Two startup chiefs at the center of the AI boom insist the prevailing narrative is off: artificial intelligence is remapping work, not erasing it. Leaders from Read AI and Lucidya argue the technology is best at offloading repetitive tasks while elevating distinctly human strengths such as judgment, relationship-building, and creative problem-solving. Their own operations—and a growing body of research—lend weight to the claim.
AI Will Automate Tasks, Not Entire Job Roles
Lucidya founder Abdullah Asiri describes what happens when customer support teams adopt his company’s tools: job descriptions shift rather than vanish. Agents who once spent hours triaging routine tickets move into supervisory roles, guide both colleagues and AI systems, and take on higher-value outreach and retention work. Freed from low-level toil, they advance the conversation with customers instead of transcribing it.
Read AI’s David Shim sees a similar pattern with meeting assistants. Automated note-taking and action-item extraction eliminate drudgery and improve recall, but they still leave decisions—and accountability—squarely with people. In his view, AI increasingly functions like GPS for knowledge work: it recommends the route, yet a human operator steers, pivots, and bears responsibility for the destination.
Humans In The Loop By Design, Not An Afterthought
The founders are blunt that some tasks will disappear, particularly in functions like ad operations or rote back-office workflows. But they emphasize this triggers a countervailing need for oversight roles to set guardrails, review edge cases, and align models with business outcomes. That human-in-the-loop design is not just ethical window dressing; it is operational risk management for systems that can be powerful and wrong in confident ways.
This split between automation and orchestration echoes what major consultancies report. McKinsey estimates generative AI could automate activities that account for 60%–70% of employees’ time in some occupations, while full role replacement remains rare. In other words, the task mix changes faster than job titles do.
Lean Teams Offer a Live Case Study in AI at Scale
Inside Read AI, a five-person customer support team services millions of monthly users—an efficiency that would be difficult without AI triage, summarization, and context retrieval. The company says its sales intelligence tool, which taps data from systems like HubSpot and Salesforce, has supported approvals on approximately $200 million in deals. Internally, Read AI reports capturing 23% more contextual detail with each update, giving reps clearer feedback loops on what moves a lead forward.
Lucidya applies the same playbook—using AI for meeting notes, content generation, and support analysis—not to slash headcount, Asiri says, but to scale outcomes without scaling payroll. The hiring mandate shifts accordingly: recruit “AI-native” operators who can compose prompts, chain tools, and build lightweight agents to accelerate their own work.
Customers Care About Resolution Speed and Accuracy
Both founders stress transparency but note that, for end users, results generally matter more than the means. Lucidya discloses when a voice bot engages and focuses on fast, accurate resolution. Read AI has found meeting participants increasingly comfortable with automated note-takers when given clear controls over recording and sharing.
Independent evidence backs this pragmatism. A widely cited study by researchers from Stanford and MIT on a large contact center found AI assistance improved agent productivity by 14% overall, with the biggest gains for less-experienced workers. Faster resolutions and more consistent answers translated to better customer outcomes without eliminating human agents.
Data Points Favor Augmentation Over Substitution
The view from these startups mirrors broader labor-market findings. The OECD has reported that workers in AI-exposed jobs are more likely to receive training and see role evolution rather than immediate displacement. Meanwhile, generative models still struggle with ambiguity, sparse data, and the social nuance embedded in sales, service, and partnership work—areas where humans interpret context, negotiate trade-offs, and build trust.
In practice, firms that adopt AI effectively re-bundle roles: automation handles high-volume pattern matching, while people tackle exceptions, empathy, and strategy. That division of labor is clearest in support and sales but is spreading across finance, operations, and product development.
The Skills CEOs Want Next for AI-Oriented Teams
Asiri and Shim expect hiring to reward employees who can orchestrate AI—building lightweight agents, integrating data sources, and turning model output into decisions. Companies will need AI operations leads, QA reviewers for model outputs, and domain experts who translate messy business goals into machine-readable tasks. The near-term risk is transitional: organizations that fail to retrain or redesign workflows may cut jobs rather than re-skill them.
The takeaway from these founders is less techno-optimism than operating doctrine: start by automating tasks, keep humans in control, redeploy capacity toward customer impact, and measure everything. AI is changing how work gets done at speed—but for those building with it, the future looks augmented, not automated away.
