Image credit: Pexels

From rapid experimentation to built-in safeguards, four entrepreneurs and consultants explain how companies are finding practical ways to harness AI without losing human oversight.

As debate over artificial intelligence (AI) intensifies, with industry leaders calling for slower development and open-source models advancing rapidly, businesses are finding ways to move forward. Rather than waiting for consensus on the technology’s future, entrepreneurs and consultants are focusing on practical adoption, responsible implementation, and measurable results. Their experiences suggest that progress does not necessarily require choosing between unrestricted acceleration and restraint. Instead, companies are learning to adapt while keeping human judgment central to their operations.

Building Fast in an Uncertain AI Landscape

For Nirajan Bom Malla, founder of Vedicmarga, AI’s development represents a turning point comparable to the invention of written language. He sees large language models as a new way to transfer and accumulate knowledge, with implications for how businesses and societies operate.

Malla shares:  “What I feel was back a few thousand years ago, when humans first introduced language, written script, things, development, everything started moving so fast, because before that, there were no written scripts, and the language was only verbal. So similarly, after this, language models are kind of scripts for the machines.”

He acknowledges concerns about unchecked development, including the possibility of AI being used to create bioweapons. 

“The way AI is developing, I think it’s not that easy to stop all because we have open web models, open source models, right? All those are also equally progressing at the speed of how this all leader model are doing.”

For businesses, Malla’s response is experimentation, not hesitation. Drawing on 18 years of content design experience, he believes human expertise remains an important differentiator.

“The strategy I’m applying for my product or for my business is building fast and testing fast, because I don’t know in the six months or one year how the market will shift.”

When Automation Creates More Work

Dan Kabakov, founder of Online Labs and author of Fixed and Scale, initially approached AI with skepticism, viewing early tools as another wave of Silicon Valley hype. Subsequent improvements changed his assessment, prompting him to incorporate automation into his business.

However, adoption brought an unexpected challenge. Despite operating more than 30 automations intended to free up his time, Kabakov found himself working longer hours. Managing, reviewing, and correcting AI-generated outputs created another layer of responsibility.

His experience highlights a productivity paradox facing businesses: automation can reduce the time spent on execution while increasing the work required to supervise its results. Rather than eliminating human involvement, AI is changing where that involvement matters most. Kabakov sees strategic and creative thinking as increasingly important responsibilities, particularly when machines handle routine processes.

Building Governance Before Problems Emerge

 Talia McCarthy, founder of Proactive AI Partners, approaches adoption through the combined perspective of enterprise experience and small-business consulting. After five years working on enterprise AI at EY, she shifted her focus towards small and medium-sized businesses (SMBs).

She describes an industry operating at two different speeds. While large organizations face complex governance challenges, many smaller businesses are still moving from paper-based records to digital processes.

“I’m moving them from paper receipts to digital processes. Like, yeah, these small, medium-sized businesses are many years behind on needing that governance yet. But there are a lot of companies that I think need those guardrails in place, especially the larger tech companies.”

McCarthy sees this gap as an opportunity to establish safeguards before adoption accelerates.

“At the very beginning of the engagement, we set guardrails, safety parameters. We set all of the metrics up for that reason and the governance of who owns what process that we’re going to automate, who double-checks it, what’s the kill switch? How do we have a back-out plan? We account for all that before we even implement.”

She believes smaller businesses have a potential one-to-two-year window to establish appropriate oversight. Larger technology companies, meanwhile, need stronger guardrails as their systems become more capable.

McCarthy’s position on slowing AI development is equally pragmatic. 

“Technologically, you can do cold calls now with the AI agent… but they don’t get a response. So those are no return on investment… there’s no reward in bypassing the safety guardrails that are built, which is great. I think that’s why it doesn’t make me so nervous that we haven’t slowed down.”

Treating AI as a Business Tool

For Quinnlan Varcoe, an AI digital forensics examiner and founder of SleuthX, much of the debate surrounding AI overlooks a basic distinction: the technology is a tool, not a person.

“AI is very fascinating, and it’s great to work with, and it’s a great tool, but it is a tool. It’s not a person; it’s not capable of intelligent thought, and it’s just a prediction of what you tell it and what Anthropic or other AI companies have programmed it to do.”

Varcoe also questions the intensity of public calls for a slowdown, arguing that some industry claims exaggerate the pace and implications of technological change.

“In my opinion, a lot of claims made by AI executives or the main people surrounding AI are trying to hype up what’s going on with it. The claim about slowing down AI progress is mostly being exaggerated right now to paint the picture that AI is going to change the world really quickly.”

His own move into AI-powered digital forensics reflects a practical business strategy: identify industries with limited innovation and use automation to improve efficiency, reduce costs, and accelerate processes.

“I think that the way that AI is being used for automation is very important, and I think day by day it should be followed because I think a lot of companies today, if they don’t use AI effectively, then they waste money and fall behind. And I think the number one advantage of any startup today is the ability to follow tech innovation and stay up to date with it to automate business processes that waste time and money.”

Varcoe sees the ability to follow technological developments and automate time-consuming operations as a significant advantage for startups.

Finding a Middle Ground

These four businesses are neither embracing AI without scrutiny nor waiting for development to slow. They are experimenting, introducing safeguards and reassessing how human expertise fits into increasingly automated operations.

The competitive advantage may ultimately belong to organizations that can move quickly without abandoning accountability. In a technology landscape that continues to evolve, practical experience, sound judgment, and responsible implementation are becoming as important as the tools themselves.