Dr. Gleb Tsipursky – The Office Whisperer
Nirit Cohen – WorkFutures
Angela Howard – Culture Expert
Drew Jones – Design & Innovation
Jonathan Price – CRE & Flex Expert
A new Stanford University study shows a striking misalignment between what AI startups are building and what workers actually need. After analyzing 1,500 employees across more than 100 job types, researchers found that nearly 41% of automation tools created by Y Combinator-backed startups fall into categories workers rank as low priority or even off-limits for automation.
This growing disconnect underscores a costly issue: while tech founders and venture capitalists pour billions into AI aimed at strategic and creative tasks, the employees who stand to use these tools say they’d rather get relief from repetitive work.
From data entry to IT troubleshooting, it’s the mundane “dirty work” that workers most want automated, not their creative or decision-making roles.
The Real Automation Wish List
The Stanford team, using U.S. Department of Labor data and a newly developed “Human Agency Scale,” found employees are overwhelmingly looking for AI to tackle routine administrative burdens, not replace meaningful or high-touch work.
Despite this, many startups continue focusing on automating high-level roles in strategy, marketing, or management, which are jobs employees often prefer to keep human-led. The result is a tech industry chasing glossy, headline-friendly AI applications while ignoring the everyday pain points that could genuinely boost productivity and morale.
Venture Capital’s Misguided Optimism
The surge of capital into AI, especially post-ChatGPT, has only fueled this misdirection. In the first quarter of 2025 alone, nearly 58% of all venture capital funding went to AI and machine learning ventures — around $73.1 billion.
But according to multiple reports, these investments frequently lack alignment with real-world needs. Some 42% of businesses report shelving AI projects due to poor fit or impact, and more than 90% of AI startups are expected to fail within five years.
Industry insiders blame this on a rush to deliver breakthrough innovations instead of practical tools. Investors are increasingly drawn to projects that promise radical transformation, rather than technologies that quietly, but effectively, improve daily workflows.
Workers Want a Say in the AI Equation
Perhaps the most telling takeaway from the Stanford research is how workers want AI to function: not as a replacement, but as a collaborator. In nearly half the occupations studied, employees preferred a balanced “H3” model — equal partnership between humans and AI.
Even when full automation is technically possible, workers often want to stay involved.
As Google CEO Sundar Pichai has said, the future of AI lies in augmentation, not substitution. In fact, organizations that design tools to enhance human input — rather than bypass it — stand to gain both in adoption and long-term success, according to Forbes.
The Bottom Line
The AI startup space is booming, but its direction may be flawed. By focusing too much on flashy applications and too little on workers’ practical needs, tech founders risk pouring resources into solutions that don’t solve real problems.
If the goal is better productivity and widespread adoption, startups may need to turn their attention to the tasks employees actually want help with, and embrace AI as a teammate.
