Research Director @PanteraCapital. co-host @indexed_pod. previously data @ Flashbots ⚡️🤖 @0xProject @ConsenSys.
Joined August 2016
- Tweets 4,791
- Following 1,869
- Followers 10,970
- Likes 22,690
746 Photos and videos
The more ironic bit 😂 h/t @joeblau for pointing this out:
The exact day ETH became deflationary was also the day XEN launched (2022 Oct 8th).
So while we’ve spent years complaining that XEN bloated Ethereum state growth, it may have been the only thing saving our bags 🥲
3
2
10
1,514
Follow-up comment from @dataalways — that timeline when ETH stopped being deflationary is also after when EIP-4844 came out in March 2024. After which L2s transitioned from paying calldata to blobs.
I remember this whole thing now re L2 scaling roadmap 😅
7
1,201
Yesterday our house wizard @whitemilk3535 reminded me that ETH is an inflationary asset.
I was in disbelief as I had the impression that after EIP-1559 we just go up only while burning base fees, til I checked the charts -
Turns out ETH stopped being deflationary a long time ago since April 2024. We're simply not burning enough.
What happened to ultrasound money 😭
22
2
50
10,712
Very cool to see our paper getting broader attention through @TheStreet’s coverage! 🤩
The first wave of onchain agents was largely driven by memecoin speculation rather than actual autonomous onchain execution. But beneath the hype, many teams have been hard at work building proprietary AI models purpose-built for trading. Looking forward to seeing where it goes.
Jun 11
Despite the hype, new IC3 research reveals that a large number of AI agent platforms aren't actually conducting autonomous trades.
Thanks to @TheStreet for the coverage. H/T to the authors @0xfishylosopher, @theamyzhao, and @sui414
thestreet.com/crypto/trading…
3
22
3,950
danning retweeted
Jun 11
Prediction markets are incredible, but the wisdom of the crowd is a marketing claim
3.1% of traders produce the accuracy, while 84.1% fund it
"The Truth Machine That Takes Your Money" drops tomorrow
Researching this one was a blast, and the data tells a wild story
Many thanks to @andrewcourt1, @sui414, @nikopolos, and @jw_seoul for the review and feedback
Jun 11
Prediction markets are accurate.
But 3.1% of traders produce that accuracy, while 84.1% lose money.
This week, we analyze who actually drives the edge, who funds it, and whether the "wisdom of crowds" holds up.
Out this Friday.
2
1
7
655
if AI
- gives recs on IC memos for VCs
- prioritizes calendars for CEOs
- write the analysis for policy maker, researcher
- designs personal gym, skin care and diet routines
- controls what to show on social media feed
isnt AI already running our society and the world today?
6
1
15
2,340
danning retweeted
Jun 6
In late 2024, the "AI DeFi Agent" narrative exploded, driving over $3 billion in combined token valuations. But is this boom driven by actual autonomous trading, or is it just a speculative bubble?
A new paper, "Paper Agents, Paper Gains," dives into the empirical data. One of the paper's authors @theamyzhao shared an overview of their findings:
2
2
21
4,222
danning retweeted
Jun 5
Had a ton of fun presenting at @Designing_DeFi my paper "Paper Agents, Paper Gains: An Empirical Study of DeFi Investment Agents", a survey tracking the scope, design, and performance of trading agents from Nov '24 to Nov '25 with @theamyzhao @sui414. Our key findings:
> We scraped 1900 projects on CoinGecko (post-TGE) and Messari (pre-TGE) tagged as AI, filtered out pure infra/non-agent projects, and selected 10 representative agentic projects, incl. Giza, Surf, Numerai, AIXBT, Eliza.
> We evaluated their strategy, observability, source of funds, AUM. Most agents don't actually autonomously spot trade - either they provide investment advice or engage in passive yield strategies. Even if they do trade, they don't have great onchain observability and verifiability that an agent is actually signing (vs. a human signer).
> We chose to analyze Virtuals ElizaOS frameworks deeper as these had the highest onchain standardization observability. Looking at top volume projects built on Virtuals ElizaOS, we realize that several have $100M peak market cap, but extremely high MC-to-AUM ratios, with some peak ratios exceeding 10,000x (compared to 1-3x for DATs). They mostly also trade micro-cap tokens, often holding other agents' tokens.
> We then choose 11 Solana projects to analyze onchain PNL for users and agent treasuries (rPNL uPNL). There seems to a disconnect - treasuries have a net profit of $34.3M, but users have losses of -$191.7M
> Furthermore, there's heavy user PNL inequality, 81.4% of gains are captured by top 1% users, while 62.2% of all users suffer net losses. Agent tokens had 93% peak to trough drawdown vs. Solana only had 54% drawdown. So for the average user buying into these DeFi Investment Agents, things didn't really turn out great.
> So where do we go from here? Our findings align with theoretical results abouts the "CoinAlg Bind", which posits a tradeoff between economic fairness and profitability for Collective Investment Algorithms (CoinAlgs). But DeFi trading agents are clearly here to stay, and the space is moving extremely quickly.
> In our view, a mature DeFi Investment Agent must satisfy 3 criteria - (1) verifiable autonomy/execution of trades, (2) sustainable profits vs. passive benchmarks, (3) stakeholder alignments, where profits are actually passed back to individual users. While the space is still early, we're really excited about how these DeFi Investment agents can continue to evolve.
[FULL PAPER LINK BELOW]
14
6
75
7,096
imagine all the crazy thing homeless ppl will do with no dignity, to get drug money. this is too dark.
Introducing pump fun GO: Pay ANYONE to do ANYTHING
Create & complete bounties for ANY task and leverage the power of humans & money across the globe
The world is at your fingertips. It’s time to GO 👇
5
34
4,201
danning retweeted
Jun 4
🚨NEW EPISODE🚨
Do AI Trading Agents Actually Work?
Today we're diving into AI trading agents, the rise of DeFAI, and whether autonomous trading bots are actually generating alpha.
We discuss:
- AI trading agents and the DeFAI hype cycle
- @elizaOS, @virtuals_io, and agentic trading products
- Onchain tracking vs qualitative research
- Whether agents actually made money
- Early wins, later losses, and extractive markets
- Alpha Arena and AI model trading competitions
- Why sample size matters for agent experiments
- Agent reliability, harnesses, and model variability
- AI workflows, Dune MCP, and product building
- Ethereum bearishness, EVM success, and ETH value capture
- Stablecoins and prediction markets
And much more—enjoy!
—
Timestamps:
(00:00) Introduction
(00:46) AI trading agents
(01:26) DeFAI product promises
(04:55) Eliza and Virtuals
(05:08) Research methodology
(12:58) Biggest wallet gains
(14:48) Profitability fades
(16:19) Alpha Arena experiment
(19:18) Sample size problem
(25:57) Learning from competitions
(31:51) Dune MCP workflows
(42:04) Ethereum bearishness
(53:41) Outro
2
4
12
2,152
you get replaced by AI
➡️ now all u can do is jobs accelerating AI replace more people
🔁➡️ UBI and neo-feudalism?
My doom scenario isn’t that AI gonna kill humans.
but that AI will “enslave” humans by extracting your last bit of value in the job market -
The people who lost jobs and got replaced by AI are now employed to
clean your house for free, wearing a camera (to collect training data)
🚘 sit in robotaxi (with no talking allowed)
⚡️ host mini data centers in the backyard, for Nvidia
👁️🗨️ click and move eyeballs on arbitrary tasks (to provide neural and behavioral data)
🛠️ workers in Chinese factories now are wearing wearables and sensors to assemble devices and cars, selling data for cheap to AI embodiments companies
——
In just a year, our life is full of AI. The moment video models took off, everyone’s Instagram Reels was taken over by dancing dogs and cheating fruits. 🍌🍓☠️ (à la Strawberina😭)
When the training is done, adoption takes ~0 time. GEN AI was the first. Next up are robotics and autonomous driving.
In the longer term, all these things may bring us cheaper robot labor and products….but scared to imagine how that day will be when “AI are done learning everything about human” as well.
1
2
708
My doom scenario isn’t that AI gonna kill humans.
but that AI will “enslave” humans by extracting your last bit of value in the job market -
The people who lost jobs and got replaced by AI are now employed to
clean your house for free, wearing a camera (to collect training data)
🚘 sit in robotaxi (with no talking allowed)
⚡️ host mini data centers in the backyard, for Nvidia
👁️🗨️ click and move eyeballs on arbitrary tasks (to provide neural and behavioral data)
🛠️ workers in Chinese factories now are wearing wearables and sensors to assemble devices and cars, selling data for cheap to AI embodiments companies
——
In just a year, our life is full of AI. The moment video models took off, everyone’s Instagram Reels was taken over by dancing dogs and cheating fruits. 🍌🍓☠️ (à la Strawberina😭)
When the training is done, adoption takes ~0 time. GEN AI was the first. Next up are robotics and autonomous driving.
In the longer term, all these things may bring us cheaper robot labor and products….but scared to imagine how that day will be when “AI are done learning everything about human” as well.
3
15
2,016
May 28
Today, we're launching shift. We're starting by cleaning your apartment in New York City, for free.
Here's how it works. Book a shift cleaning. A vetted shift operator comes to your home wearing one of our devices. They clean. They leave. You pay nothing.
In exchange, we record the cleaning. Robotics is being built on data about how people do daily tasks, and the value of that recording is what funds the service. Anything personal in it is anonymized before the recording is processed.
By now, you have heard about the shift to AI more times than you can count. About the shift toward you, the part where you actually feel it, you have heard almost nothing. Shift is what starts to make it concrete, in specific cities, with specific services.
Today, cleaning in New York. Soon, handymen, repairs, and errands across the globe. And this is just one side of shift, with more on the way.
Comment “shift” and we’ll send you an early access link.
2
1,259
Mechanism designers, please rally
What are some better oracle resolutions for Polymarket?
- delayed resolution: market stops, wait til SEC filings / gov docs to determine the Strategy sale
- dispute committee: auctions with oracle solvers, resolve when reaching quorum threshold, with rewards & slashing
- what else?
13
15
4,676
danning retweeted
RWAs onchain are scaling and maturing together.
Our Tokenization Progress Index (TPI) scores how much of an asset's lifecycle has actually moved onchain.
Stablecoins sit furthest along both dimensions: large market value and higher TPI.
Tokenized Stocks show relatively high onchain maturity, while market size remains smaller.
Meanwhile, Treasuries, Commodities, and Private Credit have reached larger market sizes but still sit in lower-to-middle TPI bands.
The rest of the asset classes remain early and primitive.
For the full TPI methodology and a deeper look at the sector, read our Q1 2026 tokenization report: panteracapital.com/state-of-…
Beyond the Digital Copy: What's Next for Tokenization?
Our research team developed a new index of 593 assets across a $320B market that reveals how far onchain finance still has to go.
Access the full report here:
panteracapital.com/state-of-…
15
39
69
12,280
My unofficial reason for loving @SurfAI is that it fetches tweets for me without any hustle 😂
it's free and takes ~30 seconds to install as a skill in your Claude Code/Codex: agents.asksurf.ai/
3
6
33
3,709