Jan 2, 2019 · 42m · a16z
a16z Podcast | AMPLab, the Power of Open Source, and the Future of Systems Software
Peter Levine · 15m spoken Michael Franklin · 12m spoken Michael Copeland · 7m spoken Haoyuan 'H.Y.' Li · 3m spoken
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
Host Michael Copeland joins AMPLab co-founder Michael Franklin, a16z partner Peter Levine, and Tachyon founder H.Y. Li to discuss how UC Berkeley's AMPLab revolutionized system software development through open-source collaboration, rapid commercialization, and venture-backed academic startups.
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 2.6 Guest teaching 4.5 Guest disagreement 1.8 The host pushing back 1.8
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 7:00 Franklin rejects premise on research focus
When Copeland suggests AMPLab succeeded by being narrowly focused rather than trying to boil the ocean, Franklin humorously counters that they were indeed trying to boil the ocean.
Hardest push from the host ▶ 15:27 Copeland presses guest on defining good ideasAfter Franklin hesitates to answer how research is commercialized, Copeland reframes and insists on defining what constitutes a good research idea.
Biggest teaching moment ▶ 11:47 Levine educates on open source value metricsLevine re-educates the panel on IP valuation, proving that original developer retention matters far more than legal patent ownership in open source ventures.
The host holds their own ▶ 6:42 Copeland attempts synthesis of AMPLab secret ingredientCopeland proactively formulates a conceptual thesis that AMPLab succeeds through targeted problem scope rather than broad unfocused research.
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 |
|---|---|---|---|---|---|---|
| The Origins and Creation of Berkeley AMPLab | 2 | 3 | 1 | 1 | Host Michael Copeland asks open background questions about the founding of AMPLab. Michael Franklin explains returning from startup sabbaticals to academia and feeling the need for faster-paced systems research. | |
| The AMPLab Open Source Model and Industry Collaboration | 3 | 5 | 4 | 3 | Copeland attempts to summarize AMPLab's secret ingredient as focused research rather than trying to boil the ocean. Franklin playfully rejects the premise by asserting they were trying to boil the ocean and explains how open source removes friction. | |
| Open Source IP Rights, Patents, and Academic Philanthropy | 3 | 5 | 2 | 2 | Copeland asks about IP distribution among corporate sponsors. Franklin and Levine educate the host on why software IP ownership matters far less than retaining original project creators and avoiding weak code forks. | |
| Driving Research Impact and Academic Recognition | 3 | 5 | 3 | 3 | Copeland presses Franklin on indicators of commercial potential in research. Franklin reframes the topic, stating commercial viability is secondary to researcher passion and student excellence. | |
| Machine Learning, Data Science, and the Shift to Edge Computing | 2 | 4 | 1 | 1 | Copeland prompts the guests on emerging trends in computer science. Levine highlights deep learning and edge computing, while Franklin clarifies the broad scope of data science across university departments. | |
| Human-in-the-Loop Analytics and the AMP Vision | 2 | 4 | 0 | 0 | Copeland asks about the human element in AMPLab's vision. Franklin details the shift from crowdsourced tasking like CrowdDB to supporting human decision-makers in analytical loops. | |
| Tachyon Nexus and In-Memory Storage Architecture | 3 | 5 | 2 | 2 | Copeland introduces guest H.Y. Li to discuss Tachyon Nexus. Levine walks through a contrarian investment thesis based on memory hierarchy flattening and removing cloud assumptions. | |
| Transitioning from Academic Researcher to Startup CEO | 3 | 5 | 1 | 2 | Copeland questions Li and Levine about transition hurdles from student researcher to founder. Levine explains the psychological shift required to lead through others rather than personally writing code. |
Statements from this episode (13)
UC Berkeley's AMPLab Is the Birthplace of Apache Spark
“The place where Apache Spark was born, UC Berkeley's Amplab has not just created a major open source software platform, it's spun out more than its share of groundbreaking companies.”
Levine: AMPLab spinouts are far more innovative than Cisco or Oracle
“The companies coming out of there are way, way different than what I found, you know, from people coming out of, let's say, Oracle or Cisco.”
Levine: Code authorship matters more than IP ownership in open source
“IP ownership is, to me, less important than who actually wrote the code”
Levine: Commercial software forks will never rival original code inventors
“The fork company is never going to be as strong as the one, ah, who has invented it.”
Levine: Commercial tech innovation has shifted from corporate labs to universities
“If you look at, ah, you know, a lot of, ah, innovation Up till, well, to me, re up until recently came from, like, Bell Labs and IBM and Microsoft Research, and that's where a lot of the new, I would say the new commercial ideas came from folks leaving those a…”
Levine: Computing and analytics will shift from the cloud to edge devices
“And I think the Internet of Things and drones and however you want to define these, this sort of range of computers that exist not so much in our pocket but out there in the world, I believe that there's going to be computing that will be fundamental to how we…”
Franklin: Human-in-the-loop analytics solve problems beyond current machine learning capabilities
“The idea was you wanted to be able to bring in people to solve those parts of these, you know, machine learning problems that the machine learning wasn't quite up, up to speed for.”
Li: Memory-centric computing is the future of infrastructure architecture
“Memory-centric computing is the future, and you have compute layer of the storage layer, and we want to be the storage layer.”
Levine: Spinning hard disk drives will be completely phased out
“Spinning disks are going away, okay?”
Levine: a16z targets investments in projects that consensus views as crazy
“When more people disagree, that's the project we want to go invest in. Because, like, you know, it's too crazy for, you know, hell, that'll never work.”
Levine: Find future opportunities by asking what replaces currently dominant technologies
“My whole philosophy around this is I take something that's very popular right now, cloud computing, and I say, what happens when cloud computing doesn't exist anymore? What fills that place?”
Levine: Transitioning from individual contributor to manager is the hardest career pivot
“I think one of the most difficult transitions a person makes in their career, whether it's from academia or from just an individual, is from individual contributor to being a manager and leader.”
Levine: Delaying open-source business models makes future monetization significantly harder
“If you don't have that early, what starts, and you don't know what that is, then potential customers and partners don't know what you're going to charge for in the future and all that, and it makes it, the longer you go on without having a model, the harder it…”