are embedding AI directly into the tools we use every day. Think of Google integrating its Gemini AI into Workspace apps like Docs and Sheets to create first drafts, or Microsoft upgrading Copilot to handle more complex workflows. This approach doesn't replace workers but augments their capabilities, making AI a helpful assistant and encouraging widespread, incremental adoption across countless roles.

Second, the hardware needed to power these advancements is becoming more available. At its GTC 2026 conference, Nvidia highlighted a massive sales opportunity for its next-generation chips like Blackwell and Rubin. Furthermore, reports show that even with export controls, access to powerful GPUs is expanding globally. This growing supply of computational power is the essential fuel that allows AI models to keep improving and scaling.

Finally, the regulatory environment is shaping AI's rollout to be more cautious and controlled. Actions from bodies like the U.S. Federal Trade Commission (FTC), court decisions on AI-generated content copyright, and Europe's AI Act all push companies toward responsible, human-in-the-loop systems. This focus on safety and oversight naturally favors a step-by-step integration over a disruptive, uncontrolled wave. This measured pace is also reflected in economic data, like the steady but not explosive labor productivity growth reported by the BLS, which aligns perfectly with a story of gradual diffusion.

O*NET: A comprehensive database maintained by the U.S. Department of Labor that contains detailed descriptions of hundreds of occupations.
Human-in-the-loop: A system design that requires human interaction, typically for validation, correction, or oversight of AI-generated outputs.
BLS (Bureau of Labor Statistics): The primary U.S. government agency responsible for collecting and analyzing data on labor market activity, working conditions, and price changes.

MIT Study Projects AI to Automate Most Text Tasks by 2029, Suggesting a 'Rising Tide' of Change

An MIT study suggests that AI's impact on work will be a gradual 'rising tide,' projecting that by 2029, AI will handle most text-based tasks with 80-95% success.

Consequently, the near-term risk to the labor market is not sudden mass displacement in specific jobs, but a broad, gradual augmentation of tasks across many different roles.

A recent MIT study suggests that artificial intelligence will transform our work lives not with a sudden shock, but like a gradually rising tide.

So, what's driving this gradual, tide-like change? Several key factors are working together.

O*NET: A comprehensive database maintained by the U.S. Department of Labor that contains detailed descriptions of hundreds of occupations.
Human-in-the-loop: A system design that requires human interaction, typically for validation, correction, or oversight of AI-generated outputs.
BLS (Bureau of Labor Statistics): The primary U.S. government agency responsible for collecting and analyzing data on labor market activity, working conditions, and price changes.