During my undergrad, a friend and I randomly showed up at Prof's office. We weren't even in his uni. We told him we want to work on game theory/ mech design/ social choice. Instead of kicking us out of his office (which he should have), he gave us stuff to work on and mentored us
One of the really fun things i got to do this summer is visit the amazing lab they're building at
@a16z
crypto research and chat with
@Tim_Roughgarden
@skominers
and others about cool game theory/ mech design questions they're facing up and down the stack... 1/n
Apropos
@ShengwuLi
's tweet about history of economic thought--- i still remember how blown my mind was when i read the first paragraph of rothschild and stiglitz:
In classic academic papers, some passages confuse the modern reader, because they are defenses so successful that the critiques they defend against are long dead. Some passages are written to be remembered; some passages are written to be forgotten
we want to design anonymous, permissionless, decentralized markets that are robust to adversarial behavior... but our social networks must stay pristine and doxxed
Old joke: Advisor tells their graduating student to leave 20$ bill in the hardcopy of thesis in the library.
Oh as a reward for the first reader, the student asks?
No, when you come back to town you have backup cash for lunch the prof replies.
Spent a lot of time trying to come up with a pithy, informative thread about a new paper I'm pretty excited about, "Censorship Resistance in On-Chain Auctions" (with
@MaxResnick1
and
@eljhfx
:
@chiarapack
I see you got answers already but if isn't available online freely, yes! Indeed, I once emailed a prof (in France) as an undergrad (in India)... he literally snail-mailed me a preprint of the paper!! I think i still have it somewhere...
I think this is a cool paper. Here's why: Econs have long had two basic models of discrimination--- taste-based (I just don't like your race/ gender/ ethnicity so I discriminate against you) and statistical (I see different information about your race/ gender ethnicity so I
Professor Rahul Deb is one of the researchers behind a new mathematical model that uses wage distribution to isolate the effects of discrimination in wage gaps.
#EconTwitter
🚨 Really excited to host a workshop on Blockchains and Decentralized Finance tomorrow, Monday July 8th at EC24
@AcmSIGecom
. Quite an exciting program up ahead:
Should authors sign statements before submitting as to their contribution? Ex post “wasn’t me” excuses that your coauthors or RAs did the data work would be better supported. And, it’ll help editors decide whether they are ok with papers produced by RAs who are not coauthors.
This is a true story. i took a grad asset pricing class.
my brother (tradfi bro): hey tell me more maybe some of this will be useful for me.
me: sure. think of an asset as a point in a hilbert space.
my brother: we will never talk about finance again. CLICK.
In one of my first weeks as a junior faculty, GJM took me out to lunch and said, "remember, when you get ref reports on your papers, bring them to me". and when i did some point later, he literally went through them (and my responses) line by line.
- Figuring out how to deal with referee reports, submission strategy, etc. is a learned skill and not something you should try to navigate without advice: ask mentors, seniors in your dept, etc.
@ben_golub
As a robustness check, the authors should prove that their approach would also work for all other viruses. The details can be put in an online appendix.
Nothing?? I asked. He said well I understand nothing in any talk. So then I said why do you go to seminars.? And he said the function is 0 pointwise, but the integral is positive :)
That was amazing analysis by
@Mylovanov
(and friends) who are somehow thinking analytically on the side while helping their fellow citizens and suffering a massive bombing campaign. Stay safe, and hopefully we can talk (and help) more whenever this ends
A senior colleague in a talk once put up her related literature slide, smiled at the audience, and said "and here is a list of papers i plan to read that appear related"
Indian cricket prospects all going to get messages from their mom tonight, "beta I know you want to do cricket but you can be like Saurabh and be engineer too, na?"
Saurabh Netravalkar, who had figures of 2-18 in his 4 overs against Pakistan, is currently an engineer at Oracle. He also has an MS in computer science from Cornell University.
I saw this shocking exchange yesterday on Twitter. Person 1 posted politically controversial but reasonable position. P2 asked reasonable question. P1 gave calm reply. P2 said thank you that makes sense, I'm going to go think about that now. Ban both now!!!
My migration story was made into a major Hollywood movie. The screenwriters took some artistic liberties (e.g., I flew over rather than took the route indicated in the movie), it captures the gist of my struggle (e.g. I did fly economy):
I just spent another afternoon perfecting a 30-minute presentation about my migration story for 9-year olds.
The presentation contains many hardships that I faced in my life such as when my parents refused to buy me a dedicated pair of giant slalom skis when I was 12 years old.
As a researcher in crypto I want to work on the important problems vexing everyone in this space: how should we design txn fee mechanisms? can we limit MEV without limiting chain efficiency? who is
@GwartyGwart
and how are they right about everything?
@ben_golub
@ElliotLip
Elliot's daughter has now experienced econ refereeing too--- have a quality insight, immediately get shot down because it doesn't generalize to infinite dimensional Banach spaces.
🚨🚨 Researching on DeFi/ Blockchains? We are seeking paper submissions for our half-day workshop at EC24 at Yale on July 8th.
We welcome research on a broad range of topics, including Transaction Fee Mechanism Design, Censorship Resistance, Mechanism Design for block
Thanks
@jasondhartline
! Delighted to bring some virtual value to
#EC2023
(heh)... More seriously, for those of you who have thoughts on things you like/ dislike about your experiences e-attending conferences, have suggestions etc: DM/ email me!
Delighted to have
@MalleshPai
as the Virtual Chair for
#EC2023
.
@AcmSIGecom
is committed to its meetings being broadly accessible and inclusive. Not able to join us in London? Block out July 9-12 to join the virtual meeting, packed with livestreams, networking sessions, etc.
As a grad student, I attended a particularly incomprehensible seminar. After the seminar was over, I ran into a senior (and very famous) prof and asked him what did you get from this talk. He said, nothing.
“Meta CEO Mark Zuckerberg says there are complex copyright questions around scraping data to train Al models, but he suggests the individual work of most creators isn't valuable enough for it to matter.”
This week
@kse_in
/
@kse_in
Seminar: the inimitable
@skominers
on "Redistribution and Marketplace Design" (joint work with
@akbarpour_
and Piotr Dworczak). Noon Eastern on 5/18. Join!! Sign up link here:
Michel Talagrand has been awarded the Abel Prize, one of the highest honors in mathematics, for applying tools from high-dimensional geometry to complex probability problems.
@jordanacep
reports:
A day in the life of a mechanism designer:
Job
#1
: "should we combine /theta and /omega? makes clear that our formulation extends to non-EU preferences."
Job
#2
: "hmmm interesting mechanism. but what if someone is absolutely aping into a shit coin"
This tbh. We need to set a spec on what is acceptable. It might suck if some willing stakers can't stake because their internet access isn't good enough, but we *already* screen people on access to 32Eth = ~80k USD of capital! Not everyone can afford that!
bullish solo staking
bullish geographic decentralization
bullish on having real bandwidth targets that are based on more than a couple people’s fingers in the air
why do people wake up and choose violence against mechanism designers? we're like the rest of you, just use more latin in everyday life and greek in our writings. oh wait i see it now.
One of the things that bothers me about blob pricing is that there's no feedback with transaction fee pricing. Blobs use basically no compute, but they still use storage and network. So when you have gas fees at 100 gwei like earlier today you still have 1wei blobs.
My colleague Konrad Mierendorff died on Saturday August 7th 2021 after a fall on the
#Wasenhorn
, Switzerland. Konrad was a gifted Economist, both as a researcher and a teacher. He loved the mountains and was on his own journey into Alpine climbing. 1/5
Rani Spiegler has a fantastic book about the culture of economic theory. Looks like a wonderful read for all, highly recommended for students interested in economic theory, and also for colleagues (EC types) who want to understand us! For now, a trailer:
Iirc If you look at the original vision that was the idea. Like there wouldn't be 10 general purpose l2s replicating the L1. Instead there might be 1 gaming L2, 1 ai L2 etc. also reduces the need for interop. Trouble is only finance stuff is profitable rn.
At the risk of either overselling or underselling this: this is a blindingly obvious paper that I really enjoyed working with
@MaxResnick1
and
@mikeneuder
on. 2 simple points.
NEW RESEARCH
"Optimizing Exit Queues for Proof-of-Stake Blockchains: A Mechanism Design Approach"
Changing validator sets over time potentially erodes proven security of fault-tolerant consensus protocols.
by
@mikeneuder
,
@malleshpai
,
@MaxResnick1
And mentors of mentors of scholars (think how many of us are students/ mentees of Krishna/ Mailath/ gul/ Jackson/...). RIP Hugo Sonnenschein. Thanks for teaching my teachers how to fish.
@DinaPomeranz
Heh for some of our research agendas, getting 20 newspaper articles can only happen if we do something very bad. So I'll take the top 5 :)
This is a lovely paper. Must read if you're interested in restaking risks. Even if you're not, it's an elegant mathematical model/ formulation with clean results.
New paper with
@n_durvasula
offers a framework to reason about the risks of restaking. Key question: Under what conditions can validators be safely reused across multiple services? 1/13
What a fun two days! Our third economic theory mini conference in the books.... Thanks for coming
@JuanDubra1
(also Léonie Baumann, Nicole Immorlica and Rakesh vohra!)
I didn't think that after 2 years of pandemic/ zoom fatigue, I would be excited about another zoom seminar but this was amazing. Thank you
@JeanTirole
!! Next up,
@DrDaronAcemoglu
(April 6, 12noon Eastern). Sign up details:
@kse_in
.
This talk convinced me that Ethereum can implement multiple concurrent block proposers
@MaxResnick1
proposes a new design called Braid, based on the simple idea of running many instances of Ethereum consensus in parallel
I think this could become a defining part of the roadmap
A cool thing about this space is that some top minds are thinking about it. And among the academics (and more generally) in the space,
@noamnisan
is top of the heap. A must listen for academics esp in the EC community wondering what there is to do round here...
🎙️This week,
@annarrose
&
@tarunchitra
chat with
@noamnisan
, exploring his journey from complexity theory to pioneering AGT, his move to blockchain and his work at
@StarkWareLtd
+ his latest work on PoS Tokenomics and more.
1/4 Biologists often interpret resource exchanges among organisms as symbiotic mutualisms. My latest in
@PNASNews
shows these can be modeled as Walrasian equilibria.
Here's a talk i gave (which they were nice enough to record) while i was there on the aggregating the wisdom of the crowd by also eliciting higher-order beliefs. (2/n)
While people look at omicron data and case numbers to glean the future, let me offer some hopeful news: liquid hand soap was in stock at the local Costco today for the first time since Feb 2020.
With so many possible roadmaps for Ethereum, how do we move forward together?
We need to clarify our optimization targets. If we care about home stakers, what does that actually mean? If bandwidth is the bottleneck, we need to have dashboards showing bandwidth usage of all the
Censorship resistance in block chains doesn't just mean that valid transactions in a blockchain get in *eventually*. For blockchains to be useful for e.g. finance they have to get in *fast*. This has been understood for a while:
3rd+ year PhD students: have a website! Formatting doesn't matter. Google sites or your dept website is fine; github if you're showing off.
- Include your CV + email
- List your fields/interests + any ongoing projects
- Post PDFs of recent publications + good working papers
Are you doing a ecoomics-related PhD?
Interested in AI/ML?
Then apply to our Summer Institute in Machine Learning!
Learn from awesome faculty
Meet other like-minded students
Get exposed to cool, new ideas
(march 24 deadline)
@ben_golub
@paulgp
you joke about this but apparently back in the day cake mix companies realized that their clientele found making a cake from a mix "cheating"--- so they added back some slight "baking experience" e.g. add an egg or mix some milk.
⏰⏰ Just under a week to go for
@AcmSIGecom
EC Preview week!!
A quick run through of the planned activities (fully virtual) below, tl;dr:
mentoring workshop registration (free!)
preview week registration (free!):
(1/n)
Tymofiy is a friend and former colleague who's been doing incredible work as the president of the KSE in Ukraine to establish the kind of democracy and durable institutions we take for granted here.
You have been asking how you can act. Here it is.
"Kyiv School of Economics" (501.3(с) nonprofit registered in Washington DC) is collecting funds for humanitarian aid to Ukraine!
(USD) Euro is also accepted, please follow the QR code.
There's a phase of research I call Not Even Stuck (by analogy with physicists' Not Even Wrong), where things don't work but you don't understand the problem well enough to know what the blockers are. If you can get from there to Stuck, you've made a lot of progress.
@instrumenthull
@wwwojtekk
sure, but at some point then is self-made just tautologically vacuous? like you went to a good uni--- oh you have a uni degree, not self-made. your parents have enough money to invest in your startup? not self-made.
One of the amazing things about this space for mechanism designers is that ideas get taken to fruition and you can see them actually get to practice (not to say we're the only people who came up with MCBP, but I'd like to believe we had a part). It's super satisfying
@_danielmarzec
@MaxResnick1
Was planning to release this ROP next week but we'd be interested to think more about it! h/t to
@0xQuintus
for sharing your note too, great stuff to get started
@ben_golub
@AlexKontorovich
Unrelated: his advisor Jacob Laurie had the best sentence in his talk at icm 2012... Something to the effect of "you might think this is theory for it's own sake... it's not... it's theory for the sake of other theory". I've always wanted to use that in a talk
In contests like rent-seeking or R&D, should the efforts of earlier movers be disclosed? The paper shows that such transparency enhances total effort, maximizing it with full transparency and minimizing it with none.
OH: If crypto has 1 contribution to society it's gathering all the r****** in one place and giving them toy problems to work on so they don't bother people who actually work on important stuff
(Funny because I was told something similar about econ theorists once)
@ShengwuLi
I have to disagree (if only because, e.g. you and your advisors etc give amazing talks and are presumably great teachers). My suspicion--- giving good talks and/or teaching well, just like research, requires effort. Too many folks substitute away...