Vercel's CTO Malte Ubl suggests a simple method for finding valuable internal automation tasks: ask people, "What do you hate most about your job?" This uncovers tedious work that requires some human judgment, making it a perfect sweet spot for the capabilities of current-generation AI agents.
Related Insights
Integrate AI agents directly into core workflows like Slack and institutionalize them as the "first line of response." By tagging the agent on every new bug, crash, or request, it provides an initial analysis or pull request that humans can then review, edit, or build upon.
Instead of replacing top performers, AI should be used to do work humans physically cannot. Salesforce targeted a backlog of 100 million 'orphan leads,' using an AI agent to work through 8,000 dormant leads in three weeks. This generated $500,000 in pipeline that would have otherwise been zero.
To discover high-value AI use cases, reframe the problem. Instead of thinking about features, ask, "If my user had a human assistant for this workflow, what tasks would they delegate?" This simple question uncovers powerful opportunities where agents can perform valuable jobs, shifting focus from technology to user value.
Cognition's CEO highlights five ideal, immediately delegable tasks for AI coding assistants: miscellaneous front-end fixes, version upgrades and migrations, documentation generation, initial incident response, and writing unit tests for existing code.
The next frontier for AI in product is automating time-consuming but cognitively simple tasks. An AI agent can connect CRM data, customer feedback, and product specs to instantly generate a qualified list of beta testers, compressing a multi-week process into days.
When employees are 'too busy' to learn AI, don't just schedule more training. Instead, identify their most time-consuming task and build a specific AI tool (like a custom GPT) to solve it. This proves AI's value by giving them back time, creating the bandwidth and motivation needed for deeper learning.
Instead of focusing on foundational models, software engineers should target the creation of AI "agents." These are automated workflows designed to handle specific, repetitive business chores within departments like customer support, sales, or HR. This is where companies see immediate value and are willing to invest.
To win over skeptical team members, high-level mandates are ineffective. Instead, demonstrate AI's value by building a tool that solves a personal, tedious part of their job, such as automating a weekly report they despise. This tangible, personal benefit is the fastest path to adoption.
A killer app for AI in IT is automating tedious but critical tasks. For example, investigating why daily cloud spend deviates by more than 5%. This simple-sounding query requires complex data analysis across multiple services—a perfect, high-value problem for an AI agent to solve.
A tangible way to implement a "more human" AI strategy is to use automation to free up employee time from repetitive tasks. This saved time should then be deliberately reallocated to high-value, human-centric activities, such as providing personalized customer consultations, that technology cannot replicate.
