Category: Web/Tech

Is AI centralizing research influence?

Increasingly, modern Artificial Intelligence (AI) research has become more computationally intensive. However, a growing concern is that due to unequal access to computing power, only certain firms and elite universities have advantages in modern AI research. Using a novel dataset of 171394 papers from 57 prestigious computer science conferences, we document that firms, in particular, large technology firms and elite universities have increased participation in major AI conferences since deep learning’s unanticipated rise in 2012. The effect is concentrated among elite universities, which are ranked 1-50 in the QS World University Rankings. Further, we find two strategies through which firms increased their presence in AI research: first, they have increased firm-only publications; and second, firms are collaborating primarily with elite universities. Consequently, this increased presence of firms and elite universities in AI research has crowded out mid-tier (QS ranked 201-300) and lower-tier (QS ranked 301-500) universities. To provide causal evidence that deep learning’s unanticipated rise resulted in this divergence, we leverage the generalized synthetic control method, a data-driven counterfactual estimator. Using machine learning based text analysis methods, we provide additional evidence that the divergence between these two groups – large firms and non-elite universities – is driven by access to computing power or compute, which we term as the “compute divide”. This compute divide between large firms and non-elite universities increases concerns around bias and fairness within AI technology, and presents an obstacle towards “democratizing” AI. These results suggest that a lack of access to specialized equipment such as compute can de-democratize knowledge production.

That is a new paper by Nur Ahmed and Muntasir Wahed.

Emergent Ventures winners, eleventh cohort

Andrew Dembe of Uganda, working on the “last mile” problem for health care delivery.

Maxwell Dostart-Meers of Harvard, to study Singapore and state capacity, as a Progress Studies fellow.

Markus Strasser of Linz, Austria, now living in London, to pursue a next-generation scientific search and discovery web interface that can answer complex quantitative questions, built on extracted relations from scientific text, such as graph of causations, effects, biomarkers, quantities, etc.

Marc Sidwell of the United Kingdom, to write a book on common sense.

Yuen Yuen Ang, political scientist at the University of Michigan, from Singapore, to write a new book on disruption.

Matthew Clancy, Iowa State University, Progress Studies fellow. To build out his newsletter on recent research on innovation.

Samarth Athreya, Ontario: “I’m a 17 year old who is incredibly passionate about the advent of biomaterials and its potential to push humanity forward in a variety of industries. I’ve been speaking about my vision and some of my research on the progress of material science and nanotechnology specifically at various events like C2 Montreal, SXSW, and Elevate Tech Festival!”

Applied Divinity Studies, this anonymously written blog has won an award for his or her writing and blogging.  We are paying in bitcoin.

Jordan Mafumbo, a Ugandan autodidact and civil engineer studying Heidegger and the foundations of liberalism.  He also has won an award for blogging.

Blog and Substack contest winners

To date there are three:

1. Anton Howes for his Substack Age of Invention.  He is a historian of invention, often but not exclusively focusing on the eighteenth century, here is Anton on Twitter.  As a separate matter, don’t forget Anton’s excellent recent book Arts & Minds: How the Royal Society of Arts Changed a Nation.

2. Works in Progress.  Here is their About page: “Works in Progress is a new online magazine dedicated to sharing novel ideas and stories of progress, and features original writing from some of the most interesting thinkers in the world.”  The major individuals behind Works in Progress are Sam Bowman, Saloni Dattani, Ben Southwood, and Nick Whitaker, all with bios at the previous link, all strong intellectual forces.

Note also: “Works in Progress is always looking for new writers for upcoming issues and our blog. Reach out if you want to talk about writing for us, with a short summary or abstract of your piece.”

3. Alice Evans, lecturer at King’s College London.  Here is Alice on Twitter.  She is working on “”THE GREAT GENDER DIVERGENCE” What explains global variation in gender relations?” and here is her related blog on that same topic.  Here is her famous post on gender relations in north vs. south India.  Her home page also links to her podcast.

I do expect there will be further awards, and I will keep you posted (here is the original announcement).  If you just started writing a blog and submitted, you may still be in the running for the future. In the meantime, congratulations to these winners!

Vitalik Buterin’s Conversation with me

He is interviewing me, and yes he does close with a bout of Overrated vs. Underrated.  Here is the YouTube link, it starts at about 7:00, give or take a few seconds.  I thought it was very interesting, on both of our sides, more of a dialogue than an interview, the points of focus being crypto and tech utopianism.

Perhaps the next time we will get to The New Monetary Economics…

Novid — a pre-exposure notification system for Covid (and other things)

I find the (short) video easiest to follow:

Best of all, it is incentive-compatible. The founder Po-Shen Loh, a mathematician from Carnegie Mellon, wrote to me:

…for each positive case, don’t just ask the direct contacts to quarantine; instead, tell everyone how many relationships away COVID just struck (e.g., “3” is a contact of a contact of a contact). Then animate this over time like a weather radar…Keep everything anonymous.

Suddenly, the main purpose of the intervention is no longer to protect others from you (quarantining after being exposed). Instead, it is to directly protect you from others, because that early warning of approaching COVID lets you know it’s a good time to wear a better mask, or to be more vigilant about distancing, because the situation is getting hot. This appeals to self-protection instincts instead of altruistic instincts. Since this app is already in deployment, we know anecdotally, for example, of a person who installed the app because his kid was going to a university that was using the app. Why? So that he could be alerted in case COVID started spreading his way from the university via his kid.

Here is his associated preprint.  As economists, ought we not to feel that appealing to self-interest and love of family sometimes works?

Threadhelper, a new method for improving Twitter

Vasco Queirós, from the 7th cohort of Emergent Venture grant winners, has, with Francisco Carvalho, launched ThreadHelper – “A serendipity engine on the Twitter sidebar”. You can get it here.

ThreadHelper is a browser extension that finds you the tweets you need. It shows as a sidebar on the right hand side of your Twitter timeline and instantly and automatically searches bookmarks, retweets, and your past tweets for tweets that are semantically relevant to the tweet being composed. It can be used as a specific search that’s faster than Twitter’s or as a fuzzy search tool.

When in doubt, ask Alex Rampell

Alex is very very highly rated by those who know him, but still in the broader scheme of things significantly underrated as a Bay Area tech and finance thinker.  Here is Alex on SPACs:

If the best fundraise (IPO included) aggregates as many potential buyers as possible to raise money at the highest price with the least dilution and lowest fees, it’s hard to understand how a SPAC represents an improvement against those constraints. When a SPAC merges with a target (“de-SPACs”), it’s tantamount to an IPO. The SPAC (already publicly traded, with lots of cash on its balance sheet) and the target company agree on a pre-money valuation for the target; the money sitting in the SPAC becomes the “money raised” (IPO equivalent) with typically a PIPE (Private Investment in Public Equity, a further institutional fundraise / large block sale) done at the same time. As an example, a $500M SPAC might merge with a private company, ascribing a $4.5B valuation to the private company, meaning a $5B post-money valuation of the combined entity. How do we know *this* is the fair price? Should a company meet with one SPAC? Two SPACs? Three SPACs or four? Where is the price discovery?

And while the fee structure of SPACs will likely change, right now it is indisputably a more expensive option with less price discovery. Bankers are paid 5%+ for taking the SPAC public; SPAC investors typically get warrants with their investment; the SPAC sponsor typically gets 20% (of the pre-merger value of the SPAC) for finding a target, so 2% in the above example; a banker is normally hired and paid to handle the merger; a mini-roadshow happens to get approval from the SPAC shareholders AND to potentially secure more cash in the form of a PIPE, for which a banker is also paid.

Most of the post is on why IPOs are less inefficient than you might think, informative and interesting throughout.

Should government improve the quality of its software?

From a GitHub repository:

This GitHub repository is a back up of my FAERSFix scripts.

The FDA Adverse Effect Reporting System is a horrifically dysfunctional quagmire of shockingly bad data. The data is not just bad for severe epistemological reasons, it is also poorly organized and riddled with flagrant absurd errors.

These scripts smooth over the very messy process of acquiring and basic debugging of the data. At the end of the process a user can arrive at a local repository of the FAERS data that is sane enough to begin to think about some kind of sensible analysis. To understand the disastrous state of the original source data, see the source code of the scripts which is designed to be a readable self-documenting manual demonstrating how to correct this mess.

Since the FDA’s gremlins never rest, these scripts will become obsolete. If you would like to contribute updates or fixes, feel free to send me a patch or a pull request. Good luck!

I thank Chris E. for the pointer.

Clubhouse

I’ve tried it a few times, and I think it will become “a thing.”  You can read about it here, though it is time someone did a more current article.  It is the best forum invented to date for mid-size, friendly, intellectual chats.  Broadly speaking imagine a Zoom call, with competing topic-named rooms, the video turned off, and better queuing and calling upon people procedures.  It doesn’t seem to induce fatigue the way Zoom calls do.  The software has a fluidity and ease of use that I hardly ever see, as usually I hate new apps that people tell me to try.

I don’t know that it will ever be “my thing,” mostly because I can read so quickly, which raises my opportunity cost of consumption.  (Though you can read with it on in the background — is it ever the case that no one in the audience is listening?  Would that even matter?)  In any case, I suspect it will take some real mind-share away from Twitter and Facebook, and now is the time for you to start learning about it.

So far it is invite-only, though I assume it will be opening up more broadly.  Furthermore, an invitation is not so difficult to get.  Arguably there is a problem with who gets invited or deplatformed, though so far this seems to bother the (non-participating) people on Twitter more than it disturbs anyone performing on Clubhouse itself.

It is much more Tiebout-like than Zoom, so someday this also may be an incredible data source on what leads to useful conversations, which are the best governance rules, and so on.

What our new Chinese overlords desire for us

Fleetwood Mac’s Rumours is a top 10 album this week — more than four decades after its release — thanks to a viral TikTok video that’s had everyone vibing along to “Dreams.” Rumours now ranks seventh on the Billboard 200 chart, the publication announced last night, the album’s first appearance in the top 10 since 1978, a year after it debuted.

Rumours’ newfound popularity is thanks to a viral video from Nathan Apodaca, who goes by 420doggface208 on TikTok, that shows him skateboarding down a road and sipping cran-raspberry juice straight from the jug, while “Dreams” plays over top. It’s been viewed more than 60 million times since being posted at the end of September, and it’s even inspired both Mick Fleetwood and Stevie Nicks to sign up for TikTok over the past two weeks. Fleetwood recreated the viral video himself, while Nicks posted a video of herself lacing up roller skates and singing along. They have a combined 35 million views.

Here is the full story, via Fernand Pajot.

The AstroZeneca saga, according to one source

This seems unconfirmed, and do note some sources in the story do not believe this account, but here goes:

AstraZeneca, whose Phase 3 coronavirus vaccine clinical trial has been on hold for more than a month, did not get critical safety data to the US Food and Drug Administration until last week, according to a source familiar with the trial.

The FDA is considering whether to allow AstraZeneca to restart its trial after a participant became ill. At issue is whether the illness was a fluke, or if it may have been related to the vaccine.

The source said the root of the delay is that the participant was in the United Kingdom, and the European Medicines Agency and the FDA store data differently.

“They had to convert data from one format to another format. It’s like taking stuff off a PC and putting it onto an Apple. They had to spend a lot of hours to get what they wanted,” the source said.

On Friday, a federal official hinted there might be some word this week on the trial’s future.

Or maybe they just fooled CNN with it?

Otherwise, good thing we are kept safe from such dangerous data formats!  Would it really not be better to move to reciprocal recognition procedures?  Not to mention a unified data format, or perhaps some FDA methods to read data produced for the EU?

For the pointer I thank Jackson Stone.

Could we detect it if we are living in a simulation?

“If quantum computing actually materializes, in the sense that it’s a large scale, reliable computing option for us, then we’re going to enter a completely different era of simulation,” Davoudi says. “I am starting to think about how to perform my simulations of strong interaction physics and atomic nuclei if I had a quantum computer that was viable.”

All of these factors have led Davoudi to speculate about the simulation hypothesis. If our reality is a simulation, then the simulator is likely also discretizing spacetime to save on computing resources (assuming, of course, that it is using the same mechanisms as our physicists for that simulation). Signatures of such discrete spacetime could potentially be seen in the directions high-energy cosmic rays arrive from: they would have a preferred direction in the sky because of the breaking of so-called rotational symmetry.

Telescopes “haven’t observed any deviation from that rotational invariance yet,” Davoudi says. And even if such an effect were to be seen, it would not constitute unequivocal evidence that we live in a simulation. Base reality itself could have similar properties.

Here is further discussion from Anil Anathaswamy.  Via Robert Nelsen.  As you may already know, my view is that there is no proper external perspective, and the concept of “living in a simulation” is not obviously distinct from living in a universe that follows some kind of laws, whether natural or even theological.  The universe is simultaneously the simulation and the simulator itself!  Anything “outside the universe doing the simulating” is then itself “the (mega-)universe that is simultaneously the simulation and the simulator itself.” etc.

What should I answer Vitalik Buterin?

Not a CWT, as he will be interviewing me, but likely a discussion too, who knows what I may fling back at him?

What do you all recommend I read to prepare?  I prepared a great deal for my CWT with Vitalik a few years ago (one of my favorite episodes), but what is new with Ethereum and blockchain and other matters since then?

I thank you all in advance for your usual wise counsel and advice.

Are Nobel Prizes worth less these days?

It would seem so, now there are lots of them, here is one part of my Bloomberg column:

The Nobel Peace Prize this year went to the World Food Programme, part of the United Nations. Yet the Center for Global Development, a leading and highly respected think tank, ranked the winner dead last out of 40 groups as measured for effectiveness. Another study, by economists William Easterly and Tobias Pfutze in 2008, was also less than enthusiastic about the World Food Programme.

The most striking feature of the award is not that the Nobel committee might have gotten it wrong. Rather, it is that nobody seems to care. The issue has popped up on Twitter, but it is hardly a major controversy.

I also noted that the Nobel Laureates I follow on Twitter, in the aggregate, seem more temperamental than the 20-year-olds (and younger) that I follow.  Hail Martin Gurri!

And this:

The internet diminishes the impact of the prize in yet another way. Take Paul Romer, a highly deserving laureate in economics in 2018. To his credit, many of Romer’s ideas, such as charter cities, had been debated actively on the internet, in blogs and on Twitter and Medium, for at least a decade. Just about everyone who follows such things expected that Romer would win a Nobel Prize, and when he did it felt anticlimactic. In similar fashion, the choice of labor economist David Card (possibly with co-authors) also will feel anticlimactic when it comes, as it likely will.

Card with co-authors, by the way, is my prediction for tomorrow.