Tim Beiko is part of the Protocol Team at the Ethereum Foundation (https://ethereum.org/en/foundation). In this role Tim orchestrates development teams focused on the core protocol—most notably for The Merge. In this episode we discuss Tim’s origin story in crypto, how Ethereum can serve as an economic testnet for new types of coordination in the world (and why that’s a valuable proposition), what he’s most excited about for the future of Ethereum, his advice for developers new to the crypto space, and much more.

Recorded Thursday August 31, 2023.

And I think if we get that, this is sort of how I think Ethereum ends up being tolerated in the world.

I think there's a lot of the core ethos around blockchain, such as fundamentally at odds with how corporations, how nation states view themselves.

Just the idea that you have money that's not tied to any like central issuing entity that people can send freely all over the world is like a pretty contentious thing.

And so I think if we want that part of blockchains, we're like, if we believe that having, you know, economies that are kind of untethered from geographical and nation state boundaries, that's great and we should do it.

But we should also think about what do we offer back to the world such that they tolerate this and they don't just see it as purity, negative, some for them.

Hey everybody, Tanner here with Wagner Ventures.

On today's episode, we have Tim Bacow, part of the protocol team at the Ethereum Foundation.

For anyone who's new, this is the Wagley Ventures podcast, where we do snapshots with interesting builders and founders from across Web3.

Check out wagleyventures.io to learn more about the syndicate behind the podcast.

But for now, let's get into it with Tim at the Ethereum Foundation.

All right, hey everybody, I'm here today with Tim Bacow, part of the protocol team at the Ethereum Foundation.

You know, I think before we get started with some of the bigger conversation, I'd love to just get a sense for your story before we really dive in.

And I was thinking maybe 1 way of getting at that is to talk a little bit about your work before the Ethereum Foundation.

So you were at Element AI and then kind of made your way to ConsenSys and ultimately the Ethereum Foundation.

I think my question was going to be, you know, what did your thinking at each point look like?

How did you end up gravitating to the opportunities that brought you to the work you're doing now?

Sure, yeah, that's definitely a good place to start.

So I've been following crypto for a while.

So I think I first heard about Bitcoin in 2013, 2014.

I heard about Ethereum when the DAO happened, but when the DAO was the project and not the hack.

I sort of knew about Bitcoin and then I'd heard about Ethereum, but the DAO was really like the first big application on Ethereum.

For people who aren't familiar, it was effectively like a VC fund that was organized as a DAO.

So if you bought tokens in the DAO, you could have a vote on which companies the fund invests in.

And so that's kind of what got me interested.

So I bought some Ether and then I bought some DAO tokens.

And then about a week or 2 after I did that, the DAO got hacked on Ethereum.

So this was the largest hack at the time.

I think it had raised something like 15% of all ETH that was circulating and somebody just managed to hack the whole thing.

And so that was really kind of a forced dive down the rabbit hole of like, okay, I have these tokens and this thing that got hacked, what happens with that?

And then after the DAO hack, there was like this whole resolution with a hard fork.

And then that led to the creation of Ethereum Classic.

So then you had all your coins being split on 2 chains.

So that was like kind of a forcing function to like, actually get a bit deeper than I probably otherwise would have.

And then after the hack, for about 6 months or maybe a bit more, I wasn't sure Ethereum was actually that good of an idea because you had a smart contract platform, but then if you couldn't write safe smart contracts, was there actually a ton of value in having that?

So I kept following what was happening, but not as closely for a while.

And in the time I was interested in AI, so I just was studying and ended up working in that as well.

But I'd say late 2016, early 2017, you start to see a bunch of new projects also launch on Ethereum.

And by mid-2017, there was this huge ICO boom.

And I think at that point, it became pretty clear to me that there would be demand for Ethereum, even if Every project so far failed, got hacked, was a scam, whatever.

Clearly there was demand to use the protocol itself.

And that got me really interested in wanting to work on the protocol directly.

So I think I wasn't confident enough to go work on any of the applications being built on Ethereum.

But I definitely wanted to get involved more in the protocol, just because as a user, you could tell that things were a bit fragile and clunky and whatnot.

And if we could improve this, then we could probably unlock way more demand.

But at the time, I was a product manager with a background in AI.

And so not only did I didn't have any experience working on blockchain stuff, I also wasn't a researcher or a software engineer.

So it took some time to find a job that worked with my background, with the stuff I wanted to work on.

And I got kind of lucky in a sense where in 2018, ConsenSys started to put together a protocol team and they were basically the first ones to hire product managers for that.

So historically before that, I think most protocol teams were just like a bunch of researchers or a bunch of engineers.

I didn't have like the structure of like a traditional, I don't know, tech startup, like team company, team structure.

So they hired me as a product manager And then that's how I got to like start working full time on Ethereum.

I was working on their Ethereum client, BASU there.

So I did that for 2 and a half years, got to understand how, you know, the protocol works, how like the software that runs the protocol works and, And the way that changes to the protocol are made in Ethereum, there are these public calls where all the different companies and entities working on Ethereum get together and chat about potential changes.

So there was somebody running them for probably 5 years when I joined ConsenSys called Hudson And in about 2020 or so, he decided he wanted to step down.

And so asked me if I'd want to take over for him running the calls.

So instead of being on 1 of the implementations of the protocols, I sort of helped coordinate all these different teams and get them to agree on what changes we make to Ethereum and how we make those changes.

So I joined the EF in late 2020, early 2021.

And that's basically what I've been doing since.

So now I help coordinate all these different teams working on Ethereum and just think a bit more broadly about like, what does it take to get a good ecosystem of contributors to the protocol?

There's elements of that I want to call out at different points in this conversation to kind of revisit.

But, you know, 1 of those things that I think maybe you may be best known for, well known for is the merge and your role in the merge.

And we're coming up here on the kind of 1 year anniversary.

And so I wanted to maybe do a little bit of a retrospective question here, which is, you know, the merge was really like the Super Bowl for crypto, right?

This transition from proof of from POW to POS.

So like, you know, everyone's tuned in to watch how it went.

That's all everyone talked about for a while.

And as 1 of the lead engineers for the merge, kind of coordinating a lot of that activity.

How do you reflect on that experience here almost a year after it was executed?

Yeah, probably the weirdest thing is that when we do changes to Ethereum, we really have to get them right.

Like, you know, we can't break the chain, but then they sort of work forever.

And we very quickly go from this mode where like, you know, before the merge goes live, it's like thinking about every single potential edge case and, you know, loads and loads of testing and will this thing ever work?

But then it's like, as soon as it's happened and it works, we, we kind of take it for granted in a way because it just runs on the network by itself.

And now, for example, you know, we run, we we're working on like, it's been already another upgrade since the merge and we're working on the 1 after that.

And so every time we stand up a new test net, a new dev net or whatever, like we assume the merge has happened.

We run the transition on it if we have to or something like that.

But it's like now it's just like almost a dependency of the stack.

It's pretty crazy that you spend all this upfront time building this thing and making sure it really works, but then it sort of chugs along kind of forever and doesn't really need any updates.

And that's very different than a lot of consumer tech or web software that gets built.

So it's almost a bit closer to building a physical thing, like a car or a plane, right?

Where you might spend a ton of time designing this engine or designing the entire plane or whatever.

But then the plane just flies all the time, and it's fine.

So yeah, I'd say that's probably the biggest thing, where the contrast between spending so much time and effort and mental cycles thinking about all the possible edge cases and how is this thing gonna work and is it even gonna work?

But then a year later, we basically fully take it for granted and it just like works in the background and no one's even thinking about it.

Where all of this work happens and then life continues, right?

I'm curious, just kind of looking, you know, that's a great look back.

I'm curious, a little bit of a look forward where, you know, again, this the merge was just such a topic of conversation, it was so important for the space.

And there's been significant work from your teams even since then.

So I don't want to minimize any of that by fixating on the merge by any means.

But what are you thinking about these days that might capture everyone's attention to the extent that the merge did?

So I don't know if we'll be able to top that because we sort of advertised the merge for what, like 6 years or something, and then it eventually happened.

But I think right now, so the thing the Merge gave us is it gave us definitely better security of the network in that we can now punish validators rather than just have incentives for miners.

It didn't quite help for scalability directly, but it gave us the scaffolding or the infrastructure that we then need to build better scalability solutions.

And I think that's really where a lot of our attention has shifted now.

So you've obviously probably heard about layer twos and all of that.

And that's really the strategy that Ethereum is thinking about with regards to scaling.

So you just can't have like a single chain that processes all the transactions for everyone in the world if you want the nodes on that chain to be able to be run with like relatively cheap hardware and bandwidth access, right?

Like, you know, I have a MacBook in front of me and I'm in my apartment.

If I want to run my Ethereum node on my MacBook and with my residential internet connection, there's no way I could process millions or even hundreds of thousands of transactions every second.

What Ethereum's plan is around that, still offers scalability to users, is we have this idea of layer twos, which are basically separate blockchains that run all their transactions outside of the main chain.

But then they post the data back for those transactions on Ethereum layer 1 without actually running the computation.

So they just post back all the data and they can do some fancy things around like compressing that data to get some more scalability gains.

But because it's basically quite expensive to run computations on Ethereum, but cheaper to post data there, you can get much more transactions using these Layer 2s.

And so the big thing we're focused on now in terms of the protocols is getting better sort of data storage infrastructure for these Layer 2s.

And what this means in practice is that today, if you post data to the Ethereum network, it sort of gets stored forever on the network.

And because of that, we need to price it quite high because you're imposing this cost on the network forever.

But layer 2s don't actually need the data to be stored forever.

They only need it to be stored long enough that if there was an issue on the layer 2, we could go back to L1, figure out what happened, and resolve it and move on from there.

And so if you're familiar with say like optimistic rollups, they assume already that like within 7 days, if you're like initiating your withdrawal from the L2, within 7 days, if something had gone wrong, somebody would have flagged it and If not, then you're good to go.

So you sort of go from a world where like Ethereum currently stores data forever to L2s needed for about a week or so.

And so this is the main thing we're working on now is introducing like temporary data storage on Ethereum.

And this is what people call data availability.

And so when you hear data availability, the idea is just that these networks can provide data storage that's still secure, but because it's not stored forever, you can charge much less for it.

And then those costs savings get passed on to the users of layer twos.

Because today, like about 95 to 99% of an L2 transaction is just posting the data back to L1.

So the cheaper we can make that, then the more scalability and the lower the cost we can get for L2.

So this might not be something that's like as flashy as the merge.

And it's also not something that will be such a like phase transition, like going from 1 thing to another.

But it's a type of thing that can actually help Ethereum, you know, onboard orders of magnitude more users.

So, you know, you're in a unique position where I feel like I can ask this question confidently, where, you know, you're entrusted with bringing to life this kind of ever unfolding vision of what Ethereum could and should be, right?

We're talking years and years of hard work to really progress each component of the vision.

And so my question is, how do you think about patience with respect to bringing about a potential vision where, you know, really kind of this balancing act between not losing morale because it's taking so long, but also not having too much enthusiasm maybe because still it's gonna take a while to accomplish each part.

How do you sort of manage that with your team and think about it yourself?

I think there's some aspect where you have to be like almost overly optimistic and like, you know, otherwise you just wouldn't do the thing.

And this was true of the merge, but I think it's true of like everything.

And it's almost like a self-fulfilling prophecy.

Like if you think you can do this thing in a year, it'll probably take 2 years.

But if you thought it would take 2 years, it might take 4 years.

So I, yeah, I don't know of like a great way to do this, except Like, I think I spend not necessarily a lot of time on this, but like, I definitely spend some time like when new people come in and like, say, want to do a big change to Ethereum, trying to like get their expectations to somewhere that's reasonable, probably getting their, moving their expectations from like being overly optimistic to maybe just optimistic so that they're not disappointed or disheartened enough that they won't do the work, but they're also not expecting to change all of Ethereum in 6 months.

I think the other part that we've gotten much better at in the past couple of years is building better testing infrastructure.

Actually, yeah, that's probably 1 of the other parts.

So having better infrastructure and being able to more confidently test stuff helps speed things along a little bit because we can just reduce uncertainty quicker.

So that's been a huge other thing where even though there's multi-year efforts, you break it up into smaller chunks and you can have like high confidence in each of those chunks moving forward and sort of, you know, get momentum that way.

But yeah, there's like no way around it, right?

Like I think, again, if you take the example of like, building physical things rather than building software, like if you're working on like a new airplane or like, you know, a new iPhone model or whatever, like these things just take time.

And you sort of have to be okay with that at some level.

You mentioned like in the scenario where someone comes in, wants to change Ethereum in some particular way.

I was curious, what does it look like for an idea to become significant enough within the Ethereum Foundation to warrant further exploration or dedication of engineering time?

The prioritization conversation, basically.

How does that, what does that look like from

And I, maybe I'll start by saying like the EF, the Ethereum foundation doesn't get like a, a veto, right.

On like what gets worked on or what gets done in Ethereum, right?

Like we help, we build some common infrastructure.

Like I help, you know, coordinate the calls where people have these debates, but I don't get to choose what goes in the upgrade.

It's not like I'm the sort of, think of like a traditional tech company where I'm the product lead and I'm like, okay, we're doing X, Y, and Z.

It's like the best we can do is get all the people around the table or on a Zoom call and ask them what they think we should do and help guide that conversation.

And similarly, in terms of just the software, the best we can do is we can write some testing infrastructure.

We can obviously do some research, but we can't necessarily make everybody go 1 way.

So what the process looks like is there is a formal process of you can write the specification for your change, get it reviewed, and have the different implementation teams give you feedback.

And if they think it's valuable enough and good enough, then eventually it might get deployed on mainnet.

And I think how complex it is to get something prioritized or get something done depends on how big of a change it is.

Because as Ethereum grows and has more value, has more users depending on it, it should be hard to change it.

If you're using Ethereum and you have your life savings on it or you've built a business on it, then the platform shouldn't just change in a completely random way under you tomorrow.

So if you have a big change, then you should expect to spend years not only working on the technical parts of it, but trying to understand the impact that it has on people and ideally mitigating that impact.

And if you absolutely can't mitigate the impact, you're going to need a pretty strong case for why your thing is so beneficial that it's worth impacting others.

And that process is sort of like more social slash political process.

But you can think of it as like going to all the relevant stakeholders, trying to get them on board and being able to articulate, you know, that this has happened and, you know, get like buy-in from those people.

And maybe 1 example of a medium-sized change recently that was done this way is there's a new EVM opcode.

So EVM is the computer, the virtual machine that runs on Ethereum.

There's a new operation that's being added in the next few months.

And this is something that was proposed years ago.

And probably a year or 2 ago, someone from Uniswap thought that this old idea could actually be quite beneficial to them.

And so Uniswap spent a bunch of time in 2021, 2022, trying to get this change in.

It was obviously beneficial for them, but they had to go and figure out, is it beneficial for all the other applications on Ethereum?

Are there applications that are harmed by this?

Is this something that the languages like Solidity and Vypr can support?

Is this something that's safe and whatnot?

So they went through this entire process and now it's scheduled to go live in the next couple of months.

I think it's, yeah, it's a good example of like a team that came in, took something that was already proposed that they saw value in, and, you know, built, like support for it and like addressed a lot of the concerns around it.

But and that's the, took on the order of like 1 or 2 years.

So just to, you know, put it in perspective, yeah, for anyone listening who might want to try this, it's not something that's impossible, but it's something like you need to be willing to put time and effort in.

So, you know, 1 question I have for you is really maybe an ideological question at its root where, you know, I think you're several years into this work with Ethereum, right?

And I think my question would be, you know, what about the vision of Ethereum at this point, maybe as distinct from, you know, some of what drew you to it initially, like what keeps you, what's What keeps you still showing up?

What keeps you most excited these days about the vision and what it's really about?

So I'd say 1 thing that's changed, it doesn't feel as broken anymore.

I feel like a lot of the stuff that we worked on in, call it like 2018 to 2022, like, you know, say with the merge, there was just like a lot of like obviously wrong things with Ethereum.

So The first big thing I worked on before the merge was EIP-1559.

So this changed the transaction fee market.

And the way the transaction fee markets worked on Ethereum before just felt so broken to me.

If you explained how Ethereum works to somebody who knows nothing about blockchain and just understands maybe a bit how consumer products or platforms work, they would intuit that it works basically the way that it does today.

And the previous version just made no sense.

But it was much simpler in terms of architecture and whatnot.

And then similarly, I think when you talk to a lot of the proof of stake researchers and designers, they sort of felt similarly about proof of work, right?

They just looked at it and they were like, man, this is so broken.

And proof of stake is so obviously better, but there's all this work we need to do to actually make it production ready.

So I'd say in terms of just basic functionality, like meeting expectations, Ethereum is in a much better spot than it was 4 or 5 years ago.

And I think the thing looking forward is just, if you look at like the impact and the scale that it's had, I'd say it's still, you know, kind of minimal, right?

Like there's not like millions or hundreds of millions of people using this.

I think most of the use cases are still sort of use cases that are kind of within the crypto community and not that those are necessarily bad.

But I think the thing that I hope we can accomplish is scaling this to a point where there is a better interface or bridge between what happens on Ethereum and what happens outside of it and the rest of the world, the sort of, you know, the like very short version of this is I think in the longterm, Ethereum can be something like an economic test net for the world, right?

Where we have this network, which stores value, has like, you know, organizations on it, a bunch of different assets on it as well.

And we can run, you know, economic experiments that for whatever reason, you sort of can't run in the real world, because there's just much higher at stake.

And you look at stuff like, you know, quadratic funding being like a great example of that, get going pioneered.

This is the type of thing where it's like, you can write an economic paper to prove that it's just better than other mechanisms, but no government is going to, you know, switch everything to quadratic funding overnight because people's food and shelter depends on this stuff.

But if we can run it on Ethereum and say, hey, look, we've allocated tens of millions, hundreds of millions, billions of dollars using this mechanism, and it's better than whatever other allocation mechanism exists, then you can actually start to get people to pay attention and think, oh, maybe this actually works.

I think there's a lot of the core ethos around blockchains that's just fundamentally at odds with how corporations, how nation states view themselves.

Just the idea that you have money that's not tied to any central issuing entity that people can send freely all over the world is like a pretty contentious thing.

And so I think if we want that part of blockchains, we're like, if we believe that having, you know, economies that are kind of untethered from geographical and nation state boundaries, That's great and we should do it, but we should also think about what do we offer back to the world such that they tolerate this and they don't just see it as purity, negative sum for them.

And yeah, I think that the best thing we have to offer is being a testbed for economic and coordination experiments that we can run at reasonable scale with relatively low risk, and hopefully prove that there are ways we can improve this and export those into nation-states and physical worlds whenever they're ready.

So, you know, this may be a little bit of a difficult question to frame the exact right way, but I'd like to try where, you know, on this podcast, for example, we've heard several founders in the crypto space basically make these offhand remarks about how the future of the space isn't what was maybe originally envisioned, this idea of decentralize all the things and everything will be better, freer.

And a part of me has been a little tiny bit disappointed hearing things like that, just because it sounds like the death knell of some of what has kept crypto idealistic and optimistic and principled, if that's even the right word.

See all WAGMI Ventures Podcast transcripts on Applepodcasts

Ethereum as The World’s Economic Testnet, with Tim Beiko (Ethereum Foundation)