Introducing the DAO DAO Telescope Experiment

We will soon be using Telescope to examine DAO2DAO dynamics among DAO DAO DAOs. For those unfamiliar with Telescope, it is an ethical Discord bot assisting our ethnographic research by automating the collection of qualitative data. It not only connects researchers with participants but also aids communities in preserving and gaining a deeper understanding of their own history.

The way this bot works is simple, the telescope emoji 🔭 is used by anyone in a participating Discord server to highlight relevant messages/discussions. The use of the 🔭 on a message triggers an automated process: message authors receive a consent request and consented collection is saved in our dataset and added to a cross-server bridge.

Here is a link introducing the Telescope bot and this is an academic paper featuring the Telescope titled, Towards a participatory digital ethnography of blockchain governance.

Our goal: surfacing DAO2DAO dynamics.

Our work with DAO DAO aims to compile a dataset of Discord posts that can be used in qualitative analysis. In particular, we are interested in surfacing DAO2DAO dynamics. Nick Merrill (Elsehow) has articulated DAO DAO’s role as a mesh governance protocol here. We aim to gather empirical evidence of mesh governance emerging throughout out the DAO DAO network using ethnographic techniques. We want to find instances of token swaps, joint ventures, co-liquidity pools, etc.

The bot will be live in several DAOs Discord servers. Members of those DAOs would use the Telescope to communicate a comment back to DAO DAO. They would be guided to use the Telescope emoji on comments/discussions that relate to how they are working with other DAOs in the ecosystem (for instance, token swaps), as well as DAO DAO products/framework attributes that facilitate this. By adding the Telescope emoji to a comment, anyone can ‘flag’ a comment for possible inclusion in the dataset (described further below).

What is ethnography?

An ethnography is social science research that uses particular qualitative research methods. What we call ‘observation’ or ‘participant observation’ in ethnographic research typically means paying attention to the everyday, getting to know a community or group, and looking at how group dynamics form through informal processes, norms, practices etc. An ethnography can challenge assumptions, provide subtle insights into social phenomena, and inform the development of more effective policies, products, and services that might better align across pluralistic and heterogeneous groups.

Some ethnographies set out to capture events and to understand what they mean in relation to the wider system. Most people think of ethnography as anthropologists who study one locally defined community at a time (going ‘to the field’). However, ethnography has had to change to accommodate our technologically mediated lives (‘digital ethnography’). When communities communicate online, and where some of the actors in a system are machines (smart contracts in DAOs, for instance), things start to get complicated.

It’s important to use transparent processes while doing ethnographic research and to have the consent of those who participate in the research, which can be difficult in a digital context. It can also help to have research participation from the communities being studied. That’s why we are developing experimental computer aided research tools. This type of research we’re calling Computer Aided Ethnography (CAE) as it enhances research methods, enables participation, and generates richer insights through automating aspects of data collection, consent, storage and analysis. See this post by Kelsie Nabben, Michael Zargham and Ellie Rennie.

Our Discord bot is one such tool designed to help us capture nuanced descriptions of interactions and processes.

The Telescope

The Telescope helps address several challenges for doing ethnography in the fast paced and transient Discord servers populated with pseudonymous identities and machine agents. In essence the Telescope bot acts as a consent form, a data bridge and a curation tool when collecting qualitative data across busy Discord servers. It is a computational assistant collecting insights into the experiences, influence, and perspectives of a range of agents within a dynamic process.

Figure 1. Control flow for the DAO DAO Discord Telescope bot

We now have a relatively simple tool that automates qualitative data gathering in an ethical and participatory way.

The DAO DAO Telescope Experiment

We’re excited to launch a new experiment using The Telescope with the DAO DAO community!

We are now inviting DAOs in the DAO DAO network to participate in our research. Our plan is to integrate The Telescope into multiple Discord servers, creating a bridge between DAO DAO and the DAOs that choose to participate. Those that do participate will be able to use the Telescope emoji in their Discord servers to flag messages of interest that will be reviewed by a curation team.

We will specifically be asking DAO members to use the Telescope for comments related to DAO2DAO events. This could include things like token swaps, DAO collaborations, and emergent networks occurring across DAO DAO DAOs.

We hope to enrol 10–15 DAO to start with. The experiment will last six months, ending in mid-September 2023. After this period we will assess the success (or failure) of The Telescope Experiment and collectively decide on next steps.

Figure 2: DAO DAO Telescope Experiment

Things to Note

We want to be clear that this is highly experimental, and it might not work! The SourceCred trial demonstrated that for the bot to be effective the following needs to happen:

• DAO members need to know why it’s there. Importantly, they need to accept the bot in their DMs (they will get a message in the server if the bot is trying to message them and their DMs do not allow it).
• DAO members need clear instructions on what to ‘tag’ with the bot.
• It helps for the community to reflect on what’s been tagged (for instance, 5 min review of messages that were tagged in a community call). We will try to advance this process by providing written summaries in the form of Discord updates/newsletters.
• Like all platforms, Discord can make changes that will cause the bot to fail. For instance, it doesn’t work with messages that are in threads.
• Our links break when channels are archived, which means that we are unable to view the context of a discussion if DAOs reorganise their channels.
• Finally, this tool is just that. It cannot do ethnography, it merely assists an ethnographer (whether that be an academic researcher or a DAO member ethnographer).

Overcoming Limitations

Let us know if a channel is archived or moved. DAO members can use the bot to alert us to that, DM us, or @ us in the bridge channel itself.

As mentioned above, posts that have been tagged in threads are not visible in the dataset. We recommend that when tagging posts in a thread, tag the first message in a thread. If the researcher wishes to use a comment further down the thread, they will need to contact the author manually for permission at this stage.

Please also reach out to us if you have things to say and would prefer to share them through a conversation/interview (anonymously or otherwise) rather than on Discord.

Data

An Airtable will be used to store the Telescope data. This table will be accessible by an approved list of researchers and DAO DAO members.

Informed consent for data collection

Communities who wish to participate in the research will receive information and necessary resources before signing on. We are available to talk to the communities if they have questions.

Throughout the project, users can access relevant links pinned in the DAO DAO bridge channel in their respective Discord server. When their posts are tagged with the Telescope emoji, users will receive automated direct messages providing details about the research. Participants have the right to withdraw their consent for individual messages and can request removal of all their content from our dataset at any point during the research.

Analysis

The Telescope is designed for ethnographic research, to be used in conjunction with observation and interviews.

The data collected by these bots can be explored in a multitude of ways depending on the researcher or community’s need and expertise. For instance, the researcher may choose to code the data according to themes to make it explorable in a qualitative research environment such as Obsidian or Roam Research. Or community members could be invited into this environment as a place to explore and define themes relevant to their community and respond to issues arising within the data.

In this project we are building an Obsidian qualitative research environment that will use data collected by the bot alongside interviews and other materials (such as blog posts etc) to undertake ethnography. If a member or members of DAO DAO are interested in being part of that Obsidian environment, we are happy to discuss that possibility once it’s established and operating smoothly.

Another possibility for exploring this kind of qualitative data is using a generative pre-trained transformer (GPT), enabling researchers to interface with the collected qualitative date much like a GPTChat bot trained on a private community dataset made up of the members in the DAO who have consented toward the content being collected. We don’t know if this would work (it would definitely not substitute for ethnography) and we are not suggesting that DAO DAO do this, although we are happy to experiment on the dataset if that’s of interest.

Authors: Matthew Green and Ellie Rennie are a small team of researchers from RMIT University in Melbourne, Australia. Prof. Ellie Rennie leads this research which sits within RMIT’s Digital Ethnography Research Centre and the Blockchain Innovation Hub and is funded by an Australian Research Council Future Fellowship grant.