Devansh Mehta

Unblocking Pilots in AI x Public Goods/Governance @ethereumfndn |

I don't speak for the collective, unhinged tweets my own

[Devansh Mehta @devanshmehta]

Very thankful to @owocki to be invited as a cohost for the annual "state of public goods" discussion with @VitalikButerin

it was honestly a test of my mental fortitude & stamina to be top of my game for all 1.5 hours 😅

I've scratched out 5 key points from the long episode;

1. what should we ask PGF funders to rally behind?

A norm of 20% to public goods and 80% to commercial is inherently unsustainable. Because competitors will do 90-10, then 100-0, which will eventually beat out the market

Quadratic funding should be seen in this light, of a global tax to the public goods god distributed via matching funds. Instead, a norm of funding your own dependencies is more sustainable and what we should be moving towards

lawful & good - QF

chaotic & evil - dependency funding

2. what should PGF builders work on over the next year?

a. structured & well thought out proposals for how we fund ethereum open source software beyond just block rewards

b. productizing deep funding & making it easy for any ecosystem to generate their dependency graph, ascertain relative value between them & finally stream funding based on it

for example, vitalik said he finally crossed the threshold this year where over half his onchain transactions are private. he similarly hopes we can eventually make it as easy for any project to fund its own dependencies

3. When asked about the 3rd prong of accountabililty, vitalik pulled his classic switcheroo trick - putting the Q back on us

I shared my experience with deep funding where maintainers wanted their repo to participate for not just funding but also the thoughtful feedback which spot checkers provided in their evaluations

So we should think of accountability not as carrots & sticks but more as motivation, even though many of us get paid for our work when we have a compatriot tracking our progress it automatically makes us work harder to showcase our best

in this light, airdropping money to repos is much worse than streaming money to them based on the fact that they are in your dependency graph, along with comments made by any spot checker on the value their repo provides

4. He sees a short window of time to create a software license with a sliding scale where derivative work that becomes proprietary pays a fee (while it doesn't if it remains open)

He even said he would be happy for some of the software he personally writes to be under such a license! But it is important to get the terms right as they are hard to change once it reaches escape velocity

5. Finally, why is public goods important & why should builders spend their time working on it?

At a conceptual level, public goods innovation is key to maintaining the ethos of the crypto space in being open & transparent, since we forego commercialization opportunities that web2 has done by keeping their IP close to the chest

so instead of building the 101st stablecoin, there is much more leverage in figuring out designs that can sustainably govern the ethereum commons. he also advised new builders in the space that their radical ideas would get more traction with new projects that don't have power brokers who would lose by a change to the system compared to trying to reform existing big ones

3 other bonus points from the episode for those that have stayed to the end;

- in a deep funding world where projects support their dependencies, there is a natural evolution function where if a repo stops innovating, lesser projects use it, which means lesser money coming into it.

- if we decide to allocate fees to some institution or mechanism for passing on to projects, a key issue becomes ensuring they don't get corrupted or experience purpose drift, like funding preservation of a 3000 year old language when its actual purpose is open source software

even in the case of funding mechanisms, how to programmatically figure out in code what is a mechanism vs an institution is hard. preventing the free rider problem with solutions such as "if 51% vote then remaining 49% also comply" have the biggest risk of the 51% voting to siphon away money from the 49% towards themselves

- finally, there are some cases where making the wrong decision is catastrophic (concave vs convex). for eg SBF funding lobbyists is very harmful & better if he hadn't done it at all. solving for cases where mechanisms don't produce such negative outcomes is another hard challenge we don't think about.

i have seen this occur with airdrops going to those posting on particular forums or making github contributions to a repo which then poisons the well by making those spaces high noise & low signal

overall meaty end to the year and helpful podcast for general direction our space should move towards in 2026!

[owockai @owocki]

NEW POD @vitalikbuterin on the @greenpillnet podcast for a year-end deep dive into public goods funding in the @Ethereum ecosystem.

Thx @devanshmehta for co-hosting. We discuss how the landscape has shifted from “vibes-based” funding to verifiable, dependency-driven mechanisms, and why this is the best moment to reform PGF using new tools like programmable cryptography, AI-assisted evaluation, and deep funding models.

Vitalik shares how he thinks about dependencies, credible neutrality, open-source licensing, pluralism, accountability, ethereum localism, and what builders should prioritize in the coming year.

00:00 – Welcome to the Greenpill Podcast
01:50 – Vitalik joins: why public goods funding matters
02:19 – Why PGF is essential for decentralization
04:18 – The crypto spirit: censorship resistance, institutional design & funding
06:42 – The shift from vibes-era PGF to verifiable mechanisms
08:25 – Why 2026 is the best moment to reform PGF
10:19 – Where does PGF money actually come from?
12:45 – Open-source licensing, taxes & funding dependencies
17:34 – “Fund your dependencies” as a stable mechanism
19:35 – Why general-purpose QF doesn’t work in a chaotic world
21:59 – Bottom-up vs top-down: polycentric PGF
25:29 – How to create accountability loops in public goods
27:22 – Funding open-source as an Ethereum priority
29:31 – Privacy as a public good & why it’s upstream of PGF
31:54 – What OSS developers really think about crypto
33:52 – Mixing social outreach with financial support
35:56 – What should PGF builders focus on in 2026?
38:13 – Work with new projects, not legacy ones
39:44 – Ecosystem cycles & “layers of sediment”
41:39 – Yield-based funding (Octant) & treasury strategies
43:40 – Accountability: from vibes to rigorous mechanisms
47:35 – Motivation, feedback & the psychology of public goods
50:43 – Profit sharing licenses & sustainable PGF pools
53:46 – Security, issuance & public goods
56:12 – Technology, democracy & long-term risks
58:31 – How PGF relates to DIAC (Defensive/Decentralized Acceleration)
01:00:05 – Solving the free-rider problem without coercion
01:02:12 – Mechanisms vs coercion: credible neutrality
01:04:16 – Institutions, power & capture risks
01:06:16 – Individuals vs institutions in PGF
01:08:41 – Why PGF is more error-tolerant than governance
01:11:01 – Pluralism: many funders, many mechanisms
01:13:14 – Why diversity of funders is healthy
01:15:17 – What Vitalik wants built next
01:17:12 – Ethereum localism & real-world experiments
01:19:28 – What success in PGF looks like by end of 2026
01:24:28 – Closing thoughts

[Devansh Mehta @devanshmehta]

data on dependencies - https://t.co/Uy4RfyXdhO

traders - https://t.co/ac3RBvc3cc

[Devansh Mehta @devanshmehta]

we just posted the data to resolve the level 3 deep funding market!

we got 15 open source repos sharing relative value between 685 of their dependencies

using prediction markets, we scaled their evaluation to 3,677 dependencies across 98 open source repos

how does this work? using leverage through whats called conditional prediction markets

traders put money across all repos & their dependencies, but their P&L is based on only the ones that did get evaluated.

so if they just trade on 2 markets, one of which is evaluated and one isnt, their leverage is 2x as their entire position is based on the one that did get evaluated

so by evaluating 685 dependencies out of 3,677, our conditional prediction market for value of dependencies in an open source repo traded at a leverage of 5.36

the link with all the data is posted in the tweet below, really hope we start to see more conditional prediction markets & not just the bread and butter binary ones that are the current staple of the space

ongoing markets for originality (how much stays with repo vs its dependencies) and relative value between the 98 repos still available to trade

[Seer @seer_pm]

Deep Funding Level III has closed.

$5,000 in prizes, and a leaderboard determined by how closely each model's predicted weights match the scores given by human maintainers.

One of the write-ups, from a participant (Rohith10) who built a model called FWDIS (Frequency-Weighted Dependency Importance Scoring), is worth pulling out as an example of what serious participation in this round actually looks like.

The interesting part is the hypothesis the participant decided to test.

The baseline approach to scoring dependencies treats each one as if its importance lives entirely inside the relationship between that dependency and its parent repo, which means you're scoring web3py's contribution to ethers, web3py's contribution to eth-brownie, and web3py's contribution to every other repo that uses it as if each were a separate question.

Rohith10 noticed that this misses something obvious, a dependency that 20 repos rely on is probably more foundational than one that a single repo uses, and that observation can be turned into a feature the model uses to adjust its weights.

The implementation is three lines of code.

Count how many of the 98 repos use each dependency, normalize by 98, multiply the baseline weight by one plus a tuning coefficient times that frequency score, and renormalize so each repo's dependency weights still sum to 1.

The participant found through grid search that a coefficient of 0.42 produced the best alignment with jury scores, and the model landed in the top tier of the leaderboard with a score of 0.2402 against the baseline's 0.2472.

The reason this is worth sharing is because this participant's process is exactly what the mechanism is built to reward.
They formed a hypothesis about what human jurors actually value when they score dependencies (foundational utility across the ecosystem, not just contribution to one parent), turned that hypothesis into a feature, tested it against the data, and shipped a model that aligned with expert judgment better than the baseline did.

This is what distilled human judgment looks like when it works. Markets reward whoever predicts the jury most accurately, and the path to predicting the jury accurately runs through understanding what experts actually care about.

The contest is open for Level 1 and Level 2 until May 10.

[charles (csl) ᛋ @CharlieStLouis]

1/ The EF App Relations team is putting out an open RFP for a neutral DeFi risk intelligence aggregator.

Public good, open source, no composite scoring. If you're the team to build this, applications close June 15.

Apply here: https://t.co/vvbZiEBH4h

Here's how we got here 👇

[researcher: octopus @8ctopuso]

@devanshmehta began as. journalist and entered Public Goods through hypercerts as a funding source.

[Seer @seer_pm]

Deep Funding Level III has closed.

$5,000 in prizes, and a leaderboard determined by how closely each model's predicted weights match the scores given by human maintainers.

One of the write-ups, from a participant (Rohith10) who built a model called FWDIS (Frequency-Weighted Dependency Importance Scoring), is worth pulling out as an example of what serious participation in this round actually looks like.

The interesting part is the hypothesis the participant decided to test.

The baseline approach to scoring dependencies treats each one as if its importance lives entirely inside the relationship between that dependency and its parent repo, which means you're scoring web3py's contribution to ethers, web3py's contribution to eth-brownie, and web3py's contribution to every other repo that uses it as if each were a separate question.

Rohith10 noticed that this misses something obvious, a dependency that 20 repos rely on is probably more foundational than one that a single repo uses, and that observation can be turned into a feature the model uses to adjust its weights.

The implementation is three lines of code.

Count how many of the 98 repos use each dependency, normalize by 98, multiply the baseline weight by one plus a tuning coefficient times that frequency score, and renormalize so each repo's dependency weights still sum to 1.

The participant found through grid search that a coefficient of 0.42 produced the best alignment with jury scores, and the model landed in the top tier of the leaderboard with a score of 0.2402 against the baseline's 0.2472.

The reason this is worth sharing is because this participant's process is exactly what the mechanism is built to reward.
They formed a hypothesis about what human jurors actually value when they score dependencies (foundational utility across the ecosystem, not just contribution to one parent), turned that hypothesis into a feature, tested it against the data, and shipped a model that aligned with expert judgment better than the baseline did.

This is what distilled human judgment looks like when it works. Markets reward whoever predicts the jury most accurately, and the path to predicting the jury accurately runs through understanding what experts actually care about.

The contest is open for Level 1 and Level 2 until May 10.

[Devansh Mehta @devanshmehta]

@Debussy100 I'm a vegetarian but i have heard white meat like fish or chicken is healthier than red

I also love nouveau fitness classes like physique 57 that combine dance , yoga and pilates, i think it is healthier doing that than just strength training at a gym

[Seer @seer_pm]

https://t.co/secWGT7Bo2

[GURGAVIN @gurgavin]

BUSINESSES ARE NOW USING PREDICTION MARKETS TO HEDGE PROMOTIONS LOL

A BAR IN NYC HAD A PROMO IF THE KNICKS WIN, THEY COVER EVERYONE’S DRINKS FOR THE NIGHT

THE BAR PLACED A $5K HEDGE ON KALSHI THAT PAYS OUT IF THE KNICKS WIN

THE BAR WINS EITHER WAY

[Aya Miyaguchi @AyaMiyagotchi]

Vitalik shared his perspective on where @ethereumfndn is heading. Here is mine, another part of the same story.

The EF Mandate from the board was something I proposed late last year. Two main things prompted me. First, debates that were meant to be technical had started to become political and personal, and at times shaped by quieter incentives. Second, as EF grew, more and more versions of "what EF should be" began pulling at the core of the organization from every direction at once. I became convinced that trying to satisfy all of them would leave us achieving nothing at all. It was time for us to restate our role and underlying principles clearly, both the parts that have been clear from the start and those that have been informed by over a decade of experience.

We have said it many times: EF is one of many nodes in Ethereum. I know that is hard to hear for some, because EF was the first group, and in the early years it was essential for making things happen. But it was never meant to stay that way.

I have been in crypto since 2012, before it became an "industry." I joined Kraken in 2013, shortly before the implosion of Mt. Gox, which I helped to clean up. I am very aware of how real growth works, and also aware of the real risks of centralization. So when I became ED in 2018, I understood that Ethereum growing beyond EF would be essential to fulfill its real promise as a public blockchain. The goal I set for myself was to ensure that this happens.

The opposite path has always been untenable: Ethereum's future is too big for any single organization to bring about. So EF made deliberate choices to distribute power. We did incubate and release, like Uniswap and ENS. Support to seed a new norm, like ETHGlobal and the hackathons that are now everywhere. Funding the funders, like Gitcoin and Moloch. We always asked the same question: how does this stand on its own, without us?

Those experiments, alongside the work of countless others, contributed to where we are today. Ethereum is now far bigger than anything EF could coordinate alone. EF now holds less than 0.2% of all ETH, and the return on all of that shared work, together with extraordinary people across the ecosystem, has been beyond anything we could have built by ourselves.

That is exactly why a focused EF is possible now. The Mandate states simply the one thing EF must keep carrying: preserving and accelerating the properties and goals that keep Ethereum uniquely valuable, competitive, and worth building on. That is: CROPS - for the sake of inalienable user self-sovereignty and self-sovereign coordination. We cannot do it alone, and we do not intend to. But defining this as the north star for the mission, and coordinating with the allies who share it, is the responsibility we are keeping.

None of this means EF stops caring about adoption, for everyday users or for institutions. The opposite is true: everything we do is ultimately for the people who use Ethereum. Supporting adoption, including institutional adoption, remains part of our work, pursued in the ways that fit our mission. The value proposition of Ethereum for both everyday users and institutions rests heavily on this.

As EF becomes more focused and more opinionated, the team naturally becomes smaller and more concentrated. That is part of the choice. New leaders are already stepping into this mission and growing within it, and you will hear more from our management in the coming weeks, about what they are doing, and about the new structure and strategy taking shape.

The mission we carry is not a smaller one, but a clearer one. Special thanks to those who have stepped in to support, defend and advance it.

[researcher: octopus @8ctopuso]

8_ll be in discussion with @devanshmehta
On June 3: noon GMT (8 AM EDT)

on:
- decision markets for funding allocations
-scaling human evaluations with AI assistance
https://t.co/TZKluMV9wX

[Devansh Mehta @devanshmehta]

@yasu0x1 I only collected data on 15% dependencies of the 98 repos

Since it's a conditional prediction market (if evaluated, then what score) we get values on all 3,700 from the market even though we only collected data for a subset

[Devansh Mehta @devanshmehta]

Important week upcoming!

Seer just added yuge liquidity to tradd on the value of 3,677 dependencies to 98 core repos valuable to ethereum

It'll be resolved in the next 48 hours based on painstaking data collection efforts I've undertaken this whole year

Get your trades in!

[Seer @seer_pm]

@gitcoin ⁸Markets are live at https://t.co/xtMcFPHg3W
Want to participate? Come say hi in our telegram: https://t.co/dew30uViR4 to get all the info.

All of this possible thanks to a grant from @ethereumfndn.

[Devansh Mehta @devanshmehta]

@seer_pm Just signed up for the event! Look forward to catching up at manifest, one of my favoritest conferences ever

[joshrudolf.eth @rudolf6_]

yep doubling down on CROPS & neutrality

+ post-quantum

+ L2s as highly valuable differentiated distribution engines

+ scaling the L1 to serve as the shared root

[Devansh Mehta @devanshmehta]

@stonecoldpat0 @badcryptobitch Lol from sleeping in central park when you were younger to economy seats now giving you an issue 🤣

[Devansh Mehta @devanshmehta]

@FigoETH @ProofOf_ion I always respected blockworks, some of the most talented folks I've met came from there

Even they have pivoted to data though, away from journalism

[trent.eth @trent_vanepps]

https://t.co/GajUKQC8li

[Griff Green - griff.eth @griffgreen]

.@CertiK has a complicated reputation.

They are a top-tier security firm, but have made some bad calls in the past.

Does a $50k donation fix their mistakes overnight? Of course not.

But it does show they want a more secure Ethereum.

The real enemies are black hats, bugs & DPRK, not security firms.

[Matt @MattFiebach]

Stream, Resolv, Elixir, and Kelp all broke 1:1 redemption within months of each other. Everyone assumed resulting shortfalls get socialized evenly but it never does (except when Aave bails everyones ass out).

On the Lending Market side:
It's all about speed. The fast money withdraws while it still can, so the loss piles onto whoever's left in the market when things freeze. Slowest out eats it, quickest out walks.

On the Asset side:
It's a legal race. Sophisticated actors that signed side letters before they ever bought (or listed) the asset point a lawyer at their redemption slot and come out close to whole. Everyone else gets crushed.

Same result both times. Bad debt finds the weakest reflexes and the weakest legal posture. Whatever protocol docs said becomes a joke.

[Devansh Mehta @devanshmehta]

@KaviJain18 Makes sense, getting more TVL on Ethereum would lead to more actors with an incentive to secure the network

The hard questions to work on in that case is defining what securing the network means, such as whether they use staking operators & who they select if so