Jan 2, 2019 · 46m · a16z
a16z Podcast | The Role of Academia in the Startup World
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of the a16z Podcast, Marc Andreessen and Chris Dixon discuss how universities, tech transfer offices, and interdisciplinary education drive tech entrepreneurship, while evaluating how computer science intersects with physical sciences and venture capital.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The host as informed peer 4.4 Guest teaching 5.9 Guest disagreement 1.9 The host pushing back 1.0
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 4:43 Dismantling contradictory media critiques
Andreessen aggressively calls out media commentators for publishing mutually exclusive critiques of technology—that tech does nothing important while simultaneously destroying society—without attempting to reconcile them.
Hardest push from the host ▶ 4:15 Challenging Silicon Valley's focus on big problemsDixon directly presents the prevailing skeptical argument against tech industry priorities, invoking the classic trope that society was promised self-driving cars and moonshots but received Twitter instead.
Biggest teaching moment ▶ 14:00 Applying C.P. Snow's 'Two Cultures' to modern CSAndreessen provides an intellectual lecture drawing on C.P. Snow's 1950s essay to reframe the tension between liberal arts and computer science into a vision for a 'third culture'.
The host holds their own ▶ 26:25 Dismantling pop-culture startup mythsDixon showcases deep VC operational expertise, rejecting pop-culture depictions of startups as single patented lightbulb moments in favor of long-term engineering execution and recruiting dynamic.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The host as informed peer | Guest teaching | Guest disagreement | The host pushing back | Why |
|---|---|---|---|---|---|---|
| Evolution of Academic Computer Science | 2 | 5 | 1 | 0 | Host Chris Dixon sets up the interview by asking Marc Andreessen about his early NSF funding for Mosaic and how university CS departments have evolved over time. Andreessen provides historical context on how CS departments historically favored academic theory over industry utility. | |
| Addressing the Critique of Big Problems | 3 | 6 | 4 | 2 | Dixon introduces the popular critique that Silicon Valley avoids big problems, prompting Andreessen to aggressively dismantle media hypocrisy. Andreessen notes that commentators contradictorily claim tech does nothing meaningful while simultaneously destroying culture and jobs. | |
| University Ecosystems and the Stanford Model | 5 | 6 | 2 | 2 | Dixon asks about Stanford and Berkeley's success in seeding startups, prompting Andreessen to contrast enlightened university philanthropy models with draconian IP offices. Dixon steps in with domain expertise to explain why tech VCs view patents as secondary to talent and open-source contributions. | |
| K-12 CS Education and the Third Culture | 3 | 7 | 2 | 1 | Dixon brings up NYC's new K-12 CS initiative and asks how to fix the CS pipeline. Andreessen educates the audience by referencing C.P. Snow's 1959 'Two Cultures' essay and proposing a 'third culture' where CS intersects with liberal arts. | |
| Computer Science Intersecting with Life Sciences | 3 | 6 | 1 | 0 | Dixon asks for specific examples of CS intersecting with other fields, leading into a deep dive on biology. Andreessen explains how generational shifts are bringing young bio PhDs who coded from childhood into computational biomedicine. | |
| Fostering Student-Led Startup Teams | 6 | 6 | 2 | 1 | Dixon asks how CS professors can foster student-led startups. Dixon demonstrates significant expertise by dismantling Hollywood myths about static algorithm inventions versus long-term team building and recruiting, which Andreessen reinforces with the Nicira case study. | |
| Developing Leadership and Business Skills in Engineers | 5 | 5 | 1 | 0 | An audience member asks about leadership development for technical students. Andreessen argues for teaching business skills directly to engineers rather than relying on Dilbert-style MBAs, while Dixon shares knowledge on successful co-op models like Waterloo. | |
| Navigating University Patents and Tech Transfer | 6 | 6 | 2 | 2 | An audience member asks about licensing university patents. Dixon demonstrates strong industry perspective by arguing that aggressive university tech transfer offices counterintuitively destroy value and revenue compared to open-sourcing research. | |
| Contrarian Investing and Counterintuitive Ideas | 7 | 6 | 2 | 1 | An audience member asks how the firm identifies contrarian investments. Dixon takes the lead in explaining how venture capital differs fundamentally from public markets because founders select investors, after which Andreessen details their 'good ideas that look like bad ideas' framework. |
Statements from this episode (25)
Andreessen: NSF rejecting support funding for Mosaic catalyzed Netscape's founding
“So we applied for NSF funding to basically build a customer support team. And they informed us very kindly that that wasn't part of what NSF funds. So that was a good catalyst to go start a company which was very helpful.”
Andreessen: CS Departments Have Shifted Focus Toward Practical Industry Impact
“The biggest change is just we have a sense for a lot of universities we deal with The computer science departments have a much, I mean, they still want to breed PhDs and professors, but a much bigger focus on practical impact in industry and, you know, the stu…”
Andreessen: Media commentators publish contradictory critiques of Silicon Valley
“Nobody attempts to ever reconcile those two critiques and the same commentators will literally write both critiques. In, like, different columns, you know, two weeks apart without ever attempting to reconcile them.”
Andreessen: Technological progress in atoms lags behind progress in bits
“It's a fair critique that there are many fields that are not moving as fast as they could. And in particular fields, you know, applications of computer science or other fields of engineering that intersect more in the world of atoms as opposed to the world of …”
Andreessen: Applying computer science is the best way to advance physical industries
“We now have the opportunity to make progress in many of those fields, and in fact, the best way to make progress in many of those fields is to apply more computer science.”
Andreessen: John Hennessy is maybe the best university president ever
“John Hennessy who, in addition to being a legendary theoretical computer scientist was also himself a former company founder. And so, and you know, maybe the best university president anybody's ever had. You know, that's worked very well.”
Andreessen: John Arrillaga donated over $1 billion to Stanford
“Because he went on to become a very successful real estate developer and has donated what has to be, at this point, more than a billion dollars back to Stanford, personally. And there's, you know, a 150 buildings on the Stanford campus that he's paid for.”
Dixon: Patents are never core to any a16z startup investment decision
“We, you know, for example, when we make investments, we ask about patents, but it's like a side item. Like, it's never actually core to any investment, because anyone who's built a company knows it's a dynamic process.”
Andreessen: Combining CS and liberal arts enables unique breakthroughs
“He proposes what he calls the third culture which are the people who can bring the two cultures together and the people who can bring, you know, physics, or now computer science, into liberal arts, and the people in liberal arts who can learn and understand, e…”
Marc Andreessen's first college job was writing code for physicists
“My job in college, my first job in college was actually to write computer code for physicists who hated computers.”
Andreessen predicts continuous bodily tracking and quantified self will become standard
“Quantified self is an example we think is going to evolve into. You know, a cornerstone of biology in the future. You know, we, you know, we one of our theories is we think we're living in the stone age today in the sense of, like, we really don't know what's …”
Andreessen: University spin-outs fail when professors found companies but remain in academia
“There's a very clear failure case that, which is the dominant failure case that I think that we see. Which is the professor has the idea. The professor starts the company but with no intention of going full-time at the company. The professor runs the company o…”
Dixon: Only great product engineers can recruit top engineering teams
“What startups really are is a dynamic process that probably, you know, 90% of it is frankly recruiting a great product and engineering team, and What we find is only, just empirically, is that only great product engineering people are able to recruit those peo…”
Andreessen: Nicira scrapped its OpenFlow tech immediately after a16z investment
“So we funded this here at NetWeQ's networking lab, and it was the OpenFlow. OpenFlow was the sort of protocol was the approach and Martine had, had, had been one of the main people doing that. And so they start in Asira, and we're all excited, because we're li…”
Andreessen: PageRank is only one of 900 factors in Google search
“It's one of, like, PageRank is, like, one of, like, 900 factors now that go into a Google search result. The other 899 happened after they left Stanford.”
Andreessen: Startups require five years of 18-hour days, six days a week
“There is no substitute. There is no substitute. The big takeaway, and this applies to everything we fund, and it's what we work on all the time, there is no substitute to the first five years of people in the office, 18 hours a day, six days a week, grinding a…”
Andreessen: Top engineers with business skills should found companies and hire MBAs
“The much better model is for the engineer to have the business skills. For the engineer to be a top flight engineer, but also have the business skills to be able to start a company. And then the engineer hires the MBA as, you know, the head of marketing, or th…”
Andreessen: Stanford and Berkeley run engineering business programs independent of business schools
“What's so striking about the Stanford and Berkeley programs is those programs are completely separate from the business school.”
Dixon: University of Waterloo engineering undergraduates complete six-plus internships
“Waterloo's a great example of that, where they, I think they have it's like they have a crazy number, like six or plus internships over their four years of undergrad.”
Andreessen: Silicon Valley tech firms target interns to lock in hiring a year early
“In the last five years alone in the Valley, the tech companies have hugely increased their focus on their internship programs. And basically the goal is to get all of your best students as interns, and then be able to basically get them locked in at least a ye…”
Andreessen: Deep hardware patents are valuable, software patents are not
“What I would argue is, like, the further down the stack, the more useful the patents are. Like, patents at the level of chips or radios you know, can be, or, you know, I don't know, memory or something like that can be incredibly valuable. Patents at the level…”
Andreessen: Aggressive university patent licensing can kill spinoff startups in the cradle
“I have seen cases, and I've lived through cases, where the universities take a very draconian view on this, and the university licensing office has this just highly inflated view of the value of what they're going to be licensing, and they can kill companies i…”
Andreessen: Legislative reform of the Bayh-Dole Act is not happening soon
“I think a really good reform would be to just eliminate that law and let universities make these decisions on their own. But that's not happening anytime soon.”
Andreessen: a16z evaluates 2,000 startups every year
“And so, you know, and we see, you know, 2000 startups a year, and, you know, many, many of them are good ideas that are just too obvious, and then many of them are bad ideas that are just bad ideas, and then every once in a while, there's the good idea that lo…”
Andreessen: a16z assumes a 50 percent failure rate on counterintuitive startup bets
“And, you know, and it's, you know, it, and by the way, it, even there we're gonna have, we assume we basically have a 50% failure rate. So, you know, we assume that half the time you know, we're gonna back the, what we think is a good idea, looks like a bad id…”