His career, which spans his time as founder of , DFJ and thehas also included high-profile missteps 鈥� most notably 鈥� underscoring the risk and volatility that goes along with making bold wagers.

A frequent personality on TV and social media, Draper is also known as a relentless champion for decentralized technology and a leading voice for bitcoin and blockchain. In 2024, he launched Draper TV, a media network, where he continues to host a global pitch competition called 鈥淢eet the Drapers.鈥� The series, which is now entering its ninth season, invites viewers at home to invest alongside him in innovative startups.

Draper exudes an almost schoolboy-like enthusiasm and passion when it comes to startups, technology, bitcoin and innovation. I recently spoke with him 鈥� while he was sporting his favorite purple and gold bitcoin tie 鈥� to get his thoughts on everything from his use of digital twins, how the current AI boom compares to previous cycles, and how he wishes policymakers approached tech regulation.

This interview has been edited for clarity and brevity.

附近上门 News: What have you been up to lately? What’s occupying your time?

Draper: We are doing something interesting with 鈥� we鈥檙e joining them for something called America’s Startup. We’re going to do a business plan competition around the country for college students. It kind of dovetails into 鈥淢eet the Drapers.鈥� is one of our sponsors, so we鈥檙e thinking about doing shows in 鈥渟mall bites鈥� for them.

We鈥檙e also doing a lot with . This is the year we turn our distribution global. We had a reach of 300 million people, with 10 million seeing each episode, but we鈥檙e focusing on building the YouTube audience now because you get more control and understand the audience better.

Then there is . We鈥檙e building relationships with various countries that send their top students or potential entrepreneurs to us. People call it a 鈥減re-accelerator,鈥� but I call it a 鈥渉uman accelerator.鈥� We accelerate the people 鈥斅爐hey have to accelerate their own business. We take them through very difficult challenges: a three-day hackathon and survival training with the Navy SEALs, special forces and the . Then they have a two-minute presentation to VCs.

You鈥檙e using “digital twins.” How are you actually deploying AI in your daily operations?

Yes, they are helping. They answer questions from entrepreneurs. On our site, they can talk with me or my digital twin, or they can send in a deck.

My team has built these in a few different ways. One is a hologram by Proto at Draper University. On our website, we have a twin created by Randy Adams that can talk to entrepreneurs. We even have an AI 鈥� built by an intern 鈥� that evaluates pitch decks and 鈥渟pits out鈥� feedback.

Beyond that, we use a tool called Seer that uses video to detect facial expressions; it can determine if an entrepreneur is passionate, lying or genuinely interesting. We鈥檙e also using a voice analysis tool 鈥� similar to how reportedly hires people based on specific 鈥渧oice models鈥� that match their desired personality types 鈥� to identify the 鈥渆ntrepreneurial voice.鈥�

What do you think feels fundamentally different about the cycle that we鈥檙e in right now compared to previous ones?

Weirdly, I don鈥檛 see a big difference. It鈥檚 as big as the dot-com boom, maybe bigger. I call it the Draper iS curve. Every industry goes through this. There is a little “i” 鈥� that鈥檚 the hype. It comes to a point (the dot on the i), and then it comes down because people are disenchanted. It sits there while engineers are hard at work, and then it grows into a big “S” that goes way bigger than the top of the i.

It happened with the internet: 1999 was the climb, 2000 was the top, and 2001 was the crash. From 2001 to 2008, it grew into a huge boom. It’s happening with bitcoin now. And AI is right at the “dot” on the i or coming down off it. People are disenchanted because of energy issues, but it will eventually be bigger than anyone imagined, especially in robotics.

What鈥檚 the trend that you think right now might be a little bit overhyped? And what鈥檚 something that’s underestimated?

The quick answer is AI is overhyped, but I don’t believe that. Under-noticed is that Big Pharma would have you believe chemotherapies are the most important thing 鈥� that you create a molecule and use it forever, and then need another molecule for the side effects. We鈥檙e moving from chemotherapies to bio-cures: stem cells, cloning and genetic engineering.

Also, companies we used to call 鈥渟pace and transportation鈥� are now called dual-use. The and governments are buying in because they realize they are way behind the commercial sector. And bitcoin is in that period where 鈥渘obody cares,鈥� but it鈥檚 slowly taking over.

Do you see bitcoin actually replacing the dollar for daily use?

For now, nobody wants to spend it because they think it will be worth more. But eventually, retailers will say, 鈥淲e only take bitcoin.鈥� If that happens, there will be a run on the dollar.

People worry about quantum computing hacking bitcoin, but they鈥檒l hack the banks first 鈥� it’s way easier. I鈥檇 be more concerned about money in a bank than on a bitcoin ledger. Bitcoin also keeps perfect records; we wouldn’t need 85,000 agents because the blockchain can just pay whoever needs to be paid.

Where do you think the biggest potential for returns in the AI space are? Tooling, vertical AI, AI-native companies?

One or two general AI companies will win big and become 鈥渉ungry giants,鈥� the way was for software or bitcoin is for tech applications. A lot of people working around the edges might just be acquired by the AGI. We鈥檝e funded companies doing vertical AI: AI for patents, AI for science.

But remember, the big winners at the start of the internet were , and , and none of them ended up being a big part of the internet later. We don’t know who will rise from the ashes yet.

If you could implement one policy to accelerate innovation, what would that policy be?

Don鈥檛 regulate in anticipation of fearful outcomes. Regulate after something bad happens. Otherwise, you put a dark cloud over every innovator. I would also sunset laws. The ’33 and ’40 Acts are just keeping the poor poor and the rich rich. We should create a free market in education, too 鈥� let the best schools thrive and the worst die.

Some would argue in the case of bitcoin, we were slow to regulate. Do you disagree?

The U.S. just decided everything was a security and made it illegal. That鈥檚 why innovators are geofencing the U.S. to protect themselves from the ‘s long arms. Countries like El Salvador, Japan, Dubai and Abu Dhabi are rocking because they say 鈥渄o it.鈥�

I say decentralize everything. The guy at the tiller of the ship knows better than the general in Washington, D.C. You don’t want a president telling you how to raise your kids; you’ll do a better job than they will.

What’s the trait you now prioritize in founders that you didn’t a decade ago?

A love for the customer. It has to be an obsession. That love becomes a viral effect; customers love the product so much they tell everyone. People will naturally follow a leader who is that obsessed with their customer.

Illustration:

And underlining the pace at which some startups are now sprinting to billion-dollar-plus valuations, four of the new unicorns are less than a year old.

In exit news, 9-year-old fintech unicorn was acquired by for $5.2 billion. That鈥檚 well below its January 2022 valuation of $12.3 billion but still marks a win for earlier investors seeking liquidity.

Of the 31 companies that joined the board, 23 are U.S.-based and two hail from Canada. Germany, France, Belgium, Israel, Japan and India each added one new unicorn to the board last month.

Among sectors, AI and AI infrastructure contributed the most new unicorns, totaling nine from those two areas. The next-leading sectors, with three new unicorns each, were manufacturing and security propelled by AI. AI was also a major contributor to new unicorns in the semiconductor, defense and autonomous driving sectors.

The largest funding last month for a unicorn company was $20 billion to 鈥檚 at an . Within a month of that funding, xAI in early February announced a merger with another Musk-led company, rocketmaker .

11 exits

Brex鈥檚 acquisition by Capital One was the largest of the four M&A deals for unicorn-valued companies in January.

On the IPO side, seven companies went public, the most high-profile of which were and , both foundation AI model companies based in China.

Here are January鈥檚 newly minted unicorns.

AI

• , an AI research lab focused on human collaboration, raised a $480 million seed funding led by and 1. The less than 1-year-old Redwood City, California-based company was valued at $4.5 billion.
• , an AI scientific research lab, raised a $180 million seed round led by , and . The less than 1-year-old San Francisco-based company was valued at $1.5 billion.
• AI evaluation platform raised a $150 million Series A led by 2听补苍诲 . The less than 1-year-old San Francisco-based company was valued at $1.7 billion.
• Voice AI startup raised a $143 million Series C led by France-based . The 10-year-old San Francisco-based company was valued at $1.3 billion. As part of its announcement, Deepgram disclosed the acquisition of , a voice AI startup for restaurants and drive-thru ordering.
• , an infrastructure company for voice AI, raised a $100 million Series C led by . The 5-year-old San Jose, California-based company was valued at $1 billion.

AI infrastructure

• , an AI networking company, raised a $200 million Series A led by , and . The 1-year-old Santa Clara, California-based company was valued at $1 billion.
• GPU marketplace raised a $150 million Series B led by . The 2-year-old Palo Alto, California-based company was valued at $1 billion.
• , for secure AI run locally on devices, raised a Series A extension funding of an undisclosed sum. The 6-year-old Austin-based company was valued at $2.5 billion.
• , which manages a GPU marketplace, raised a Series C led by . The 6-year-old company was founded in Lithuania and is now headquartered in Miami. It was valued at $1 billion.

Manufacturing

• , a builder of factories for defense and the aerospace industry, raised a $131 million private equity funding led by . The 5-year-old Hawthorne, California-based company was valued at $1.6 billion.
• , a developer of no-code applications for manufacturing, raised a $120 million Series D led by . The 11-year-old Somerville, Massachusetts-based company was valued at $1.3 billion.
• 惭辞苍迟谤茅补濒-产补蝉别诲 , a manufacturing automation company utilizing modular robotics, raised a $90 million Series D led by . The 9-year-old company was valued at $1.2 billion.

Security

• , provider of security for cloud services in real time to protect from hackers, raised a $250 million Series B led by . The 3-year-old San Francisco-based company was valued at $1.5 billion.
• Tel Aviv-based , an AI security platform that integrates with existing security platforms to provide context on incidents, raised a $140 million Series D led by . The 6-year-old company was valued at $1.2 billion.
• Belgium-based , a developer-oriented security platform, raised a $60 million Series B led by . The 3-year-old company was valued at $1 billion.

Semiconductor

• , an AI chip developer to run transformer models, raised a reported $500 million funding led by . The 3-year-old Cupertino, California-based company was valued at $5 billion.
• , an AI chip design company, raised a $300 million Series A led by . The less than 1-year-old Palo Alto, California-based company was valued at $4 billion.

Cryptocurrency

• Stablecoin payments platform raised a $250 million Series C led by . The 4-year-old New York-based company was valued at $2 billion.
• Crypto payments network raised a $75 million Series C led by . The 5-year-old San Francisco-based company was valued at $1 billion.

Healthcare

• Maternity healthcare provider, raised a $92 million Series C led by Stripes. The 4-year-old New York-based company with plans to expand healthcare services to women and children was valued at $1.7 billion.
• , a co-ordination platform for medications across doctors, pharmacies and patients, raised a Series B led by . The 3-year-old New York-based company was valued at $1 billion.

Defense

• Paris-based , an autonomous drone maker, raised a $200 million Series B led by aircraft manufacturer . The 2-year-old company was valued at $1.4 billion.
• , a builder of secure software for the defense industry, raised a $136 million Series B led by . The 4-year-old Colorado-based company was valued at $1 billion.

Fintech

• Tokyo-based brokerage infrastructure provider raised a $150 million Series D led by . The 11-year-old company was valued at $1.2 billion.
• India-based , a payment infrastructure provider, raised a $50 million Series D led by . The 13-year-old company was valued at $1.2 billion.

Fitness

• , an owner of physical fitness brands and the parent of , raised a $785 million private equity financing led by . As part of the transaction it announced a merger with . The San Luis Obispo, California-based company was valued at $7.5 billion.

Autonomous Driving

• Toronto-based , a self-driving technology company, raised a $750 million Series C led by and ,valuing it at $3.8 billion. The 5-year-old company announced a partnership with to support robotaxis.

Social media

• , an AI-powered video generation platform for social media, raised an $80 million Series A extension funding which brings its Series A funding total to $130 million. The 3-year-old San Francisco-based company was valued at $1.3 billion.

Education

• Online tutoring platform raised a $150 million Series D led by at a $1.2 billion valuation. The 14-year-old Brookline, Massachusetts-based company was founded by Ukrainians and maintains a team in Ukraine.

Compliance

• ESG compliance software platform raised a $100 million Series C led by , a joint venture between and . The 7-year-old Baden-Wurttemberg, Germany-based company was valued at $1.1 billion.

Energy

• , a developer of a residential energy storage device for electricity and electric vehicles, raised a $163 million funding. The 7-year-old San Francisco-based company was valued at $1 billion.

Related 附近上门 unicorn lists:

• (1,684)
• (596)
• (37)
• (186)
• (115)
• (102)
• (868)
• (494)
• (226)
• (38)
• (470)

Related reading:

• High-Spending Investors Give 2026 An Active Start
• In The Era Of Unicorn Valuation Escalation, A Trillion Dollars Isn鈥檛 What It Used To Be
• Global VC Investment Surged In January, With U.S. Dominating Funding, But A Pair Of AI Model IPOs In China
• A Growing Share Of Seed And Series A Funding Is Going To Giant Rounds

Methodology

The 附近上门 附近上门 is a curated list that includes private unicorn companies with post-money valuations of $1 billion or more and is based on 附近上门 data. New companies are as they reach the $1 billion valuation mark as part of a funding round.

The unicorn board does not reflect internal company valuations 鈥� such as those set via a 409a process for employee stock options 鈥� as these differ from, and are more likely to be lower than, a priced funding round. We also do not adjust valuations based on investor writedowns, which change quarterly, as different investors will not value the same company consistently within the same quarter.

Funding to unicorn companies includes all private financings to companies that are tagged as unicorns, as well as those that have since graduated to .

Exits analyzed here only include the first time a company exits.

Please note that all funding values are given in U.S. dollars unless otherwise noted. 附近上门 converts foreign currencies to U.S. dollars at the prevailing spot rate from the date funding rounds, acquisitions, IPOs and other financial events are reported. Even if those events were added to 附近上门 long after the event was announced, foreign currency transactions are converted at the historic spot price.

Illustration:

This is a weekly feature that runs down the week鈥檚 top 10 announced funding rounds in the U.S. Check out last week鈥檚 biggest funding deal roundup here.

Venture investors’ thirst for AI isn鈥檛 close to quenched yet. That鈥檚 the takeaway from this week鈥檚 lineup of large U.S. funding rounds, which was mostly a mix of AI pure-plays and companies with a heavy focus on the technology.

The week鈥檚 largest round however, a $600 million financing for drone delivery provider Zipline, offered evidence that investors are also keen on platforms and technologies with applications in the physical world. The second-largest round, a $480 million seed deal for upstart AI lab Humans&, meanwhile, showed there鈥檚 also still appetite for ultra-ambitious newcomers.

1. , $600M, drones: Drone delivery unicorn Zipline it closed on over $600 million at a $7.6 billion valuation from investors including , , and . South San Francisco, California-based Zipline also says it expects to expand into at least four new states this year, with initial plans to begin service in Houston and Phoenix.

2. , $480M, AI: Humans&, an AI lab working to apply the technology in ways that are centered 鈥渁round people and their relationships with each other,鈥� secured $480 million in seed funding. The company was founded in September by top researchers from , , , and .

3. , $300M, AI infrastructure: AI infrastructure startup Baseten reportedly $300 million with backing from , and . The financing set a $5 billion valuation for the 7-year-old, San Francisco-based company.

4. , $250M, medical AI: OpenEvidence, an AI platform for doctors, announced that it picked up $250 million in a Series D funding round that doubled its valuation to $12 billion. and co-led the round, which marks the fourth fundraise for the Miami-based startup in less than a year.

5. , $215M, rare earth magnets: San Marcos, Texas-based Noveon Magnetics, a manufacturer of sintered rare earth permanent magnets, it secured $215 million in Series C funding, including $200 million from . The money will go toward expanding the company鈥檚 rare earth magnet manufacturing capacity.

6. , $200M, AI infrastructure: AI networking infrastructure startup Upscale AI $200 million in Series A funding led by , and . The financing set a valuation of more than $1 billion for the Santa Clara, California-based company, which was founded less than two years ago.

7. (tied) , $150M, online tutoring: Language learning marketplace Preply raised $150 million in Series D funding led by . The financing reportedly sets a $1.2 billion valuation for the 14-year-old, Brookline, Massachusetts-based company.

7. (tied) , $150M, AI inference: Inferact, a startup founded by creators and maintainers of open-source LLM inference engine vLLM, announced its launch along with $150 million in initial funding. and led the financing, which set an聽 $800 million valuation for the company.

7. (tied) , $150M, cybersecurity: Security provider Claroty picked up $150 million in Series F funding led by . The 11-year-old company, founded in Israel and now headquartered in New York, has raised close to $900 million in equity funding to date, per 附近上门 data.

10. , $115M, geothermal energy: Salt Lake City-based Zanskar, a startup applying AI to geothermal exploration, raised $115 million in Series C funding led by and joined by a long list of new and existing investors.

Methodology

We tracked the largest announced rounds in the 附近上门 database that were raised by U.S.-based companies for the period of Jan. 17-23. Although most announced rounds are represented in the database, there could be a small time lag as some rounds are reported late in the week.

Illustration:

So far this year, global edtech-focused startups have raised around $2.8 billion in seed- through growth-stage funding, per 附近上门 . That鈥檚 roughly flat with 2024 levels, pointing to stabilizing investment, albeit at a fraction of the peak a few years ago.

In the U.S., this year鈥檚 funding numbers are a bit stronger relative to 2024, with $1.2 billion invested in edtech startups so far. While still far off of the pandemic-era highs, 2025鈥檚 funding figures puts this year roughly on par with 2023.

What鈥檚 in and what鈥檚 out

Edtech is a vast space, covering everything from preschool lesson-planning to corporate upskilling. Given this, it鈥檚 not uncommon to see a downturn in one subcategory while another remains a funding favorite.

If we were to generalize trends looking at this year鈥檚 larger rounds and exits, it appears investors are particularly keen on opportunities in healthcare education and training. At the K-12 level, VCs are also backing startups deploying AI tools to customize lessons for individuals and free up teachers from routine, repetitive tasks.

As for what鈥檚 not hot, we鈥檝e seen a movement away from coding academies and teaching platforms, with the rise of coding automation tools. We鈥檙e also seeing a paucity of jumbo-sized funding rounds and not a lot of deals that look like pre-IPO financings.

What the biggest rounds tell us

So who is getting funded?

, a Berlin-based startup offering a tool to learn about and research medical information, raised the largest round, securing nearly in a March financing. The company started with a focus on medical students but now also markets to practitioners.

, a provider of content and online learning activities for young children, secured the next-biggest funding, a September round led by .

Other larger rounds this year include an financing for , developer of a chatbot to help students navigate college life and boost retention, and a Series B for , a provider of AI-enabled time-saving and productivity-enhancing tools for educators.

For a slightly broader view, below we put together a list of eight of the larger funding recipients in the education sector this year.

Buyers too

Edtech is also seeing some exit activity. This is coming in the form of M&A, as the IPO market has been quiet this year.

Most recently, , a platform for healthcare workers to learn new skills and obtain certifications, sold to , a software platform for senior living and home care providers, for an undisclosed sum.

Founded in 2013, Cambridge, Massachusetts-based Care Academy previously raised at least . The company, founded and led by -trained educator , carved out a niche offering upskilling opportunities to health workers like home care aides and nursing home staffers, opening a path to advancement for what are typically lower-paid positions.

Also in the health sphere, , an Austin startup focused on online learning tools for medical students and educators, sold this spring to exam prep provider , a portfolio company of private equity firm . Previously, 11-year-old OnlineMedEd had raised at least .

And in the post-secondary education space, , a Toronto-based provider of software tools for colleges to attract and retain students, sold a majority stake to PE firm in August.

The optimist case

Looking ahead, the optimist case is that founders, investors and acquirers alike will find plenty of appealing opportunities in ed tech.

Longtime education startup investor considers the education and training market to be one of the fastest-growing sectors in the global economy. In its 2025 , the firm projects the global education market is on track to surpass $10 trillion by 2030.

In terms of growth, Owl unsurprisingly points to AI as the largest ed tech driver. In recent years, the report notes, AI in the classroom has moved beyond the experimentation stage and is already proving vital in saving educators hours of work, providing personalized tutoring to students, and helping craft compelling lesson plans.

Eventually, it鈥檚 likely we鈥檒l see the impact of AI innovation in edtech also showing in the form of more funding for startups in the space.

Related 附近上门 queries:

Related reading:

• Cursor鈥檚 $2.3B Financing Reminds Us: Coding Automation Is Still Ultra-Hot

Illustration:

Those were once again the findings of our annual look at which U.S. colleges and universities graduate the highest number of funded startup founders, based on 附近上门 data.

As usual, four universities 鈥� , , and 鈥� hung on to the top four slots. The remaining names on our list mostly include a mix of large state research universities, Ivy League institutions and private schools known for tech and business.

For a more detailed overview, we posted the full ranking below. It tracks the number of founders affiliated with each school whose startups raised seed through growth-stage funding this past year.

Leading schools are also hard to get into

When we look at the top-ranked schools on our list, all are known for providing a rigorous education. What they also share in common are highly competitive admissions processes.

This is particularly true for undergrads. For the , for example, Stanford admitted just 3.6% of applicants, followed by Harvard at 3.7%, and MIT at 4.6%.

Many of those accepted are likely people well-poised to succeed in entrepreneurship regardless of degree. For evidence, look at the history of prestigious university going on to found valuable companies such as , and . That said, of course, we also see co-founders in areas that rely on their academic expertise, particularly among AI and biotech unicorns.

Public university outperformers

Public universities, which have higher acceptance rates for in-state residents, are harder to measure for admissions competitiveness. But when it comes to funded-founder track records, it鈥檚 clear that a few names stand out.

The far-and-away leader in this category is UC Berkeley, which benefits from its highly regarded STEM programs as well as its location in the San Francisco Bay Area, the global capital of the venture industry.

ranks a distant second, with just over 100 grads going on to launch companies funded in the past year. and are close behind, with 97 and 93 funded founders, respectively, followed by , with 86.

Ivies and hard-to-get-into private schools round out the list

All eight schools in the made our list, which is what we鈥檝e seen in past years as well. After Harvard, the next-biggest Ivies for funded startup founders were , and .

Of course, there are also other private, non-Ivy universities that are notoriously hard to get into or known for particularly demanding STEM curriculum. Not surprisingly, these churn out a lot of funded founders as well. In this category, beyond the aforementioned Stanford and MIT,聽 , and also ranked high on our list.

Since size is a consideration here as well, we also ought to give a shout-out to some smaller schools that produce a disproportionately high number of funded founders. For instance, , with only about 2,500 students, outperforms many much larger well-regarded schools.

Business schools play an outsized role

Several of the schools on our list also got there to a large degree as a result of business school grads.

Per 附近上门 data, standout business schools for funded founders include , , Northwestern鈥檚 and Penn鈥檚 , among others.

Business schools have also factored in favorably in prior research looking at universities that produce the highest numbers of startup CEOs.

No big changes

While it may seem like the world is changing at a faster rate than ever, funded founder and university rankings are one of the more stable indicators. Over the years we鈥檝e been crunching these numbers, the component universities of our list, and their respective rankings, haven鈥檛 budged dramatically.

Bottom line: If your plan is to be a unicorn founder, attending Stanford is still probably not a bad idea.

Related 附近上门 query:

Related reading:

• Harvard鈥檚 Still A Startup Heavyweight
• The Unicorn Founder Myth: Why Education Actually Matters
• These Schools Graduate The Most Funded Startup CEOs

Graduates and famous dropouts have launched companies that are collectively valued in the trillions today. And alongside and , Harvard regularly ranks in the top three schools for graduating founders of funded startups.

But how is it measuring up lately?

With graduation season upon us, and Harvard garnering headlines in its with the administration, we thought it would be timely to look at the school鈥檚 recent record in launching funded founders and CEOs. This includes a look at students鈥� roles in founding and leading unicorns, launching new funded startups, and taking companies to exit.

Still going strong

First, we looked at Harvard attendees鈥� role in founding some of today鈥檚 most valuable private, venture-backed companies. To illustrate, we put together a list of 12 of the most heavily funded startups with a Harvard-affiliated founder.

Notably, the company at the top of our list 鈥� 鈥� was co-founded by , who studied at Harvard, but chose building a startup over completing a degree. He joins a famous league of Harvard dropouts-turned-billionaire-founders, including and .

Another heavyweight in the news this week is human resources software provider , which just raised $450 million at a $16.8 billion valuation. Co-founder and CEO is a Harvard grad.

The other mega-funding recipient on our list 鈥� Medicare benefits provider 鈥� counts Harvard grads and brothers and as co-founders. Ed Park also serves as CEO.

$200B+ in funding to Harvard-affiliated startups

Private, venture-backed companies founded by former Harvard students have also raised copious sums over the years.

Per 附近上门 data, there are currently over 1,900 such companies that have raised over $1 million in equity investment. To date, those companies have collectively in equity funding over the years.1

Much of that funding is rather recent. Per 附近上门 data, companies with a Harvard-affiliated founder or co-founder have raised over $54 billion in the past year. However, this is only if one includes nearly $47 billion for , which counts 鈥� who attended Harvard but later transferred to MIT 鈥� as co-founder and current CEO.

Harvard is a CEO factory too

Harvard is also pretty good at churning out future CEOs of high-valuation venture-backed companies.

Per 附近上门 data, there are 158 Harvard alumni of current unicorns, emerging unicorns or exited unicorns. Many of them are also founders or co-founders of the companies they lead.

Many of the best-known names on the list are onetime unicorns that are now public. This includes several that made their market debuts several years ago, such as , , and .

In prior surveys, we鈥檝e found that , in particular, has a strong track record for graduating future CEOs. That鈥檚 the alma mater of many unicorn CEOs as well. (Although it鈥檚 tempting as an aside to note that the success of OpenAI, Stripe, and , gives credence to the notion that perhaps people who matriculate at Harvard but later leave are the most fundable entrepreneurs.)

The path ahead

Whether Harvard continues on its well-worn path of serving as a launchpad for startup founders and leaders, of course, remains to be seen. In coming days, we鈥檒l also be taking a look at how the school compares to other well-regarded institutions for matriculating future funded founders. Stay tuned.

Related 附近上门 queries:

Related reading:

• These US Universities Graduate The Highest Number Of Funded Founders
• The Unicorn Founder Myth: Why Education Actually Matters

Illustration:

And now, in the same surreal timeline, we鈥檝e got signing 鈥� directing the and the to prioritize funding for AI-related research and grants.

You truly can鈥檛 make this stuff up and even if you could, you no longer have to.

To be clear: I鈥檓 not anti-technology. AI has a role to play in education. Personalized learning, intelligent tutoring systems, data-driven insights 鈥� these are powerful tools when used thoughtfully. But let鈥檚 not kid ourselves. We鈥檙e living through a moment where the Trump 2 administration is taking a DOGE chainsaw to the very foundations of public education. And instead of confronting that, we鈥檙e being told to get excited about chatbots in the classroom.

This isn鈥檛 leadership. It鈥檚 deflection.

Funding for future success

The truth is, AI is not the lifeline our education system needs. Certainly not right now. What we need 鈥� what we鈥檝e needed for decades 鈥� is serious investment in teachers, classrooms, infrastructure and support services. And we鈥檙e getting the opposite.

The Trump administration is proposing deep cuts to key education programs, gutting federal support for public schools, and pushing policies that favor privatization and deregulation over student success. Amid all that, we鈥檙e supposed to believe that some AI-powered lesson plans are going to move the needle?

Please.

Let鈥檚 start with the obvious: AI doesn鈥檛 fix underfunded schools any more than A1 sauce would. You can鈥檛 put an algorithm into a building with no heat, no internet and no functioning restroom and expect a miracle. You can鈥檛 expect a teacher managing 35 kids on her own to suddenly have the time and training to integrate AI into daily lesson plans (if they even have the time to make one actual lesson plan a week). And you can鈥檛 tell communities that are already struggling to get basic resources that what they really need is machine-learning software.

This executive order assumes that what鈥檚 missing in American education is innovation. But we don鈥檛 have an innovation problem 鈥� we have a priorities problem. Our students aren鈥檛 falling behind because teachers aren鈥檛 tech-savvy enough. They鈥檙e falling behind because our country refuses to treat education like a public good.

What鈥檚 broken

We鈥檝e normalized schools with outdated textbooks, overworked staff and dilapidated facilities. We鈥檝e made it acceptable for teachers to buy their own supplies, for students to skip meals, and for mental health crises to go unanswered.

And now, in the middle of that, this administration wants to convince us that the real problem is that we鈥檙e not moving fast enough on AI.

Let鈥檚 also be honest about what AI in schools usually means. It doesn鈥檛 mean teachers getting sophisticated tools that make their jobs easier. It means more standardized testing, more data collection, more screen time and more surveillance, especially for kids in low-income communities.

It means feeding student information into systems built by private companies, often with little oversight or transparency. It means potentially outsourcing educational decisions to algorithms that don鈥檛 understand context, nuance (sidebar: do any of us get nuance anymore?) or humanity.

It鈥檚 a far cry from the glossy pitch the administration is selling.

Widening the digital divide

And let鈥檚 not ignore the inequity intentionally baked into all of this. AI-enhanced education requires reliable internet, up-to-date devices, tech-literate staff and digital infrastructure 鈥� things that affluent districts are more likely to have. For schools in underserved areas, this push risks widening the digital divide under the guise of modernization.

What鈥檚 being framed as progress is actually an elegant Trojan horse for deeper inequality. The schools that most need real, human-centered support are the least likely to benefit from this initiative.

It鈥檚 particularly galling that all this is being rolled out with a heavy dose of PR spin. The A1 comment might鈥檝e been a gaffe, but it was also revealing. It showed just how deeply unserious this administration is about the reality on the ground in American schools. It was meant to sound cool, forward-thinking, maybe even meme-worthy. Instead, it became a symbol of how disconnected Trump鈥檚 appointee is from what鈥檚 actually happening in classrooms across the country.

What schools really need

Teachers aren鈥檛 asking for AI. They鈥檙e asking for manageable class sizes, fair pay, mental health resources and the ability to teach without being completely buried by bureaucracy. Students aren鈥檛 crying out for machine learning 鈥� they鈥檙e asking for support, stability and a system that sees them as more than test scores or data points. And parents aren鈥檛 begging for the latest edtech. They want to know their kids are safe, challenged and cared for at school.

AI is a tool. That鈥檚 it. It鈥檚 not a savior, it鈥檚 not a substitute, and it鈥檚 certainly not a replacement for public investment. If the Trump administration were serious about improving education, it would be fighting to expand school funding, not slash it. The administration would be making college more affordable, not reversing progress on student debt. It would be strengthening teacher pipelines, not weakening them. And it would be protecting public schools, not undermining them.

Instead, we get a photo op and a tech policy wrapped in buzzwords.

So yes, Secretary A1, AI has its place. But until we鈥檙e ready to fund schools like they matter, treat educators like professionals, and address the real, systemic issues at the heart of this crisis, all the artificial intelligence in the world won鈥檛 save us.

And that鈥檚 not artificial. That鈥檚 just reality.

is the chief strategy officer for . He holds a law degree and has taught entrepreneurship at and the , and was elected to Fastcase 50, recognizing the top 50 legal innovators in the world. His writing has been featured in , , , , , , , and many other publications. He was nominated for a Pulitzer Prize .

Related reading:

• Can We Please Still Have Nice Things That AI Hasn鈥檛 Touched?
• Stealing Ghibli: How AI Has Crossed The Line Into Creative Theft
• Who Owns Your Face? The Legal Fight for Identity In The Age of AI
• When Legal Tech Fails: The California Bar Exam Disaster And The Risks Of Blind Innovation

Illustration:

AI is making its way into every part of our lives, and education is no exception. A recent survey 88% of U.S. parents think AI is essential to their children鈥檚 education, but most aren鈥檛 sure if it鈥檚 being used in their child鈥檚 classroom.

One of the most promising uses of AI is personalized tutors. Considering the estimated shortage of K-12 teachers globally by 2030, there鈥檚 a big opportunity for AI to help close this gap.

It鈥檚 not, however, as simple as plugging in some APIs and writing prompts. Building a kid-safe AI platform has required years of development and millions in fundraising efforts by me and my team.

Along the way, we鈥檝e learned some valuable lessons that could serve others looking to do something similar.

Language learning apps aren鈥檛 made for kids

After moving my family from Siberia to California in 2014, I saw how difficult the language barrier made things for my daughter Sofia. She was struggling to learn English and make friends, which was affecting her confidence.

As we tried different solutions some problems became clear:

1. Language learning apps required reading skills (which she didn鈥檛 yet have);
2. They also didn鈥檛 allow her to practice speaking (a crucial part of language acquisition); and
3. Private online tutors were expensive and often worked with scripted lessons.

The scripted nature of the lessons made it clear to me that they could be automated in far more affordable and fun ways. So I pitched the idea to my co-founder and we dove into developing a product.

We found demonstrating the features of AI tutors that worked best for kids and knew we needed to replicate them. Effective AI tutors are animated characters with voice recognition and conversational AI capabilities as these reflect the experience of a live teacher.

True safety is made in-house

Of course, building a children鈥檚 conversational AI app came with unique challenges. First, existing automatic speech recognition, or ASR, systems struggle to understand children鈥檚 voices. Because children鈥檚 voices are higher-pitched, they also mispronounce words, emphasize the wrong parts of a sentence (sometimes screaming or singing), and use incomplete grammar, systems trained on adults get thrown off.

Consequently, we had to build our own ASR just for kids, which we trained on more than 5,000 hours of children鈥檚 voice data from around the world.

Products for children also come with regulations like the Children鈥檚 Online Privacy Protection Act. Prioritizing safety meant we also needed to build our own in-house models for our avatar鈥檚 language and behavior.

Our models are more limited in their output than some of the other powerful LLMs and text-to-video models available, but this is by design. You don鈥檛 want to risk an avatar veering off into irrelevant conversations.

There鈥檚 a massive untapped market for future innovation

The critical teacher shortage is worldwide, and in countries such as , and , English fluency is often considered a ticket to better career opportunities. As demonstrates, a child鈥檚 ability to learn a new language peaks before age 7, so there鈥檚 huge global demand for accessible early language learning. And a product with a global audience not only increases your total addressable market, but also makes establishing partnerships and pilot programs easier.

Building an AI app for kids began as a personal journey for me, but over time it鈥檚 grown into a greater mission. If I have any other advice, it鈥檚 to pick a problem you care about deeply and be prepared for the heavy lifting required up front. It鈥檚 tough, but if you鈥檙e working on something that can make a difference, every challenge is worth it.

is CEO and co-founder of , the first multimodal, conversational AI tutor designed to teach young children English.

Illustration:

However, it now looks like Vision Fund 2 is coming back strong. In the last nearly two quarters, the fund has participated in 13 funding rounds 鈥� many of them large and many of them involving artificial intelligence 鈥� per 附近上门 .

The baker鈥檚 dozen of deals, which included funding big rounds for the likes of and , are more than the fund participated in the previous four quarters combined. In the 12 months through Q3 2024, the fund only took part in 10 deals.

Back on track?

Of course, just three years ago, 13 deals in five-plus months for the Vision Fund 鈥� once known for shrewd investments in startups including and 鈥� would hardly have been news. In Q3 2021 alone, the Vision Fund participated in 66 rounds.

But that was a very different time for venture investing. Interest rates were low, money was basically free, and venture capitalists couldn鈥檛 keep their checkbooks in their pockets for longer than five minutes.

When the market came back to reality in 2022, SoftBank founder said it would pull back on investments as the Vision Fund suffered big losses.

The respite seemed destined not to last long. By mid-2023, Son told investors he would again shift from 鈥渄efense mode鈥� as the firm wanted to be a leader in AI and robotics.

However, as Son made investments directly from SoftBank, the Vision Fund stayed quiet. For the two-year period between Q3 2022 and Q3 2024, the fund took part in only 33 deals.

That was not all that surprising considering the fund continued to get markdowns. Just last month, in fact, reported a net loss of nearly $2.4 billion for its fiscal third quarter, including a loss of about $2 billion for its once-heralded Vision Fund unit as shares of and fell.

Regardless, late last year the fund started to spark back to life. In the final quarter of 2024, the fund made six investments including participating in AI-powered search startup 鈥檚 in December.

That was not its only AI-related deal. The Vision Fund also took part in data center firm 鈥檚 massive $1.2 billion Series B.

However, while AI has played a role in the fund鈥檚 comeback, it has not ignored other sectors. The fund took part in New York-based clinical-stage biopharmaceutical startup 鈥檚 $215 million Series B led by and , as well as India-based 鈥檚 $150 million Series F. The startup provides executive education programs.

The fund also made a strategic investment of an undisclosed amount in cloud security startup in November.

The new year

SoftBank has continued that investment cadence into 2025. Through early March, the Vision Fund made seven investments. Similar to Q4, the fund has taken part in some big rounds.

In January, it participated in the $100 million Series C for Seattle-based , a developer of in vivo cell therapies, as well as a $200 million Series E for Spain-based business travel management platform .

The fund also co-led 鈥� along with and 鈥� the $425 million Series F for fusion startup . That round valued the energy startup at $5.4 billion, as investors look to pour cash into new energy sources as power needs increase due to AI and other advances.

Last month, the fund took part in Ireland-based no-code workflow automation platform 鈥� $125 million Series C, but co-led Boston-based neutral-atom quantum firm 鈥檚 $230 million convertible note offering along with .

What it means

Despite those deals, the fund likely will never jump back to its level of investment back in 2021.

However, it is interesting to note that even as some of the large crossover funds like and that helped lead the charge in venture in 2021 have yet to really come back to the industry, the Vision Fund seems more curious.

It is also interesting that while SoftBank itself seems laser focused on all things AI 鈥� and the proposed $500 billion AI , for example 鈥� the Vision Fund has not narrowed its scope to just solely AI, but also deals in spaces such as biotech and cyber.

Of course, with the fund鈥檚 recent record, perhaps it would be wisest to keep its cash in its pockets. For startups, however, it may be exciting to see a larger investor re-emerging.

Related 附近上门 Pro list:

Related reading:

• Eye On AI: As SoftBank Piles Up Losses, It Doubles Down On AI
• 鈥嬧�婼oftBank Vision Fund Hit With More Losses
• SoftBank鈥檚 Vision Fund Posts $1.7B Loss
• SoftBank May Invest Up To $25B Into OpenAI 鈥� Report

Illustration:

The data shows that they developed their skills at tech giants, elite universities and even military organizations. Four in 10 unicorn founders previously founded other companies, and former personnel are 3x more likely to build U.S.-based unicorns than average.

From employees to entrepreneurs

One quarter of unicorn founders previously worked in scientific research or technology development. Another 22% had already led companies as CEOs, while 9% served as CTOs before launching their unicorns.

Engineers (21%), software engineers (17%) and product managers (14%) round out the most-common background roles 鈥� a clear pattern showing that leadership, technology and marketing skills form the foundation of unicorn founder success.

Source:

This professional background matters significantly. The data shows unicorn founders previously worked at 6,109 different organizations, but only 33 entities produced 15 or more future unicorn builders. It’s no coincidence where these future unicorn builders come from 鈥� almost half of the top 33 organizations that produce them had VC funding themselves.

Tech giants serve as particularly effective training grounds. Google alone produced 96 unicorn founders, followed by (64) and (42).

Elite universities also function as powerful launchpads, with (43 founders), (40) and (33) leading the pack (these are for founders who worked at these universities rather than those who studied). Financial powerhouses such as (27 founders) and government agencies including (19 founders) complete the picture.

Serial entrepreneurship: Practice makes perfect

Our research shows that 40% of unicorn founders had previously started other companies, and 60% of unicorns have at least one serial entrepreneur on their founding team. The path to a billion-dollar success often involves previous ventures 鈥� both successes and failures.

Still, the majority (60%) of unicorn founders hit it big on their first attempt. , for instance, co-founded without any prior experience as a founder.

One quarter needed a second try, like who created after first building a social platform for mobile games. Another 9% succeeded on their third attempt 鈥� including , who co-founded after starting two B2B firms 鈥� while 6% required four or more ventures before achieving unicorn status.

We also found interesting patterns across industries: financial services companies show the highest rate of serial entrepreneurs (61%), while healthcare unicorns have fewer prior founders (31%) compared to information technology (38%).

Unexpected unicorn factories: Military and government experience

Some of the most intriguing findings come from less obvious unicorn talent sources.

Founders with Israel Defense Forces experience are 3.1x more likely than average to build U.S.-based billion-dollar companies. Other government and research organizations also outperform expectations, including (2.1x), the (1.7x) and (1.6x).

These organizations likely instill valuable skills that transfer to entrepreneurship 鈥� strategic thinking, mission focus, technological expertise and performance under pressure.

The path to success: What founder backgrounds tell us

If you’re dreaming of founding the next unicorn, our findings deliver a mix of good news and reality checks. On the one hand, certain professional backgrounds clearly correlate with unicorn-building success. Time at top tech companies, elite universities or specific government organizations appears to provide valuable preparation for future founders.

At the same time, the significant percentage of founders who succeed after multiple attempts demonstrates that persistence matters. For those whose current ventures aren’t performing as hoped, the data suggests that learning from failure and trying again often leads to eventual success.

Our research paints a clear picture of what it takes to build a unicorn. Yes, where you worked matters 鈥� plenty of founders come from Google and Stanford 鈥� but the door isn’t closed to others who gain the right experience and don’t give up after failure.

is the foremost academic expert on venture capital. As the founder of the Venture Capital Initiative and a professor of private equity and finance at , where he teaches a popular class on venture capital, his research has been widely published in leading academic journals and featured in , , and the . He frequently leads workshops and executive sessions for senior business and government leaders around the world and has consulted for companies and investors on the venture industry trends and corporate innovation. In 2023 he was named a Top Voice on . ().

Note on methodology and sources

For this study, we define unicorns as VC-backed, U.S.-based companies that achieved a confirmed $1 billion-plus post-money valuation in a primary private round or had a liquidity event (such as an IPO or an acquisition) at a confirmed $1 billion-plus valuation between 1997 and 2021. To construct our unicorn list, we started with 鈥渦nicorn candidates鈥� from well-known sources such as and , as well as from datasets that report private funding round and liquidity event details, such as 附近上门, and . We then manually confirmed and cross-checked the location and funding details to decide on the inclusion of each company in our final unicorn list. This process resulted in a total of 1,110 unicorns. For each unicorn we also identified a peer U.S.-based VC-backed company that raised its first venture round in the same year. We call the sample of such peers a random sample. For each of the companies in the unicorn and random samples we identified founders and co-founders (we use 鈥渇ounder鈥� and 鈥渃o-founder鈥� interchangeably) from all the sources mentioned above, LinkedIn, public filings and many others. In total, we identified and confirmed 4,975 founders (different data exercises may use subsamples of this data).

Illustration: