Global from Day One
Israel offers perhaps the clearest example of how a relatively small country can become a global technology powerhouse by building for the United States, not just its domestic market.
With a population of only around 10 million, Israeli startups could never rely on their home market alone to build generational companies.
Instead, Israeli founders learned to treat the U.S. as an extension of their home market. Raising capital from American venture funds, building sales and leadership teams in the U.S., selling to American enterprises, and ultimately pursuing acquisitions or going public on U.S. public markets.
Companies like Check Point, Mobileye, CyberArk, and Wiz are some of the strongest examples of this playbook.
The same shift happened in venture capital. As more Israeli founders expanded into the U.S., a new generation of Israeli venture funds emerged specifically to back Israeli startups with global ambitions and help them succeed in the American market.
This created a powerful flywheel: more Israeli startups entering the U.S. attracted more cross-border capital and investors, which in turn made it easier for the next generation of Israeli founders to follow.
Israeli talent builds companies → those companies expand into the U.S. → American capital and markets enable them to reach global scale → successful founders and employees bring their capital, experience, and networks back into Israel → the next generation builds even bigger companies.
Companies that once began as startups now form the backbone of Israel’s technology economy, with the sector accounting for roughly 18% of the country’s GDP and 58% of its exports, a testament to how profoundly this model has transformed Israel’s economy.
Korea Needs More Startups to Go Public in the U.S.
More than 250 Israeli companies have gone public on Nasdaq since the 1980s. Korea, despite having more than five times Israel’s population and one of the world’s largest economies, has produced only a fraction of that number.
Korea already has many of the same ingredients Israel started with: world-class engineers, leading universities, deep industrial expertise, and ambitious founders. Arguably, an even stronger starting point.
Korea does not have a startup problem. It has a globalization problem. While Israel built a pipeline from Tel Aviv to Nasdaq, Korea’s best startups have historically been built toward KOSDAQ.
If Korea wants to become the next Israel, that needs to change. The goal should not simply be to create more Korean unicorns. It should be to build globally dominant technology companies and see more of them go public in the United States.
Korea has twice as many unicorns as Japan, yet Japan has sent more than twice as many companies to the U.S. public markets.
The Old Model
The traditional Korean startup funding model has three major problems.
First, pre-seed rounds are too small. A typical Korean startup might raise only $200K–$300K in its first financing round at a $1.5M–$2M valuation, often from a single venture fund. Compared with the U.S., both the round size and valuation can be significantly smaller.
Many Korean founders and investors underestimate how this can be perceived by top U.S. investors. A very small early round at a very low valuation can become a negative signal. It can raise questions about the founding team’s experience, the level of competition for the deal, and whether other credible investors were willing to back the company.
In simple terms: “Who wants to invest in a company nobody else wanted?”
Second, Korean early-stage rounds are often too concentrated. A single venture fund may prefer to take the entire first or second round rather than bring in other investors. In the U.S., strong early-stage rounds often include a network of investors who can each bring something different: future fundraising relationships, customers, talent, industry expertise, and more.
Angel investors are too often treated as cap-table clutter rather than strategic assets. Some Korean VCs discourage founders from bringing angels onto the cap table or even ask founders to remove early individual investors. This misses the point of angel investing. The right angels can become some of a startup’s most valuable early supporters, opening doors to customers, talent, and future investors long before a traditional fund can.
Third, Korean venture funds rarely support their portfolio companies changing their country of incorporation, a process commonly known as a “flip.” When a Korean startup flips to the U.S., the U.S. corporation becomes the parent company and the Korean entity becomes its subsidiary. For existing Korean investors, this can introduce tax, regulatory, and legal complications, particularly around how they hold and eventually exit their investment. As a result, many Korean funds have strong incentives to keep their portfolio companies incorporated in Korea.
I’ve met many great Korean founders who wanted to flip their companies to the U.S., but almost all of them eventually gave up because their existing investors opposed the move.
This creates a fundamental problem because corporate structure can become a major barrier to raising from top-tier U.S. investors. For tax, legal, and administrative reasons, many American venture funds prefer investing in a U.S. corporation, typically a Delaware C-Corp, rather than directly into a Korean entity.
The result is a frustrating cycle: Korean founders need U.S. capital to compete in the U.S., U.S. investors often want them to flip before investing, and existing Korean investors may resist the flip. Promising Korean startups can therefore become trapped in a corporate structure built for the Korean market, even when their customers, investors, and ambitions are global.
What Changed
Things have changed, but not enough.
A new generation of venture funds has emerged to support Korean founders going global. BASS Ventures, Krew Capital, and Sazze Partners are some examples. They tend to back Korean founders with international backgrounds or repeat founders who have built companies outside Korea. Importantly, these firms have U.S.-based fund structures, giving them more flexibility to back founders who incorporate in the U.S. from day one or plan to “flip” later.
At the same time, more Korean startups are joining U.S. accelerator programs such as Y Combinator and a16z Speedrun. YC in particular has almost become a movement in Korea. For many young Korean founders, getting into YC has become one of the clearest symbols that they have “made it” globally.
I think this is wrong.
YC today is very different from the YC of the Paul Graham era. Its batches have grown dramatically, and simply being a YC company no longer carries the scarcity it once did. The program is also heavily concentrated in software and B2B SaaS startups, with far fewer companies in robotics, semiconductors, manufacturing, and hardware.
More importantly, Korean founders should not confuse getting into an American accelerator with succeeding in America. The goal is not to collect the YC badge. The goal is to build a company so exceptional that the best investors, engineers, and customers in the U.S. want to be part of it.
Some will argue that being backed by a top-tier venture fund does not determine whether a startup succeeds.
That is true. Capital alone does not build great companies.
But the best investors can provide far more than capital: they can help recruit exceptional talent, introduce customers, attract future investors, and give founders access to networks that would otherwise take years to build.
This is where I believe Korea still has a fundamental problem. We have spent years asking how Korean founders need to change, but far less time asking how Korean investors need to change.
My thesis is simple: investors have to change first. Founders will follow, because investors shape the incentives, structures, and ambitions of the founders they back.
What Needs to Be Done
1) We need more family offices
Korean institutional LPs are also part of the problem. A significant portion of Korean venture capital is backed by public institutions, most notably the Korea Fund of Funds (KFoF), commonly known as the Motae Fund (모태펀드). Rather than investing directly in startups, these institutions act as LPs in venture funds, providing capital to Korean VCs that then invest in startups.
Public capital often comes with geographic restrictions. Funds backed by these institutions are required to invest their capital into Korean companies, making it difficult to back Korean founders who choose to incorporate in the U.S. from day one. In practice, the system can incentivize Korean VCs to keep Korean founders incorporated in Korea, even when a U.S. structure would better support the company’s global ambitions.
Surprisingly, Korea has fewer than eight family offices that actively invest as LPs in venture funds. Even among them, allocations to Korean venture funds remain limited, with many preferring to invest in established global funds instead.
Korea’s high net-worth individuals still allocate the majority of their wealth to traditional assets such as real estate and public equities, with limited exposure to venture capital and angel investing.
This is why Korean venture funds remain dependent on government-backed LP capital.
If we want Korean VCs to think globally, we need LPs who allow them to invest globally. Korea needs more family offices and private investors willing to back venture funds without the same geographic constraints, giving Korean VCs the flexibility to support companies that are incorporated in Seoul, Delaware, or anywhere else in the world.
2. We need better GPs
Korea has more than 400 venture investment firms, yet only a small number of funds are willing to write meaningful checks into pre-revenue startups at the earliest stage. Simply put, there is plenty of capital (for late-stage), but very little appetite for real risk.
The angel ecosystem is even weaker. There are very few active angel investors, and many venture funds prefer not to have individual investors co-invest alongside them. This is fundamentally different from Silicon Valley, where successful founders, operators, and executives frequently become the earliest backers of the next generation.
Many Korean funds also position themselves as U.S. GTM partners, promising to help founders expand into America. But very few have meaningful relationships with top U.S. venture funds, founders, or talent. Helping a company go global requires more than having an office in Silicon Valley. It requires being part of the ecosystem.
And this points to a deeper problem: Korea has yet to produce a venture firm that has truly earned a place in the U.S. startup ecosystem. A fund that not only invests in Korean founders going to America, but is itself sought after by the best founders in America.
Japan produced SoftBank, founded by Masayoshi Son, which became one of the most influential technology investors in the world.
China produced ZhenFund, led by Anna Fang, an early investor in companies including Manus and Genspark. Sky9 Capital, founded by Ron Cao, which has backed companies like Moonshot (Kimi) and Reactor.
Korea has yet to produce its equivalent.
We need more U.S. venture funds.
Today, there are only 4 Delaware Limited Partnership (LP) structured venture funds investing in both Korean and American founders: Sazze Partners, Krew Capital, Patriot Fund, and BlueBrown Partners.
This matters more than it may seem.
More U.S. venture funds means more Korean founders building U.S. based companies. We need investors who are willing to be the first check, encourage founders to incorporate in the U.S. from day one, and help them build on American soil.
That means introducing their first U.S. customers, employees, co-founders, and follow-on investors, helping them become American companies founded by Korean entrepreneurs. Yet there are remarkably few funds built to do this.
If Korea wants founders to compete globally, its investors need to compete globally first.
3. We need more US venture funds and global startups in Korea
South Korea was the 13th-largest economy in the world in 2025, with a GDP of approximately $1.87 trillion.
Despite having a population of just 52 million, Korea is the world’s #2 market for paying ChatGPT subscribers and #2 for Gemini iOS revenue, while ranking among the top five countries globally in Claude usage.
Koreans are clearly early adopters of global technology. Yet remarkably few global startups treat Korea as a priority market.
More importantly, there is virtually no global venture capital infrastructure dedicated to investing in Korean founders.
Other major Asian startup ecosystems developed alongside global venture firms that built dedicated local franchises.
China had Lightspeed China, MiraclePlus (formerly Y Combinator China), and HongShan (formerly Sequoia Capital China). India had Peak XV Partners (formerly Sequoia Capital India). In Japan, Peter Thiel’s Founders Fund backed Coral Capital, a dedicated Japan-focused venture fund.
Korea never built an equivalent.
And this is not because Korea lacks founders or startups worth investing in.
Dealroom counts 48 Korean unicorns compared with 21 in Japan, while valuing Korea’s startup ecosystem at approximately $431 billion, nearly twice Japan’s $229 billion. Annual venture investment is also comparable, at roughly $8.3 billion in Korea versus $7.1 billion in Japan.
This is remarkable considering Korea has less than half of Japan’s population and an economy less than half its size.
Japan has spent decades building bridges between its technology ecosystem and Silicon Valley. Korea, despite producing an extraordinary amount of technology and startup value relative to its size, remains far behind in building those bridges.
We need more U.S. venture funds built specifically around Korea. And we need more global startups to see Korea not as an afterthought, but as one of the first markets they enter in Asia.
The talent is already here. The startups are already here. The users are already here. What is missing is the bridge to the rest of the world.
4. We need more fellowships
In recent years, fellowships have become one of the most powerful talent engines in venture.
The best fellowships do not wait until someone has already built a company and raised a Seed round. They find exceptional researchers and technical builders before they become obvious founders, surround them with equally ambitious people, and give them access to the networks, capital, and confidence to build something much bigger.
Prod is one of the best examples.
Founded in 2021, Prod is a student-run nonprofit startup fellowship originally centered around ambitious students from Harvard, MIT, and Stanford. Its year-long program helps unusually talented students go from pre-idea builders to full-time founders while taking zero equity.
In just a few years, Prod has become one of the strongest signals for young technical founders in the United States. Its alumni include founders of Cursor, Mercor, and Etched.
Prod produced 8 unicorns in 3 years. Combined $100B+ valuation.
Z Fellows takes an even earlier approach.
Founded by Cory Levy, Z Fellows identifies exceptional young technical builders, often before they have even committed to starting a company, and connects them with some of Silicon Valley’s best founders and investors. Its alumni have gone on to build companies including Cursor, Cognition, Etched, Whop, and Mintlify.
These fellowships demonstrate something important. Great founders can be discovered before great companies exist.
Other countries are beginning to recognize this.
Sweden is building its own version through Velocity Fellows, a nonprofit fellowship that selects up to ten of the country’s most ambitious young builders and flies them to Silicon Valley for a fully funded week with leading founders and investors.
The goal is not simply to teach them how to start companies. It is to expose Sweden’s most talented young people to Silicon Valley’s level of ambition and bring that ambition back home.
Its early fellows include Max Junestrand, co-founder of Legora, and Alfred Wahlforss, co-founder of Listen Labs. Legora has since reached a $5.6 billion valuation, while Listen Labs has reached $1.5 billion.
Korea needs more fellowships like these.
Today, AttentionX is the only fellowship in Korea built specifically to help researchers and technical builders become founders.
Since 2024, more than 40 startups have emerged from the AttentionX community and collectively raised over $70 million from investors including Sequoia, Andreessen Horowitz, South Park Commons, and Y Combinator.
We have not produced a unicorn yet, but the early signals are beginning to appear.
AIM Intelligence recently raised a Series A led by Samsung Ventures, with participation from Hyundai, NAVER, and LG. Typa and Nari Labs was accepted into Y Combinator. Dentronic joined South Park Commons, while Philyron was accepted into a16z Speedrun.
Korea has some of the best engineers, researchers, and students in the world. What we lack are enough institutions dedicated to finding these people early, surrounding them with equally ambitious peers, connecting them directly to Silicon Valley, and convincing them that they should build for the world.
5. We need more frontier labs
Korea needs a stronger support system for researchers and technical founders, and we need to direct more of our attention and capital toward deep technology rather than industries we have already mastered. Korea is already a global powerhouse in entertainment. Now is the time to catch up with the United States and China in the AI race.
People often ask why South Korea has not produced a single globally recognized frontier AI lab.
I have wondered the same thing for years. It is frustrating, and frankly embarrassing, because Korea has some of the best technical talent in the world. There is no fundamental reason why Korea should not be capable of producing an AI lab that competes with the best labs in the United States and China.
In fact, a few Korean companies are already trying.
Upstage develops Solar, its proprietary family of foundation models, alongside enterprise document AI products. Motif Technologies builds Korean foundation models across language and image generation, with a focus on efficient training and inference. Trillion Labs trains foundation models from scratch, with an increasing focus on the compute, energy, and infrastructure required to power AI at scale.
However, none of them are close to competing with the leading American and Chinese frontier labs in model performance, capital, or global talent density.
The problem is not simply technical capability. It is an ecosystem problem.
These companies are headquartered in Seoul and operate with a fraction of the capital available to their American competitors. That makes competing for the world’s best researchers extraordinarily difficult. Upstage has publicly discussed offering compensation packages worth as much as $10 million to attract exceptional global AI talent. But even that illustrates the scale of the challenge. The best researchers are being recruited by OpenAI, Anthropic, Meta, Google DeepMind, and other frontier labs with enormous amounts of capital, compute, equity, and research infrastructure behind them.
This is where the “Korea discount” becomes more than a stock market problem.
Korean companies have historically received lower valuations than comparable companies in the United States. For an AI startup, valuation is not just a number on a term sheet. Lower valuations mean less capital. Less capital means less compute, fewer researchers, smaller compensation packages, and ultimately a weaker ability to compete at the frontier.
There is another problem that capital alone cannot solve: Korea’s lack of connectivity to the U.S. AI ecosystem. There are very few people who can genuinely bridge a frontier lab in Seoul with researchers in San Francisco, investors in Silicon Valley, and the global technical community.
China has shown that geography does not make competing with Silicon Valley impossible. DeepSeek and Moonshot AI proved that globally competitive frontier models can emerge outside the United States. What matters is concentrating enough talent, capital, compute, and ambition behind the right teams.
Korea does not need hundreds of frontier labs. We need one or two that are truly world-class.
If Korea can produce just one or two frontier labs capable of competing at the level of DeepSeek or Moonshot AI, the impact would extend far beyond those companies. They would attract global researchers to Korea, create generations of new technical founders, pull international capital into the ecosystem, and prove that the world’s most important AI companies do not have to be built only in San Francisco or Beijing.
One globally competitive frontier lab could change how the world sees Korean technology.