Hello, this is Real World!
In the AI era, people who accomplish bigger tasks with small teams
Last Thursday evening, entrepreneurs gathered at Merih Here in Seongsu-dong, where Real World Seongsu is located. It was the EMA Networking Day hosted by MYSC.
From manufacturing, biotech, content, food, to climate tech. CEOs from different industries came together, each carrying their own concerns.
"How should entrepreneurs work in the AI era?"
This day was organized with keynote speeches and networking sessions. Now, AI is not just affecting specific technologies or certain industries, but is actually changing the way startups organize themselves, build products, meet customers, and make decisions. So there were natural conversations about how to integrate AI into their respective industries' businesses and organizations, and how investors are viewing these changes.
Go crazy!
MYSC CEO Kim Jung-tae's words: "Go crazy"
At first hearing, it might sound a bit harsh. But within those words was a very realistic sense of how to view this current era.
CEO Kim Jung-tae said that within the next three years, many of the current VCs might face difficulties. The reason was actually because of AI. It may no longer be an era where you can only accomplish big things by receiving large investments, hiring many people, and taking a long time to scale, as it was before.
Thanks to AI, it's now possible to achieve far greater results with much fewer resources than before. This means that for entrepreneurs, the formula "you must receive a large investment" will inevitably be shaken. And from an investor's perspective, the existing methods alone may no longer be sufficient.
So his point was that ordinary methods won't work anymore.
It meant that it's difficult to keep up with the current changes with the same speed, the same methods, and the same thinking as before.
That day's "go crazy" didn't sound like a simple slogan.
It felt like quite urgent advice from someone who has long watched the startup ecosystem, now conveying it in front of this change.
Who has a higher success probability—AI reviewers or human reviewers?
CEO Kim Jung-tae's story naturally led to the issue of investment decisions.
Is investment truly intuition or statistics?
And to what extent can AI replace that judgment?
MYSC has accumulated data from investing in over 400 companies over the past 15 years. CEO Kim Jung-tae shared how MYSC has been viewing companies throughout that process, and how they're currently experimenting with how human reviewers and AI reviewers can work together.
What was particularly interesting was the story of the 'A Folder' and 'C Folder' that MYSC has been using internally.
The A Folder consists of the most convincing teams at the current point in time. The business structure is clear, the numbers are organized, and the team's completeness is high. In a word, they're teams that look good to invest in right now. On the other hand, the C Folder contains teams that haven't been organized yet. They often lack explanation, their market hypotheses haven't been fully validated, or it's difficult to instill immediate confidence. Most investors naturally focus more on the A Folder to reduce risk.
But when MYSC looked back at their past investment data, the results were slightly different from intuition.
A significant number of the companies that achieved the highest results were initially in the C Folder.
This part was very striking.
Investment is not only about choosing the most complete team right now, but also about reading how far a team that hasn't been organized yet can move.
So what about AI? Will it make investment decisions similar to human reviewers? Or not?
In time, with unrealized potential still ahead, how will the future of companies, as seen by humans and AI, ultimately be determined?
AI Reviewer 'Merry'—adding questions rather than replacing judgments
Within this consideration, MYSC reportedly designed their own AI reviewer called 'Merry'.
But the starting point was not "AI replaces people." The principle was clear.
Humans make the judgments, and AI organizes.
AI doesn't automatically reject companies or make final decisions. Instead, it helps structure and compare financial information, market data, team composition, similar cases, and so on. It functions as an auxiliary device that shows patterns a human reviewer might miss in a busy schedule.
In fact, MYSC compared the evaluations of human reviewers and AI across approximately 400 startups. But the agreement rate between the two judgments was only 30-40%. At first, this was a result that could make one question whether the system was flawed.
But after repeatedly examining it, the conclusion was different.
The fact that they don't agree is precisely what makes it meaningful.
If AI and human judgments were completely the same, there might not be much reason to have AI in the first place. What matters is the point where different judgments emerge.
Why did the human reviewer rate this team highly?
Why did AI view the same team poorly?
What risk did AI see, and what possibility did the human perceive?
In the process of interpreting that gap, the depth of review becomes greater. Rather than being something that provides answers, AI becomes a questioning mechanism that makes human judgment deeper.
Of course, there were clear limitations.
While AI shows strength in quantitative data, it still has limitations in areas that aren't easily quantified by numbers, like social impact, the entrepreneur's narrative, or the depth of vision. In particular, he noted that we should be cautious about potential biases when evaluating female entrepreneurs, companies targeting vulnerable populations, and vision-focused companies.
Ultimately, final responsibility still rests with people.
The point was that AI should not be the subject of judgment, but rather should play a role in asking questions and creating structures so that people can make better judgments.
However, something different was happening inside.
An increase in people didn't necessarily lead to increases in productivity and revenue. Rather, the larger the organization grew, the slower decision-making became, and the pace of work became heavier. Scale was supposed to be a strength, but at some point, that scale began to become a burden.
Then, a project that the company had prepared with all its might stalled. Business conditions deteriorated rapidly, and the company had to go through a time where it had to reduce the 45-person organization down to 8 people.
Any entrepreneur would feel their heart grow heavy just imagining such a moment.
But strangely, the turnaround began right there.
The remaining 8 people began to embrace AI very actively. It wasn't just about using it as a supplementary work tool. They redesigned the way they work altogether. They deeply integrated AI throughout the process of meeting, planning, creating, executing, and receiving feedback.
And then something remarkable happened.
Tasks that had been difficult even with 45 people, 8 people began to accomplish again. Or rather, in some respects, they began to accomplish even more.
The company, by previous standards, was no longer a small-scale organization with proportionally small power.
He vividly told the story of how, by completely reworking it with a new way of working with AI, the company remarkably took flight.
All attendees conveyed that it was truly a time of deep empathy and inspiration.
While it had unavoidably become smaller due to the crisis, that shrinking actually became the catalyst for the fastest transformation.
AI Transformation Is a Leader's Job! Never Delegate It!
There was a part that CEO Song In-hyuk particularly emphasized.
The point was that now, leaders must be the best at handling AI.
When it comes to implementing AI, it's easy to imagine practitioners finding and using the tools they each need. Of course, doing only that can boost individual productivity to some degree. But if that experience isn't accumulated within the organization, it's difficult for the company to change significantly.
It connected with what CEO Kim Jung-tae said.
The reason AI reviewer 'Merry' was meaningful wasn't because AI made judgments alone, but because it was designed together with the judgment structure of human reviewers. Similarly, within startups, AI shouldn't remain just a tool that individuals use separately. It should be connected in a way that builds organizational knowledge and judgment together.
There was another way to be careful about.
It's the case where a leader doesn't set the direction but dumps the transformation itself on members, saying "Try changing it with AI once."
But where the company needs to go, what customers it needs to meet, and what problems it needs to solve first—these are ultimately things the leader must understand most deeply. If such a leader steps back from AI transformation, the changes will inevitably scatter.
So AI transformation wasn't simply about implementing tools.
It was a process where leaders first used it deeply, integrated it into the organization's way of working, and built that experience together with members.
The reason UniqueGood Company can now accomplish the work of over 45 people with just 8 people lies here.
Rather than each person using AI separately, they made it a way for the entire organization to work together.
Again, "Go Crazy"
On the way back after the event, that thought occurred to me.
The AI era is opening up more possibilities for small teams.
But it seems that's not just a story of efficiency—"doing more with fewer people."
Rather, it's more like an era where entrepreneurs can now get closer to the work they really wanted to do, the changes they wanted to create, and the missions they wanted to hold onto till the end.
There were things in the past that were put off because there weren't enough people, not enough money, or not enough time. But now the tools have changed. The way we work is changing. How investors view companies is also changing.
So the remaining question is ultimately just one.
What do we want to accomplish?
AI doesn't completely replace people.
But it shows much more clearly how people work, how they make judgments, and where they waver.
The same applies to the field of investment. Whether AI replaces reviewers might not be the important question. What matters more is how we can better handle uncertainty when both human intuition and AI analysis are present together.
Entrepreneurship is ultimately similar.
It's not about starting because you're certain, but about creating the next scene even amid uncertainty. And AI is becoming a tool that allows us to imagine that next scene faster, experiment more modestly, and push further ahead.
As our scale has shrunk, so have our excuses.
As our tools have grown stronger, so have our alibis.
Now it's time to choose more clearly, try faster, and immerse ourselves more deeply.
That's why the one phrase CEO Kim Jung-tae threw that day stayed with me.
Go crazy.
For entrepreneurs passing through the AI era, and for investors watching those entrepreneurs, perhaps there is no more accurate word than this!