Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: [$20K in Prizes] AI Safety Arguments Competition, published by Dan Hendrycks on April 26, 2022 on The AI Alignment Forum.

TL;DR—We’re distributing $20k in total as prizes for submissions that make effective arguments for the importance of AI safety. The goal is to generate short-form content for outreach to policymakers, management at tech companies, and ML researchers. This competition will be followed by another competition in around a month that focuses on long-form content.

This competition is for short-form arguments for the importance of AI safety. For the competition for distillations of posts, papers, and research agendas, see the Distillation Contest.

Objectives of the arguments

To mitigate AI risk, it’s essential that we convince relevant stakeholders sooner rather than later. To this end, we are initiating a pair of competitions to build effective arguments for a range of audiences. In particular, our audiences include policymakers, tech executives, and ML researchers.

Policymakers may be unfamiliar with the latest advances in machine learning, and may not have the technical background necessary to understand some/most of the details. Instead, they may focus on societal implications of AI as well as which policies are useful.

Tech executives are likely aware of the latest technology, but lack a mechanistic understanding. They may come from technical backgrounds and are likely highly educated. They will likely be reading with an eye towards how these arguments concretely affect which projects they fund and who they hire.

Machine learning researchers can be assumed to have high familiarity with the state of the art in deep learning. They may have previously encountered talk of x-risk but were not compelled to act. They may want to know how the arguments could affect what they should be researching.

We’d like arguments to be written for at least one of the three audiences listed above. Some arguments could speak to multiple audiences, but we expect that trying to speak to all at once could be difficult. After the competition ends, we will test arguments with each audience and collect feedback. We’ll also compile top submissions into a public repository for the benefit of the x-risk community.

Note that we are not interested in arguments for very specific technical strategies towards safety. We are simply looking for sound arguments that AI risk is real and important.

Competition details

The present competition addresses shorter arguments (paragraphs and one-liners) with a total prize pool of $20K. The prizes will be split among, roughly, 20-40 winning submissions. Please feel free to make numerous submissions and try your hand at motivating various different risk factors; it's possible that an individual with multiple great submissions could win a good fraction of the prize. The prize distribution will be determined by effectiveness and epistemic soundness as judged by us. Arguments must not be misleading.

To submit an entry:

Please leave a comment on this post (or submit a response to this form), including:

The original source, if not original.

If the entry contains factual claims, a source for the factual claims.

The intended audience(s) (one or more of the audiences listed above).

In addition, feel free to adapt another user’s comment by leaving a reply⁠⁠—prizes will be awarded based on the significance and novelty of the adaptation.

Note that if two entries are extremely similar, we will, by default, give credit to the entry which was posted earlier. Please do not submit multiple entries in one comment; if you want to submit multiple entries, make multiple comments.

The first competition will run until May 27th, 11:59 PT. In around a month, we’ll release a second competition for generating longer “AI risk executive summaries'' (more details to come). If y...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Mid-career people: strongly consider switching to EA work, published by Ben Snodin on April 26, 2022 on The Effective Altruism Forum.

Or alternatively: “A message for 2018-me: quit your job and start doing ‘EA work’”

Note: I wrote this post quite quickly (~9 hours), and it’s quite rough. In particular, it’s more likely than normal that some bits are confusing, that where I’ve provided external links these aren’t the most relevant links, etc. If this turns out to be really useful for people, I can try to give some better advice — another reason to get in touch if this resonates with you!

In this post I want to provide encouragement and information for mid-career people who are sympathetic to EA ideas but haven’t seriously tried doing EA work. Basically, I think there’s tons of amazingly impactful, fun, well-compensated work that skilled mid-career people could do, and that this is maybe much less obvious from the “outside” than it is for a relative EA-insider like me.

Note that I focus on longtermist EA work here because this is what I know about. But I imagine an EA-insider who is in another EA area might share similar sentiments. Similarly, while I focus on mid-career people here, I imagine a lot of what I’ll say will be relevant for people at other career stages.

Here’s what I’ll say:

Doing longtermist EA work right now is extremely valuable, and there’s a wide variety of roles and skills needed and really exciting projects to work on.

I had a lot of misconceptions before I moved into EA work. I’ll describe what these were and what I now think the reality is as a relative EA-insider.

Finally, I’ll admit that there are some downsides to switching to EA work right now.

If you’re a mid-career person interested in switching to EA work, I’d be interested to chat and maybe help. Please get in touch by sending me a private message on the Forum or emailing me at hello[at]bensnodin dot com.

Doing longtermist EA work right now could be very fun and well-compensated, as well as very valuable

There is an absolute ton of very valuable work that needs to happen right now, such as:

Helping to deliver some technology or infrastructure that makes a meaningful dent on biorisk. (Some examples here.)

Making progress on technical AI safety (e.g. by working at Redwood Research).

Doing high-level cause prioritisation work (like Open Philanthropy's worldview investigations).

Helping incubate longtermist megaprojects (e.g. by working at Rethink Priorities).

Based on my experience working with EAs, I’d expect that you’d do this work with incredible colleagues: people who are driven, passionate, kind, very capable, and who share your goals.

Similarly, many EA organisations seem to put a large emphasis on employee wellbeing and on creating an excellent working environment, and I’ve personally been very impressed by this at the EA organisations I’ve worked for so far.

Also, in case you haven’t heard, there is now a lot of money available to fund longtermist EA work, which allows longtermist EA organisations to pay more than in the past. My impression is that salaries are often in the $60k-150k range (for a concrete example, see the researcher pay ranges on Rethink Priorities’ website). (Side note: if salary is blocking you from doing longtermist EA work, please talk to someone, e.g. me.)

In addition, I think mid-career people have a huge amount to offer thanks to having pre-existing skills and being able to hit the ground running when starting out — this is especially valuable at the moment given the relative lack of management time available. There seems to be a need for a wide range of skills, and I expect this need to increase as we start seeing more longtermist EA projects that are “doing things in the world” (like Alvea). You don’t have to be a researcher!

Finally, I think EA community cu...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: How many EAs failed in high risk, high reward projects?, published by mariushobbhahn on April 26, 2022 on The Effective Altruism Forum.

The ideas of high risk, high reward projects, value in the tails, etc. are quite common EA positions now. People are usually reminded that they have a low probability of success and that they should expect to fail most of the time. However, most people I know/have heard of who started ambitious EA projects are doing quite well. Examples would be SBF, Anthropic, Alvea, and many more. My question, therefore, is: Is the risk of failure lower than we expected, or do I just not know the failures? Do I just know the selection of people who succeeded? Is it too early to tell if a project truly succeeded? If so, what are concrete examples of EAs or EA orgs not meeting high expectations despite trying really hard? Is it possible that we just underestimate how successful someone with an EA mindset and the right support can be when they try really hard?

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Participate in the Hybrid Forecasting-Persuasion Tournament (on X-risk topics), published by Jhrosenberg on April 25, 2022 on The Effective Altruism Forum.

Tl;dr: Phil Tetlock's research team is running a tournament that will team up superforecasters with subject-matter experts to produce accurate forecasts and informative rationales on existential risk topics. If you would like to express interest in participating, please answer some initial questions about your views on X-risk in this form by May 13th.

More on the tournament

The Hybrid Forecasting-Persuasion Tournament will explore potential threats to humanity in the next century, with a focus on artificial intelligence, biosecurity, climate, and nuclear war in the short-term (<3 years), medium-term (10 years), and long-term (30+ years). Participants will also work in teams to craft high-quality explanations of their forecasts that persuade others to change their views—in the right direction. Sample questions are here.

We are looking for participants who:

Have studied existential risk topics, formally or informally. If you're unsure whether you qualify, we encourage you to submit the form below anyway. We intend to include participants with a wide range of relevant experience.

Can devote at least 3 hours per week to forecasting for 12-16 weeks beginning in May.

We will provide more detail on compensation for participation after selection process but it will range from $2,000 up to $10,000, depending on your level of engagement and accuracy.

To express your interest, please fill out this form by May 13th. Ten respondents who complete this survey will receive a $2,000 prize (and the chances of winning will rise as a function of the accuracy of your answers to certain forecasting questions in the survey). Note that space is limited and we will follow up with as many people as we can accommodate.

Why participate?

Impact. We aim to improve society’s understanding of topics that could hugely affect many generations to come.

Will MacAskill says, “I’m very excited to further work on forecasting existential risks, and think that this is an excellent team to drive this project forward. I think that participating in this tournament is a very high-impact activity.”

Luke Muelhauser says, “I read [an internal summary] of the hybrid forecasting+persuasion tournament. The tournament wasn't funded by Open Philanthropy, but based on that summary I think it's an exciting project, and has the potential to produce valuable forecasts of variables related to x-risk, and test the utility of some recent innovations in forecasting methodology (e.g. reciprocal scoring). I also expect to fund later x-risk-related forecasting work that would ideally learn from the results of this tournament. I say all this to help subject matter experts assess the value of contributing their expertise to this project in small or large ways."

Phil Tetlock says, “This tournament will explore how much predictive value we can extract from human judgment in ultra-high-stakes domains that some thoughtful skeptics have written off as inherently indeterminate.”

Curiosity and engagement. The tournament will be fun. You first work independently and then on teams of other talented people. Expect lots of teachable moments.

Professional development. Forecasting skills are widely useful. This tournament will include a training module (CHAMPS KNOW) that was shown in a randomized trial to boost forecasting accuracy—plus participation itself will provide valuable practice.

More information

A more detailed write-up about the tournament is here. If you have questions, please email [email protected].

We would also greatly appreciate if you share this with others who may be interested.

Thanks for listening. To help us out with The Nonlinear Library or to learn more...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: The Fabian society was weirdly similar to the EA movement, published by MathiasKB on April 26, 2022 on The Effective Altruism Forum.

I recently ran into this book review by Scott Alexander. While reading the book review I kept being struck by passages that could just as well have been describing the EA movement. The Fabian movement was an influential British socialist movement. They managed to get their members into various important positions in society to lobby for their, at the time, radical socialist agenda demanding free public education, women's suffrage, eight hour work days and more.Even its critics sounded familiar, arguing that the Fabian society was nothing more than a talking-club the privileged!

A few highlights from the review:

Whatever the Fabians’ other advantages, they arose at a really good time to be a socialist thinker. There was a sort of feeling in the air that socialism was the wave of the future, that there were literally no good arguments whatsoever against it, that you were either an intellectual (in which case it was obvious that socialism was better) or you were just so thoughtless that you had never even considered the matter at all

He didn’t expect forceful action – recruitment campaigns, branch organizations, or the like – to have any effect. Instead, he favored a soft touch. Have the sort of intellectual atmosphere that talented people would be attracted to. Gradually draw them in with interesting social and intellectual activities. Once they’re attached, get them in on the first rung of some ladder or other – local politics, informal debate, small-time pamphlet writing. Have a few geniuses around who can recognize other geniuses. Then have positions to put people in once they’re worthy of it – whether it’s the lecture circuit, the propaganda business, or a university for them to teach at.

The Society’s members tended to be strange people, mercurial and hard to keep on task. Occasionally they would get distracted and forget about communism entirely, running off to worry about art or philosophy or ghost-hunting or something.

Pease was generally unimpressed with how these worked out. He felt experience had proven most of the people based outside London to be intellectually second-rate and without much to contribute.

It seems to me there is a ton to learn from the history of the Fabian society. Both from its successes and shortcomings. I'm curious to hear what you think.

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: EA coworking/lounge space on gather.town, published by Arepo on April 26, 2022 on The Effective Altruism Forum.

PSA: I'm trying to build a community of online EA coworkers with longer-lasting connections than Focusmate and more flexibility than Discord. To this end I've created a general EA gather.town space.

My hope is that it can eventually become an always-active meeting grounds for anyone who wants to both cowork with and actually get to know other EAs to join at any time of day - perhaps like a neverending EAGx.

To help build up the community vibe, some of the hardcore users are keeping it open in the background whenever we're working to increase the probability of you finding at least one person in there. So if you see anyone with a 'happy to chat' status or similar, feel free to come over (username and indeed real name = Sasha) - and if you want to support it, it would be great if you could just hang out on there yourselves! Hope to meet you there soon ❤

Update

We've had very positive feedback so far, and seen surprising growth over the past week, so here's a more in-depth look, to encourage more people to join and keep the snowball rolling!

Features & Norms

The setup and norms are constantly evolving, but at the time of writing, here's a taster (though there's plenty more easter eggs to explore):

Plans & Support

We have started expanding the space to make sure we have plenty of room to grow without feeling crowded, and we should be able to source funding for more users so please spread the word!

We have two loci for talking about the project: a Facebook group and a subchannel of the EA Co-Working Discord. Feel free to use these places if you want to schedule coworking with people ahead of time. We also have some terminals set up in the actual town that links to google docs with information, and where you can add suggestions and introductions.

We also want to encourage people to host talks or other virtual events in the space, both to help spread awareness of it and as a convenient forum for discussion afterwards:

Please get in touch if you wish to help out or have questions, comments or suggestions! :)

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: dalle2 comments, published by nostalgebraist on April 26, 2022 on LessWrong.

(i.)

On April 6, OpenAI announced “DALLE-2” to great fanfare.

There are two different ways that OpenAI talks to the public about its models: as research, and as marketable products.

These two ways of speaking are utterly distinct, and seemingly immiscible. The OpenAI blog contains various posts written in each of the two voices, the “research” voice and the “product” voice, but each post is wholly one or the other.

Here’s OpenAI introducing a model (the original DALLE) in the research voice:

DALL·E is a 12-billion parameter version of GPT-3 trained to generate images from text descriptions, using a dataset of text–image pairs. [.]

Like GPT-3, DALL·E is a transformer language model. It receives both the text and the image as a single stream of data containing up to 1280 tokens, and is trained using maximum likelihood to generate all of the tokens, one after another. This training procedure allows DALL·E to not only generate an image from scratch, but also to regenerate any rectangular region of an existing image that extends to the bottom-right corner, in a way that is consistent with the text prompt.

And here they are, introducing a model (Codex) in the product voice:

OpenAI Codex is a descendant of GPT-3; its training data contains both natural language and billions of lines of source code from publicly available sources, including code in public GitHub repositories. [.]

GPT-3’s main skill is generating natural language in response to a natural language prompt, meaning the only way it affects the world is through the mind of the reader. OpenAI Codex has much of the natural language understanding of GPT-3, but it produces working code—meaning you can issue commands in English to any piece of software with an API. OpenAI Codex empowers computers to better understand people’s intent, which can empower everyone to do more with computers.

Interestingly, when OpenAI is planning to announce a model as a product, they tend not to publicize the research leading up to it.

They don’t write research posts as they go along, and then write a product post once they’re done. They do publish the research, in PDFs on the Arxiv. The researchers who were directly involved might tweet a link to it. And of course these papers are immediately noticed by ML geeks with RSS feeds, and then by their entire social networks. It’s not like OpenAI is trying to hide these publications. It’s just not making a big deal out of them.

Remember when the GPT-3 paper came out? It didn’t get a splashy announcement. It didn’t get noted in the blog at all. It was just dumped unceremoniously onto the Arxiv.

And then, a few weeks later, they donned the product voice and announced the “OpenAI API,” GPT-3 as a service. Their post on the API was full of enthusiasm, but contained almost no technical details. It mentioned the term “GPT-3″ only in passing:

Today the API runs models with weights from the GPT-3 family with many speed and throughput improvements. Machine learning is moving very fast, and we’re constantly upgrading our technology so that our users stay up to date.

It didn’t even mention how many parameters the model had!

(My tone might sound negative. Just to be clear, I’m not trying to criticize OpenAI for the above. I’m just pointing out recurring patterns in their PR.)

(ii.)

DALLE-2 was announced in the product voice.

In fact, their blog post on it is the most over-the-top product-voice-y thing imaginable. It goes so far in the direction of “aiming for a non-technical audience” that it ends up seemingly addressed to small children. As in a picture book, it offers simple sentences in gigantic letters, one or two per page, each one nestled between huge comforting expanses of blank space. And sentences themselves are . well, st...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: A major update in our assessment of water quality interventions, published by GiveWell on April 26, 2022 on The Effective Altruism Forum.

As we continue to grow, GiveWell seeks to maximize both the cost-effectiveness of the funding we direct and the likely room for more funding of the programs we support. We think we’ve identified a category of interventions that rates really well on both: water treatment, such as chlorination.

This is a major update for us. Before 2020, based on the available evidence, we didn’t believe that water quality interventions had a large enough effect on mortality to make them a competitive target for funding. We’ve since seen new evidence that has led us to significantly increase our estimate of the mortality reduction in young children that’s attributable to these interventions: a 14% reduction in mortality from any cause,[1] up from around 3%.

Though we have remaining uncertainties about these numbers, we’ve substantially updated our view of the promisingness of water treatment. Where we previously found that Evidence Action’s Dispensers for Safe Water program was about as cost-effective as unconditional cash transfers, we now believe it’s about four to eight times as cost-effective, depending on the location. That was a primary factor in our decision to recommend a grant of up to $64.7 million to Dispensers for Safe Water in January 2022.

We’re sharing this news in brief form before we’ve published a grant page, because we’re excited about the potential of this grant and what it represents. It’s an area of work we haven’t supported to a significant degree in the past, but one that we now think could absorb hundreds of millions of dollars in funding for cost-effective programming.

The problem and the intervention

In low-income settings, contaminated water is a major cause of diarrhea, a leading cause of death in children under five years old.[2] Several interventions exist to either purify water or protect it from contamination in the first place, but chlorination has a number of features that made it an attractive intervention for us to explore: it is inexpensive and widely used, several charities are already set up to implement it, and the technology behind it is well established.[3]

Chlorine is well-known as a disinfectant; it reacts with disease-causing microorganisms in water, inactivating viruses and bacteria.[4] There is evidence that chlorination programs, such as distributing chlorine to households, reduce diarrhea in children, but there had been scant evidence that such interventions reduce mortality, which is the single largest driver of cost-effectiveness in our models of water quality interventions.[5] Before 2020, we estimated the effect chlorination had on the prevalence of diarrhea, and then extrapolated from that effect to arrive at an indirect mortality reduction estimate of roughly 3%.[6] This translated to a low cost-effectiveness estimate relative to the programs GiveWell typically funds.

What led us to update

In mid-2020, Michael Kremer, one of the researchers who conducted the initial trial studying dispensers for safe water, and now a researcher at the University of Chicago who has extensively studied global health interventions, shared with us preliminary results from his team’s new meta-analysis of mortality data from randomized controlled trials (RCTs) of water quality interventions.[7] (The working paper has since been published, here.) Pooling studies allowed Kremer’s team to obtain statistically significant results from studies that weren’t individually large enough to produce such results on their own. These results suggest that water treatments have a significant effect on mortality from any cause (“all-cause mortality”) in children under five—Kremer et al. currently estimate a roughly 25% reduction.[8]

T...

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Why pessimism sounds smart, published by jasoncrawford on April 25, 2022 on LessWrong.

Pessimists sound smart. Optimists make money.–Nat Friedman (quoted by Patrick)

I’ve realized a new reason why pessimism sounds smart: optimism often requires believing in unknown, unspecified future breakthroughs—which seems fanciful and naive. If you very soberly, wisely, prudently stick to the known and the proven, you will necessarily be pessimistic.

No proven resources or technologies can sustain economic growth. The status quo will plateau. To expect growth is to believe in future technologies. To expect very long-term growth is to believe in science fiction.

No known solutions can solve our hardest problems—that’s why they’re the hardest ones. And by the nature of problem-solving, we are aware of many problems before we are aware of their solutions. So there will always be a frontier of problems we don’t yet know how to solve.

Fears of Peak Oil and other resource shortages follow this pattern. Predictions of shortages are typically based on “proven reserves.” We are saved from shortage by the unproven and even the unknown reserves, and the new technologies that make them profitable to extract. Or, when certain resources really do run out, we are saved economically by new technologies that use different resources: Haber-Bosch saved us from the guano shortage; kerosene saved the sperm whales from extinction; plastic saved the elephants by replacing ivory.

In just the same way, it can seem that we’re running out of ideas—that all our technologies and industries are plateauing. Technologies do run a natural S-curve, just like oil fields. But when some breakthrough insight creates an entirely new field, it opens an entire new orchard of low-hanging fruit to pick. Focusing only on established sectors and proven fields thus naturally leads to pessimism. To be an optimist, you have to believe that at least some current wild-eyed speculation will come true.

Why is this style of pessimism repeatedly wrong? How can this optimism be justified? Not on the basis of specific future technologies—which, again, are unproven—but on the basis of philosophical premises about the nature of humans and of progress. The possibility of sustained progress is a consequence of the view of humans as “universal explainers” (cf. David Deutsch), and of progress as driven fundamentally by human choice and effort—that is, by human agency.

The opposite view is that progress is a matter of luck. If the progress of the last few centuries was a random windfall, then pessimism is logical: our luck is bound to run out. How could we get that lucky again? If the next century is an average one, it will see little progress.

But if progress is a primarily matter of agency, then whether it continues is up to us.

Thanks for listening. To help us out with The Nonlinear Library or to learn more, please visit nonlinear.org.

Welcome to The Nonlinear Library, where we use Text-to-Speech software to convert the best writing from the Rationalist and EA communities into audio. This is: Solving the replication crisis (FTX proposal), published by Michael Wiebe on April 25, 2022 on The Effective Altruism Forum.

Here's my rejected FTX proposal (with Abel Brodeur) to solve the replication crisis by hiring full-time replicators. (I left out the budget details.)

Please describe your project in under 100 words.

We will actually solve the replication crisis in social science by hiring a “red team” of quantitative researchers to systematically replicate new research. Currently, there are few penalties for academics and journals that publish unreliable research, because few replications are attempted. We will fundamentally change academic incentives by making researchers know that their work will be scrutinized, which will motivate them to improve research design, or else face a loss of reputation. By fixing scientific institutions now, we can reap the compounding benefits of reliable knowledge over the long-term future.

If the project has a website, what’s the URL?

Please describe what you are doing very concretely—not just goals and long-term vision, but specifically what you are doing in the next few months.

Currently, the Institute for Replication is using volunteers to systematically reproduce and replicate new studies from leading journals in economics and political science. With funding from FTX, we can hire a Project Scientist (Michael Wiebe), post-docs, and research assistants to massively scale-up reproductions and replications. We can also launch a cash prize for completed replications, to incentivize even more replications. This can be implemented in several ways; for example, giving a prize of ~$1000 for high quality replications completed using the Social Science Reproduction Platform, as judged by a panel of experts.

What’s the case for your project?

Social science is facing a replication crisis. Researchers produce unreliable findings that often do not replicate, and the root problem is the lack of replications.

Academics have basically no incentive to perform replications, since they usually do not yield original findings, and are not valued by journals. Since they do not lead to publications, replications do not help academics get tenure, and hence few are attempted. The replications that are done are conducted by volunteers in their spare time, and can even have negative career effects if they upset powerful academics.

The rareness of replications causes peer review to be an inadequate form of quality control. Knowing that research won’t be closely scrutinized, journals and referees have little incentive to check for data quality issues, coding errors, or robustness. If a paper with unreliable findings gets published, the journal suffers no loss in reputation, because no one will replicate the paper to expose its flaws. Hence, referees take empirical results at face value, and focus instead on framing the research question and appropriately citing the literature.

Knowing that their work will not be reproduced nor replicated, most researchers don’t invest time in preparing replication packages, and don’t check for data or coding errors. The result is entire fields with serious reproducibility problems.

We can fix these incentives by investing heavily in reproduction and replication, and making a big push to systematically replicate new research. With a team of full-time replicators and cash prizes for completed replications, researchers will now expect their work to be immediately scrutinized as a regular practice. This scrutiny will put researchers’ reputations on the line: if their findings are not robust, their work will not be cited (or worse, be retracted), ultimately affecting their promotion and tenure outcomes. At the same time, high-quality work will be rewarded. A big push will attract widespread attention to amplify these reputati...