AI is quickly becoming the next major leap in technological progress. It’s accelerating how people work, create, and solve problems—just as computers and even copy/paste once did. This was part of the central message from Dan Chuparkoff, a leading keynote speaker on innovation and AI, at TMT’s IT Sales and Marketing Boot Camp last month.
Reflecting on his early career, Chuparkoff explained how learning AutoCAD reshaped his entire perspective on technology. “We got PCs, then spreadsheets, then the internet, then mobile phones, then cloud and data science and remote work. Now we have generative AI; just the next stair step in this staircase toward the future.”
Each of these innovations fundamentally altered how people work, and AI is poised to do the same at an even greater scale. “By the end of this year, if you’re not already, you’re going to be using generative AI to help you do work as well. Artificial intelligence is the biggest deal ever in humanity. It’s going to create amazing new things. It’s going to help us solve problems that we never could have before,” he said.
Despite its power, AI is not autonomous. Like any tool, it requires human direction. As Chuparkoff noted, it “needs a person driving it, making the decisions, and solving the weird edge cases that popped up that weren’t expected.”
AI doesn’t think, it predicts
To use AI effectively, it’s critical to understand how it actually works. While many know GPT stands for Generative Pre-trained Transformers, far fewer understand the implications of that design.
GPTs don’t know answers in the human sense; instead, they generate responses based on probability. That probabilistic approach means AI tends towards consensus over originality. “It hears the loudest answer,” Chuparkoff said. “Someone might have said something more creative, more interesting, more unique, but AI [provides] the average answer. It only knows the middle of the bell curve, the thing that most people think.”
Because of this, outputs are not absolute truths; they are just the statistically likely responses. “When you ask AI a question [or] give an AI agent a job to do, there’s probability [driving its output],” Chuparkoff said. “If you don’t have a longer conversation with your AI engine, you won’t understand its confidence.”
This also explains why the so-called “AI hallucinations” people talk about occur. They aren’t a bug that will be patched out, but rather a natural result of how predictive systems function. “There’ll never be a day when we get to a place where AI is 100% certain about almost anything,” said Chuparkoff.
AI isn’t a copilot, it’s an assistant
Given these limitations, AI cannot operate independently of human oversight.
“You should disagree with your AI system a lot, just like you would disagree with [your peers or competitors],” Chuparkoff said. “That’s what drives innovation and collaboration. That’s what drives higher tiers of customer excellence.”
He cautioned against thinking of AI as a peer-level collaborator. It has capabilities, but also clear limits. Consider AI “more like an intern. [It] has a capability ceiling, and you won’t know until you give it some work and review. [Repeat] that loop until you get better at giving your AI intern instructions and understanding the kind of work that they’re not capable of.”
The six kinds of work
To clarify where AI fits into your business, Chuparkoff outlined six types of work, ranging from routine to highly complex:
- Communicate—training, asking questions, determining team needs, and sharing information
- Process—following instructions and completing tasks
- Investigate—identifying issues, improvements, and successes
- Solve—creating solutions for brand-new problems
- Decide—making judgement calls based on human context and values
- Imagine—creating new innovations and directions
“If you think back over the last month, you’re probably spending a lot of your time doing those [first] three,” Chuparkoff said. While necessary, these tasks often consume time that could be better spent on higher-value work.
AI really excels in those first three areas, but struggles with the last three. “Problems can be new and weird and different every single time. [But] AI doesn’t know anything about your new weird problem, because it’s pre-trained,” he explained. “Your job is to come in and solve the new weird problems.”
Likewise, “your decision-making is crucial to you being part of the work equation… You should never delegate any decisions to AI.” And when it comes to innovation, imagination “isn’t in AI’s training data. AI can’t do it.”
Where AI can move the needle
The real value of AI lies in how it redistributes effort; not how it eliminates it. By automating communication, process, and investigation tasks, AI frees up your time for more complex, human-centric tasks.
“The only thing in your life that isn’t growing exponentially is the number of hours in your week. There will never a day where we make our teams so efficient that we don’t have to come in on Friday, because you have an infinite amount of problems and an infinite amount of decisions to make,” said Chuparkoff.
That shift enables teams to operate at a higher level, allowing them to tackle complex challenges, make better decisions, and innovate consistently. “Give those things to AI so that your team can focus on solving harder problems at scale, before they pile up and create bigger ones. If you do that, you’ll be able to spend more time imagining new ways to power the future of MSPs—a better future for yourself, your team, your customers, the whole entire industry. Imagining is what innovation is.”
The next phase of AI in action
This philosophy is already shaping how companies build AI-powered tools. Kaseya, for example, is actively advancing AI beyond assistance and into execution. With the release of Kaseya Intelligence, the company introduces a layer capable of autonomously handling tasks like triaging tickets, managing security threats, and verifying backup recovery without manual intervention.
As AI continues to evolve, its role will become clearer. It is not a replacement for human thinking but a force multiplier for it; handling the predictable so you can focus on what’s possible.
Read more about the industry’s first agentic IT management platform here.
