open science framework

Grants-in-Aid for Scientific Research → Public Goods → Open Science

  • Providing data is not an achievement.
    • evaluation system
      • Getting the data ready to be released, hassle
      • If you don’t make a profit by putting it out, there is no incentive to put it out at a cost.
    • Being the only one with data is a source of competitive advantage.
  • Journals that publish data, etc.
    • Gradual changes in the ecosystem

gpt.iconThe Allen Foundation, established in 1988, is an American charitable organization that supports research and programs primarily related to health and nutrition. The founder is Paul G. Allen, and the Foundation primarily provides grants for scientific research and educational programs in the areas of health improvement and nutrition. In particular, the Foundation supports projects aimed at improving dietary habits and helping to promote health among children and the elderly.

  • The scope of the Foundation’s activities is international, with a particularly large number of grants to universities and research institutions in the United States. It also focuses on topics such as improving food quality, nutrition education, and reducing health disparities.

  • Paul G. Allen is best known as the co-founder of Microsoft, and his foundation was established as part of his personal philanthropy. His other philanthropic endeavors include the Allen Institute for Brain Science, which has made significant contributions to brain research and the medical field

  • National Brain Research Centre - Wikipedia

  • Paris Brain Institute | Sorbonne UniversitĂ©| SantĂ© Allen Brain Atlas - Wikipedia

  • The Allen Institute for Brain Science is known for its detailed research and data provision on the mouse brain. In particular, the Allen Mouse Brain Atlas is a high-resolution 3D atlas that provides detailed mapping of gene expression throughout the mouse brain and is used to elucidate the genetic characteristics of cells in each region of the brain. This data has contributed significantly to the advancement of neuroscience and is an important resource for understanding brain structure and function.

  • The Allen Institute’s efforts are also being used in research related to Parkinson’s disease, Alzheimer’s disease, and neurological disorders such as autism, using mouse brain models as a basis for exploring the genetic and neurophysiological basis of these diseases. It should also be noted that the atlas is published as open data and freely available to researchers worldwide (

  • Scientists unveil first complete cellular map of adult mouse brain - Allen Institute

  • A new high-resolution, 3D map of the whole mouse brain - Allen Institute

  • de facto standard of how to divide a rat’s brain, achieved by taking a large amount of data and making it public with 100 billion~ in funding.

  • POC and PMF are different.

Need [data sharing infrastructure

I want to see the system solve problems.

People, Tools, Data

Data that cannot be made public

  • For example, information on the location of people in need of assistance who are unable to walk alone at the time of an earthquake is important for rescue efforts.
  • On the other hand, since it is personal information, it cannot be disclosed, and since there is a risk of criminal use, the disclosure must be to a limited number of people with verified identities.
    • So this is where it leads to the Reserve Self-Defense Force.

Visualization of Contributions

Co-Living

People who were originally doing blockchain in China are now going to Chiang Mai because it’s too hard to do.

  • Since the basis of life is in the crypto world, the location in the physical world doesn’t matter much.

  • development camp

  • When the “development camp” concept first started to gain momentum, accommodations with Internet access were rare.

  • Even if the sign says “Internet access available,” it may be ISDN.

  • Or that the router is so crappy that if 20 people connect to the Internet, it will die.

  • So the engineers who were originally active in sharing information through OSS and other means, “Let’s share information on lodging that can be used for development camps.

  • Young people today feel that if they stay in a hotel, they should have wifi, so they don’t recognize it in the “lodging available for development camps” bracket.

Data Utilization Infrastructure

Protocol Labs and Neurotech.

Crypto Subsidy

  • Several grants are done in different ways, but no comparison has been made.

    • It would be nice to know “what way is more efficient” if we can compare them.
  • Gitcoin

  • Optimism

    • They elect fund managers.
    • gpt.iconIn Optimism’s Governance Fund, the community votes to determine the allocation of funds. Specifically, users holding Optimism tokens (OP tokens) can vote for themselves or delegate their voting rights to other delegates (Delegates). This determines the allocation of funds to support project development and ecosystem growth
    • The Token House plays a central role in voting, and its primary role includes the discussion and voting on governance proposals and the allocation of funds for the governance fund. This voting is based on the amount of OP tokens held, and a certain percentage of votes and approvals are required for a proposal to pass. For example, a governance fund proposal requires a minimum of 30% of the token supply to participate in the voting, of which at least 51% must vote in favor of the proposal
    • Through this mechanism, the Optimism community helps projects grow through transparent and decentralized decision-making.
  • DAO Treasury

    • The DAO Treasury, on the other hand, is a pool of funds managed within a DAO (decentralized autonomous organization), where the allocation of funds is determined by token holders voting on proposals. The unique feature of DAO Treasury is that it is fully decentralized, allowing the entire community to be involved in how funds are used. The voting system and governance model may vary from project to project, for example, Quadratic Voting. Flexible operations according to the needs of the community
  • Impact Evaluation

  • Database required A story not limited to crypto

  • EBPM] in the United Kingdom.

    • gpt.iconThe introduction of Evidence-Based Policy Making (EBPM) in the United Kingdom was emphasized after the Labour Party’s Tony Blair government took office in 1997 to improve the quality of policy. The goal is to provide more effective and efficient public services by creating policies based on statistical data and scientific evidence, not just experience and feeling.
  • COVID lockdown could be considered an intervention.

Transparency through on-chain is a lie.

  • Somewhere in the ocean there is a diamond ring.

    • You can take it if you find it.
    • Anyone is free to take it!
  • Cost to obtain is an issue.

  • Evaluation systems are also valued.

  • that’s so

COO needed for visionary issue

  • Create trusting relationships
  • Listening to problems in the field Organizational problems are usually people problems. Skill in appointing good aides

Polis moderation costs are high.

  • Lawyers and such did it.
  • Moderation Automation
    • That’s exactly what Audrey was happy about when Talk to the City was born.
    • Emotional clashes, sudden violence
    • Long monologues out of context
      • I’m sure you have more to tell me.”
        • A sense of being heard
        • Not a very productive narrow view, though,
          • Participants’ “sense of being heard” is increased and self-disclosure that they normally do not engage in takes place.
          • Listening by example enhances participants’ listening attitudes.
        • and other effects make the meeting as a whole a more productive place.
          • interestingnishio.icon
      • Was the listener a community member?
        • Wrong.
        • It’s not that I leaned on him because I knew him originally.
      • San Marino judge comes from Italy
        • Possibly valuable for small communities to be heard by people outside the community
      • gather for discussion
      • I feel like I’ve got it all out.
    • The weight of making one’s point gradually decreases as one gets tired. - It looks like annealing method, doesn’t it?
    • AI has unlimited cognitive fitness.
    • Forgotten Japanese
      • The author (1907-81), a lifelong traveler under the tutelage of Kunio Yanagida and Keizo Shibusawa, has painstakingly researched folklore throughout Japan, unearthed forgotten aspects of Japanese life from the histories made by people with writing, and lovingly describes the nature of the communities that produced and passed on “folktales. A masterpiece of Miyamoto’s folklore, it consists of thirteen stories, including “Tosa Genji” and “Onna no Seken” (A Woman’s World). (Commentary by Yoshihiko Amino)

Independent researcher should be able to eat

  • Incentive to publish results promptly.
  • Direction close to DeSci as well.

Money, ability and will

  • Which resources are missing? - Resources for the enthusiastic
  • I don’t trust people who have money as an incentive.
    • At least not at the stage of looking for a second.
    • If there are 100 people and you’re the 101st, fine.

Pagoda

  • www.4seas.io
  • 4seas
  • FOUR SEASONS
    • They rent out hotels for 3 years, interesting. A high INTENSITY community of about 30 people
  • Community Manager is a scarce resource
  • trauma work
    • Ask how hard it is.
    • Rwanda
  • gpt.icon - Community-based trauma therapy has played a major role in Rwanda since the 1994 genocide. Particular attention has been paid to the "sociotherapy" approach, which is a group therapy aimed at psychological healing and restoring trust in the community. The program provides an opportunity for perpetrators and victims to share the healing process and understand each other in the same space. - [Healing Trauma and Building Trust and Tolerance in Rwanda(PDF)](https://neveragainrwanda.org/wp-content/uploads/2021/11/Healing-Trauma-and-Building-Trust-Draft-report.pdf) - [Societal Healing in Rwanda: Toward a Multisystemic Framework for Mental Health, Social Cohesion, and Sustainable Livelihoods among Survivors and Perpetrators of the Genocide against the Tutsi – Health and Human Rights Journal](https://www.hhrjournal.org/2021/06/societal-healing-in-rwanda-toward-a-multisystemic-framework-for-mental-health-social-cohesion-and-sustainable-livelihoods-among-survivors-and-perpetrators-of-the-genocide-against-the-tutsi/) - Resilience Healing Spaces, which focuses on trauma treatment, is also implemented to promote emotional recovery and strengthen social bonds through group therapy. - [How Community-Based Healing Spaces are Strengthening Mental Resilience in Rwanda - Interpeace](https://www.interpeace.org/2023/02/how-community-based-healing-spaces-are-strengthening-mental-resilience-in-rwanda/) - Such treatment contributes to the regeneration of not only the individual but also the community as a whole, and has been recognized as an important method for mitigating the effects of trauma. - These efforts promote social cohesion and contribute to sustainable peacebuilding.

If there is a mechanism, we can make it a community. - confidence building

Vocabulary to talk about Japanese politics in English

  • Translation in LLM for written text
  • Is there an AI if you want to target people who prefer to talk?nishio.icon
    • Why not create an AI Audrey to answer questions about Plurality in any language (abrupt)?

AOI - vTaiwan - OpenAI Grant

  • [Democratic Inputs to AI

Improvement of TTTC

  • Good with Polis
  • Visualization → a place for people to voice their opinions
  • A “discussion space” that is not a threaded forum
    • Would it really provide customer value?nishio.icon
      • Most people haven’t adapted to the new UI.
      • Is there a need for a system that will listen to you complain for hours on audio?
      • Audrey’s “Talk to the City cluster so we can have a conversation.” Want to have a conversation?

Even if you are enthusiastic, you can’t participate if the onboarding hurdles are too high.

  • image

House of Councillors election

  • Next July?

Process Evaluation

Well-being Well-being

toothbrush issue.

  • I don’t want to use what others have used.
    • With incentives like that, mixed things get made.
    • Interoperability degradation
  • Put it all together and UMAP

Consider the relationship with broad listening.nishio.icon

  • A normal survey (with questions to be answered on a 3 or 5-point scale) requires people to input N natural sentences to obtain an N-dimensional vector.

  • Polis is the same.

    • The difference is that (depending on the setting) people make additions to the question text itself
      • When multiple surveys are bundled together, there is a strong correlation
        • So you want to compress questions into smaller dimensions.
        • If you bundle this with UMAP, it’s almost identical to the clustering portion of Talk to the City.
      • Similarly, there is a strong correlation in Polis when people add new sentences
        • I’m sure there was some mechanism to determine this and present the information in the order that maximizes the amount of information that can be acquired

        • Is it possible that the mechanism here is actually a concept that bridges multiple fields of application?
        • I’d like to get a proper grasp of this with the paper and the code

  • Definition of Happiness I used to think about “Why can we say we need digital democracy?” and came up with the idea that “there is a need for people to define their happiness.”

  • from Let’s Talk with Cybozu Plurality Plurality Practices and Expectations

    • Key Issues: Importance of AI governance vs.
      • If AI becomes smarter than humans, AI should rule.
      • However, the definition of people’s happiness needs to be done by humans.
        • It is not up to one person to decide, but must be datamined that there are diverse forms of happiness
        • Without it, if an AI rules, it will impose someone else’s predetermined form of happiness.
      • So we should go in the order of digital democracy, then integrated technocracy.
  • Well-being Index is the very definition of happiness.

I learned that surveys and open data are available for each region. Regional Happiness Indicators

I like that it’s multidimensional x multiregional, rather than being reduced to a one-dimensional scale.

  • Which axis is more important for each person varies, and even for the same person, it depends on his/her life stage.

    • When children are born, they prefer municipalities where childcare is easier.
  • People with different preferences can move to their preferred municipality if they are visualized in multiple dimensions x multiple regions. - mobile voting

  • This is, in essence, a model in which humans have a vector of preferences v and regions have a vector of happiness u, and the inner product of these vectors is an individual happiness scalar

    • Given a vector of personal preferences, AI can answer the most appropriate region for that individual.
  • The change of mindset here is rather interesting.

    • If you try to design a function that returns a scalar value of “people’s happiness” as the objective function, the question becomes what do you do?
    • Consider first a “personal happiness” function that takes personal data as a parameter and returns a scalar, pending how to make it a single scalar
      • How to aggregate multiple people’s “personal happiness” scalars
        • Is the sum/average of happiness better? Is the median better?
        • Minimizing the number of unhappy people is good?
          • Would you prefer area under the curve?
          • That’s consistent with “sum of happiness” if the horizontal axis is people.
          • So the question becomes, can we treat all people equally as “1”?
            • Is it okay to be 1 for a birth and 0 for a fetus born tomorrow?
      • Such a “mechanism for combining many people’s scalars into a single scalar” is not necessary to consider at first, since the current society (whether it is optimal or not) has already established collective decision-making through voting and the market.
    • If we separate them, the optimization in the first half will be matching, recommendation, etc. - Understand complex things as they are complex.
  • In this case, “where you live” is a choice between multiple options, but in general, you can participate in multiple online communities, and social interactions in such places also contribute to your happiness.

    • Then just which of the N online communities to join becomes a choice, making it difficult for a human to make a decision!
    • A future where AI “recommends” it to you, thereby reducing loneliness and increasing happiness.
  • Polis-like visualization of the 2022 House of Councillors election

  • Interesting story about blu3mo’s question of what is the second principal component axis.

    • News from all sides
    • America may be a simpler conflict.
      • If so, the contribution of the first principal component should be larger
      • In other words, whether the distribution of arguments is plurality/multidimensional/smooth or unidimensional/stepwise should be quantifiable by principal component analysis
        • The idea here reminds me of an exchange I had in visualization of bio-data when I was a PhD student: “No, I’m observing this in 2 dimensions with PCA, but the contribution ratio is not much different in 8 dimensions, and looking at it in 2 dimensions is throwing away 3/4 information.nishio.icon

The Way of Happiness

  • East Asia emphasizes [connection
  • Different cultures have different forms of happiness.
  • Rather, it depends on the age.
  • It’s going to change in the process of college students job hunting and finding jobs.
    • Life events may change the shape of happiness.
    • It’s true, childbirth changes the priorities of things so violently.
  • You believe too much in hypotheses when you only have individual N=1 experiences (+ a few familiar people’s experiences).
    • Let’s collect a lot of them and meta-science them!

prediction market

  • Futarchy - Wikipedia
  • Polymarket
    • Presidential elections are a lot of fun.
      • GPT.iconThe Iowa Electronic Market (IEM), established in 1988 by the University of Iowa, was a pioneer in these forecasting markets, especially during the 2000 U.S. presidential election.
  • A forecasting market needs a diversity of people using it to make good forecasts.
  • economic incentive
    • Predictive markets are a mechanism for using the market for collective knowledge.
    • What is suitable and what is not
    • Is this good policy?”
      • no settlement is reached
    • Often used in presidential elections because there is a clear result when the deadline arrives.
      • So the bet by its outcome is valid.
      • No one wants to make a bet with an unknown fixed date.
  • Sales Forecast Report
    • Reporting in person creates a bias to be high spirited.
    • Forecast Market at IBM
  • A story about assassination plans leaking out of the market.
    • Incentive design that is anonymous and benefits from the correct sharing of information that everyone else does not know and that you know.
  • Timing of results
    • Forecast Market Horizon
    • Kane’s Beauty Contest
      • It is impossible to verify whether plan A or B has a higher probability of success
        • You can’t do both, so you need to make a decision to choose which one you’re going to do.
      • If the person with the most votes is chosen and rewards are given to those who had chosen it, then it becomes a popularity contest.
    • Decision making may not be a predictive market.
      • Beneficial decision making changes society.
        • A story related to the inability to separate subject and object, or the fact that an observation affects the subject when it was intended to be an unaffected observer.
      • If market-like mechanisms are used for decision-making systems.
        • People who want to make decisions throw decision targets and offer rewards for them?
          • Currency is a scalar, so the band width seems too narrow.
          • As for the deliberative system, I’d prefer to send it in natural language.
          • It IDEA CLOUD is a
        • Decision makers often don’t want to disclose what they want to make decisions about.
          • Well, we’ll just have to make do with what we can disclose.
      • The implicit assumption that people predict
        • What if an AI (or rather, a program) makes a prediction?
        • So that’s algorithmic trading at HFT.

Code for Japan Summit


This page is auto-translated from /nishio/æ€è€ƒăźç”çŻ€ç‚č2024-09-14 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.