Barbara Liskov

1. Key Themes

From Benign to Byzantine: The Leap That Enabled Blockchain

Liskov's research trajectory moved from handling simple, "benign" failures (machines either run or go silent) in ViewStamp Replication to handling actively malicious, lying nodes in PBFT. This leap required moving from 2F+1 to 3F+1 replicas and adding cryptographic certificates — and it is precisely this harder problem that blockchains needed solved.

"We only worried about the simple failures, which were in fact the main failures that happened at that time because it was just the days of the ARPANET and we were all friends. And we didn't have the malicious attacks that showed up in the 90s once we had an internet." [00:15:50]

State Machine Replication as the Hidden Substrate of All Blockchains

Turing-complete blockchains like Ethereum and Solana are, in the most literal sense, implementations of the general State Machine Replication (SMR) problem first articulated by Liskov's group in the 1980s. Smart contracts are simply special-case applications running inside a general SMR protocol.

"Turing Complete blockchain protocols, like Ethereum, Solana, et cetera, they're almost like literal implementations of the fully general state machine replication problem... These blockchain protocols are in some sense one of the more literal embodiments of the general state machine replication problem that we've seen today." [00:31:08] — Tim Ruffgarten

DARPA as a Civilizational Lever: Government Funding Created the Internet Stack

PBFT — now foundational to almost every major blockchain consensus mechanism — only happened because DARPA issued an RFP on malicious attack tolerance. Without that signal and funding, the work might not have been done when it was.

"Between DARPA and NSF, I mean, they were the reasons that research was getting funded all those years. They are one of the very important backbones of why we have the Internet that we have today. Without that research funding, it would have been much more difficult to have done all that work." [00:22:47] — Barbara Liskov

The Decade-Long Lag Between Research and Adoption

ViewStamp Replication was published in the late 1980s, yet waited roughly ten years before commercial systems used it. PBFT similarly preceded the blockchain era by over a decade. Deep foundational research consistently has long but enormously impactful adoption curves.

"For view stamp replication, it was a delay of about 10 years before people started to use it. And then along came blockchains. And, well, that was very funny." [00:31:46] — Barbara Liskov

ViewStamp Replication and Paxos: Independent Parallel Discovery of the Same Protocol

Liskov's ViewStamp Replication and Leslie Lamport's Paxos were independently developed and were essentially identical protocols. Neither group realized this for years — it was only discovered when a former Liskov student at Google noticed it while reviewing the Google File System paper.

"The story I heard was that Bill Weil, who had been a student of mine in the 80s, happened to be at Google. And he was looking at what was going on in the Google File System. And he said, oh, he said, that's View Stamp Replication. And they really were the same protocol developed in two different places." [00:20:38] — Barbara Liskov

Modularity as Universal First Principle — From Programs to Theorems to Blockchains

Liskov's conviction that modularity is the master organizing principle extended far beyond programming languages — she draws a direct parallel to mathematical proofs (lemmas as modules), and Tim Ruffgarten independently arrived at object-oriented programming as the mental model for explaining blockchain execution to undergraduates.

"Modularity is everything in building large programs... In a theorem, you start off with a statement of what is supposed to be true... you have lemmas and the lemmas each have a statement of what they are. And then you prove their correctness independently... this is exactly what's going on in a modular program." [00:07:28] — Barbara Liskov

AI Is Reshaping the Role of the Programmer — Verification and Abstraction Become the Core Skill

Liskov argues that as AI writes routine code, the value shifts entirely to higher-level design, specification, and verification — precisely the skills she built her career teaching at MIT since the late 1970s. She cites an op-ed by Mary Shaw as crystallizing this shift.

"The students really need to understand how to write those programs themselves or otherwise how will they know that AI did them right or wrong... You're now managing more of the coding and you're working at a higher level. That's a good job." [00:34:09] — Barbara Liskov

The Separation of Consensus and Execution Is a 40-Year-Old Idea

The modern blockchain architectural distinction between a consensus layer and an execution layer (e.g., Ethereum's separation of the beacon chain from the EVM) was already conceptually embedded in ViewStamp Replication and PBFT from the beginning — Liskov's group explicitly designed the replication protocol to be application-agnostic.

"We didn't really care. We just wrote into the ledger what the operation was that was being requested, but we were not in the least bit interested in what it meant to execute that operation. So, it just seemed like a good separation of concerns." [00:29:28] — Barbara Liskov

2. Contrarian Perspectives

The Theoretical Computer Science Community Undervalued Its Own Breakthroughs

Work on Byzantine fault tolerance had existed in theoretical CS for years before PBFT, but the broader community treated it as academically interesting but practically irrelevant. It took an empirical systems paper with working benchmarks to change that perception — not the theorems themselves.

"Sometimes, good systems work that have benchmarks that show good actual practical results, they change people's mind. They say, oh, this technology can actually be used. There's a difference between having a theoretical paper saying it's possible and seeing good benchmarks." [00:27:23] — Itai Abraham

Most Important Innovations Come From Crossing Fields, Not Deepening Within One

PBFT was born from Liskov's student scanning DARPA RFPs outside of academic computer science, not from an internal research agenda. ViewStamp Replication imported transactions from database systems. Cryptographic certificates came from theory. The breakthroughs consistently came from boundary-crossing.

"I suggested to Miguel that he look at the RFPs that DARPA had put out and see if anything interested him. And he found an RFP that was looking for ways to handle the malicious attacks that were going on on the Internet." [00:21:54] — Barbara Liskov

CS Education May Be Training People for Jobs That No Longer Exist

Liskov is openly worried — not bullish — about whether a computer science degree still makes sense for undergraduates, given AI's ability to write routine code. She sees the field entering a genuinely disorienting transition, not a simple upgrade.

"I am a little worried about computer science as a field for young people going to college and should they major in computer science and so forth. And I don't understand the limits of what AI can do." [00:33:40] — Barbara Liskov

Bitcoin's Novel Consensus Design Was Actually a Detour, Not the Foundation

Bitcoin's consensus mechanism looked very different from PBFT and the decades of distributed systems work that preceded it. The blockchain community spent years not realizing that PBFT-family protocols were the right tool for the problems they were trying to solve.

"Blockchains were sort of launched with Bitcoin protocol, which actually looks rather different than PBFT and most of the other consensus protocols anyone had thought about to that time. And it took, I think, the blockchain community a number of years to realize that PBFT and the sort of subsequent protocols in that family were exactly the right tool for a lot of the problems that they're trying to solve." [00:32:10] — Tim Ruffgarten

3. Companies Identified

Google

Large-scale internet company and operator of the Google File System. Mentioned because it independently used ViewStamp Replication-equivalent techniques (Paxos) in the Google File System, and a former Liskov student at Google identified the equivalence of the two protocols.

"The Google File System paper was published. And that paper used replication and talked about using Paxos... Bill Weil, who had been a student of mine in the 80s, happened to be at Google. And he was looking at what was going on in the Google File System. And he said, oh, that's View Stamp Replication." [00:20:10] — Barbara Liskov

4. People Identified

Turing Award–winning computer scientist at MIT. Pioneer of data abstraction (CLU language), distributed systems (Argus), ViewStamp Replication, the Liskov Substitution Principle, and Practical Byzantine Fault Tolerance (PBFT). Her work is the direct intellectual foundation of modern blockchain consensus protocols.

"We came up with a protocol that if the primary seemed to not be doing its job, the backups then carried out another protocol in which a different replica became the primary." [00:00:25] — Barbara Liskov

Miguel Castro

PhD student under Liskov at MIT. Co-inventor of PBFT. Identified the DARPA RFP on malicious attack tolerance that launched the PBFT research program.

"I had a student, Miguel Castro, who was looking for a PhD thesis... he came to me and he said, why don't we see whether we can figure out a way to do replication that handles these malicious attacks?" [00:00:00] — Barbara Liskov

Brian Oakey

PhD student under Liskov at MIT in the mid-1980s. Co-developed ViewStamp Replication as his thesis project.

"I had a student, Brian Oakey, who was looking for a thesis topic. And so the idea of a replicated file system seemed like a really interesting thesis topic. This is in the mid-80s." [00:10:05] — Barbara Liskov

Leslie Lamport

Distributed systems theorist and creator of Paxos. Independently developed the same protocol as ViewStamp Replication without either group realizing it for years.

"Leslie Lamport was working on Paxos. And I actually heard him give a talk in the 80s about Paxos, and I didn't understand what he was talking about. And I had certainly no idea it was the same thing. But this was a mutual lack of understanding." [00:19:43] — Barbara Liskov

Bill Weil

Former Liskov student who ended up at Google. The individual who finally identified that ViewStamp Replication and Paxos were the same protocol upon reviewing the Google File System.

"Bill Weil, who had been a student of mine in the 80s, happened to be at Google. And he was looking at what was going on in the Google File System. And he said, oh, that's View Stamp Replication." [00:20:38] — Barbara Liskov

Bob Kahn

Internet pioneer. Mentioned as the author of the paper that inspired Liskov to pivot into distributed computing in the early 1980s.

"I read a paper by Bob Kahn in which he talked about the dream of distributed computing and how there would be distributed programs that had components at different nodes in a network and communicated over the network. And only nobody knew how to build them. And so I thought, there's a great problem." [00:04:02] — Barbara Liskov

Ron Rivest

Co-inventor of RSA cryptography, MIT colleague of Liskov. Cited as an important part of the MIT intellectual milieu that made cryptographic tools available to systems researchers.

"I was fortunate to be at a place where Ron Rivest was and his colleagues, you know, and so forth." [00:17:21] — Barbara Liskov

Butler Lampson

Pioneer systems researcher. Mentioned as representative of the unified systems research community at SOSP in the early era, alongside top database researchers.

"The people you think of in the systems area like Butler Lampson and so forth were there. But so was Jim Gray. So was Bruce Lindsay. The top database people were all there." [00:10:59] — Barbara Liskov

Jim Gray

Database pioneer. Mentioned as a participant in the unified early SOSP conference community, illustrating how tightly coupled systems and database research were.

"So was Jim Gray. So was Bruce Lindsay. Well, the top database people were all there." [00:10:59] — Barbara Liskov

Mary Shaw

Carnegie Mellon computer scientist. Co-authored a New York Times letter arguing that programmers still need to understand how to write code themselves in order to evaluate AI-generated code — cited by Liskov as crystallizing the challenge for future CS education.

"There was a letter in the New York Times recently by Mary Shaw and somebody else about what's the future for coders that made the point that the students really need to understand how to write those programs themselves or otherwise how will they know that AI did them right or wrong." [00:34:09] — Barbara Liskov

John Guttag

MIT colleague of Liskov. Co-developed the foundational MIT course on building large programs — focused on design, modularity, specification, and verification — starting in the late 1970s. That course is now arguably a template for the future of CS education in the AI era.

"The course that I developed at MIT in the 19, starting in the late 70s with John Guttag, which was a course about how do you build big programs? And it was all about design and modularity and specifications and verification." [00:34:39] — Barbara Liskov

5. Operating Insights

Scan RFPs and External Problem Statements Deliberately as an Innovation Input

PBFT — one of the most consequential distributed systems papers ever written — only happened because Liskov told a PhD student to scan DARPA's open solicitations for interesting problems. Deliberately scanning government, corporate, or open-source problem statements (not just following internal roadmaps) is a systematic way to surface important, well-resourced problems before they become obvious to competitors.

"I suggested to Miguel that he look at the RFPs that DARPA had put out and see if anything interested him. And he found an RFP that was looking for ways to handle the malicious attacks that were going on on the Internet." [00:21:54] — Barbara Liskov

Understand What You Don't Understand — Incompleteness Is the Research Agenda

Liskov's explicit operating principle for her research group was to identify precisely where their understanding was incomplete and treat that gap as the next problem to solve. This is directly transferable to product and company building: map your own blind spots as rigorously as you map your strengths.

"It's also very important in research to understand what you don't understand. I always tell my students that's where you get the insight into where you have to move. And you don't want to have an incomplete understanding of why something works. You have to really understand it completely." [00:18:27] — Barbara Liskov

Stepping Stones Compound: Prior Proprietary Knowledge Builds Durable Advantage

Liskov's group was able to develop PBFT faster than anyone else precisely because they had deep, internalized knowledge of ViewStamp Replication that was not widely understood outside MIT. Proprietary mastery of a prior-generation technology is a genuine, underrated competitive moat when the next hard problem arrives.

"We simply started from view stamp replication. My group all knew what view stamp replication was. And I don't think it was understood at other places. And so we had that stepping stone that other people didn't have, and we used it." [00:23:12] — Barbara Liskov

6. Overlooked Insights

Accountability Protocols May Be the Next Major Primitive in Distributed Systems and Crypto

Tim Ruffgarten briefly mentioned a body of work from the last five to seven years on "accountability" in Byzantine fault-tolerant systems — the ability to cryptographically identify and prove which specific nodes misbehaved even after a safety violation has occurred. This is distinct from just preventing violations. Liskov said she was unaware of it. For investors and builders, this is a significant signal: a research area that even the field's founding figure hasn't tracked is early, and it has direct applications to slashing mechanisms, validator accountability, and protocol forensics in live blockchain networks — areas where there is currently no dominant solution.

"Accountability says the only way to create a consistency violation is by signing lots of inconsistent things. And then through inspection, you can actually identify bad actors who double signed on different conflicting states... analyzing PBFT-style protocols, not just in the regime where the number of faulty nodes is small, but also where it's actually bigger than the threshold." [00:27:55] — Tim Ruffgarten

The Verification Tools Gap Is an Enormous, Underserved Market Being Created by AI Code Generation

Liskov flagged almost in passing that AI-generated code will be frequently wrong, and that what will be urgently needed is a new generation of formal verification tools to check that AI output is correct. She frames this as a research area, but it is equally a product and investment opportunity: as AI coding agents proliferate, the tooling layer for verifying correctness of AI-generated code at scale does not yet exist in a mature commercial form. The person or company that builds the "spell-check for correctness" for AI-generated code could capture enormous value across every software-producing industry.

"I can easily see how AI can write a little program for you if you give it a specification. I can easily see how that program might be wrong. And it's very easy to see that we're going to need a lot of verification tools in order to make sure that program is right." [00:33:40] — Barbara Liskov