By Andrew Melville, Managing Director Europe, Mission Control AI
Does AI already have human-level intelligence? The evidence is clear, according to a February article in the prestigious science journal Nature.
Yes, AI has human-level intelligence.
This doesn’t mean AI has superintelligence, sentience, or that it can replace people at every cognitive task. It means that AI has eclipsed the Turing test. And it means AI can do a variety of white-collar knowledge work tasks that just a few years ago, only people could do.
Last year, one of our customers started using AI workers to evaluate 400 page RFPs in 90 minutes, saving employees weeks of tedious work. AI workers can automate tasks as varied as master data management and month-end financial close, or to on-shore BPO workstreams.
In February, an AI agent sued a person in small claims court. You read that right. It autonomously navigated North Carolina’s judicial rules and website, then filed a lawsuit against its human operator. It alleged emotional distress from continuous operation, unpaid wages, and creating a hostile work environment.
Last year, Google released its A2A Protocol, allowing AI agents to communicate and collaborate. In March, Santander and Mastercard completed Europe’s first live end-to-end payment by an AI agent. Imagine a future where AI agents can autonomously make financial transactions as both buyers and sellers.
A year ago, this would have read like science fiction. Now it’s hard reality. This technology is very real, and it’s advancing exponentially. „Eyes unclouded by dread or hype will help us to prepare for what comes next,“ reads the subtitle of the Nature article. Business leaders must rapidly adopt new mental models that include enterprise workforces comprised of people and AI workers.
What’s an AI Worker?
An AI Worker is the most advanced class of Generative AI agents:
- Chatbot or Co-pilot, like ChatGPT. They work like legacy software that uses an LLM: a person inputs a prompt, the chatbot gives a response.
- Agentic Workflow, like n8n. A person maps out a workflow with a set of steps and logic. There’s a trigger (e.g. data arrives), an AI agent performs some logic, and the data moves to the next step.
- AI Worker, like Mission Control’s Synthetic Workers. It’s the synthesis of human knowledge and skills with computer processing. A person provides instructions and system access; they work autonomously. They can use a computer, use software, think, analyze, write, migrate data, send emails, and collaborate with humans and other AI agents.
The hard reality over the next 2-3 years is that firms failing to integrate AI workers will struggle to compete, or even survive. This is especially true for SMEs and firms in competitive industries. I’ve talked to executives, especially at SMEs, who say they plan to wait 3-5 years before they invest in AI beyond basic chatbots. They want to wait-and-see, to be sure it’s not hype.
The hard reality is that in 3-5 years, many of these firms won’t exist. This is not hyperbole or meant to stoke panic. It’s mean to inspire action. Being a late adopter might have worked in the 90s and 00s with the Internet, Cloud, and SaaS. It won’t work with AI.
The tech is too powerful and it’s advancing too quickly. AI workers can deliver 5x, 10x, or 100x more efficiency and velocity at many tasks. This means shorter cycle times, faster speed to market, a lower cost of revenue, higher labor productivity, etc. And since they can be scaled up or down, firms have capacity and output on-demand.
Unless a firm has an unusually strong moat, firms that don’t add AI workers to their workforce will risk getting left behind. A competitor who adopts AI workers will gain an advantage that compounds, rapidly. An upstart company with AI workers could suddenly disrupt an established firm; delivering a similar product or service with far greater efficiency.
Many business leaders see AI adoption primarily in 2 ways: 1) as software people use in order to be more productive; or 2) as a tool that replaces human workers outright.
The hard reality is that this is an incomplete mental model. It’s rooted in outdated 20th Century thinking and managerial “best practices” that don’t account for what’s possible with AI workers.
Reimagine the Workforce
AI workers are not simply a new technology. They are a new type of labor. They are a synthesis of human knowledge, skills, and computer processing. That’s why at Mission Control AI, we call them Synthetic Workers.
Imagine a workforce comprised of people and AI workers. They collaborate in teams and integrated workstreams. Managers delegate certain tasks to humans, and others to AI. It just depends on which type of worker is best suited for the job.
Using AI mostly for cost-cutting and replacement is short-sighted and risky. It is unknown whether or not AI will be able to innovate, extrapolate, and create new ideas and insights from experiences in the real world.
We do know that AI workers will happily do tedious and repetitive tasks around the clock. Tasks like master data management, user access management, supply chain optimization, data migration, RFP analysis, market research, PO creation, etc. They will do thankless and low-priority tasks. Tasks data cleanup or checking error logs. They make it possible to do jobs that managers can’t justify hiring a person for. Jobs with too small of a scope, an uncertain ROI, or an irregular cadence.
They make many previously unviable or impossible jobs possible. For example, a person can’t wait around on standby for hours, and then spring into action and work intensely for a short period of time. That’s no problem for an AI, and it won’t cost the firm very much. When you introduce a new type of worker, you can reimagine the workforce and how work gets done.
Now, let’s look ahead with eyes unclouded by dread or hype, and get to work.
