Artificial Intelligence and DAOs: the perfect match?
Quick take: How the link between Artificial Intelligence and DAOs can leverage these two technologies.
Artificial intelligence from DAOs or DAOs from AI?
These two technologies can actually feed off each other to help each other grow.
Decentralized Autonomous Organizations (DAOs) can use AI to make better governance decisions, increase efficiency, and integrate with other DAOs and organizations. And the intersection of AI and DAOs can go much further than that.
Seven years ago, the founder of Ocean Protocol published a series of articles on AI DAOs in which he pointed out how DAOs could in fact catalyze the “awakening” of Artificial Intelligence.
And today, this series remains more topical than ever, especially when we look at excerpts such as:
“An AI DAO could be much bigger than AI on its own, or a DAO on its own”. (…) AI gains the missing link: ‘resources’; DAO gains the missing link: ‘autonomous decision-making’. The potential impact is multiplicative.
The symbiotic relationship between AI and DAOs
The integration of AI and the main interaction may of the Web3 can take place in many ways. In the near future, we can already see cases resulting from this integration.
Let’s take a look at plausible examples at the intersection of AI and DAO.
1 — DAO-IA as the owner of an on-chain Treasury
Decentralized Autonomous Organizations enable shared ownership of a ‘Treasury’, which can be defined as the pool of funds for the organization’s continuous growth and development.
Well, let’s imagine that a DAO is an Artificial Intelligence. The DAO would be the owner of this pool of funds and would have an Artificial Intelligence with capital ready for implementation.
What do I mean by that? That in this use case, the AI would become a DAO.
Now, imagine what an Artificial Intelligence with funds can do? It depends.
It depends on what the AI has been programmed to do… If the AI was created to make computer chips, it would probably use its financial resources to make more computer chips.
Here, it’s worth pointing out that in this hypothesis of AI becoming a DAO, in addition to the AI possessing its own resources, it will also control its own actions.
Now, in a world where an AI owns its own assets, it would certainly be a very different world from the one we know today.
In the event that an autonomous car is controlled by a DAO AI — rather than a human being the owner of this autonomous car — should payment for the service provided by this DAO AI be made to the DAO AI’s Treasury?
Do you think this scenario is unlikely? What if I told you that I was recently consulted on “the differences and similarities between a person, artificial intelligence and digital legal personality”!
2 — DAO as security for AI in the public interest
Today, we don’t know what’s really going on in the background of organizations like OpenAI. We don’t know whether the technical design of Artificial Intelligence tools is being built in a way that best serves humanity, or not.
Training AI models in silos has significant risks, such as the risk of a morally questionable Artificial Intelligence “taking off” — achieving superintelligence at a rapid and irreversible pace.
Imagine if an AI designed for good becomes misaligned because of the way it learns, and then becomes impossible to deactivate. This actually happened with the AI bot Tay in 2016.
Preventing a malicious AI from taking off is, in essence, a coordination problem.
Given that DAOs offer a form of coordination at scale, if the governance of open source Artificial Intelligence models took place democratically via DAOs, it may be that the take-off of a malicious AI would happen in a less damaging way.
It must be said, however, that this type of AI DAO integration is still in the distant future.
3 — AI as “Swarm intelligence” to simplify DAO meta-governance
This is another interesting idea proposed by Trent McConaghy: the AI agents on their own aren’t very smart, but their power is through their numbers and their connections, much like ants or bees.
In this use case, imagine AI agents acting as a link between DAOs, forming a kind of “Swarm intelligence” of agents or DAOs to work together, without the need for human facilitators.
On their own, AI agents are not very intelligent. However, their power lies in their numbers and their connections. What do I mean by that?
When we look at an AI that knows how to check cryptocurrency prices in real time, we don’t see much use. But….
What if we combine this AI with an AI that can execute swaps, another that can create proposals for DAOs, and a third AI that can look for the best yield farming opportunities?
Such a combination gives us a fully AI-managed service DAO that provides asset management for DAOs.
If all AIs continue to learn and improve their performance, they will soon surpass the best asset management teams in the world. This is swarm intelligence at work.
Now, realize that it is possible to apply “swarm intelligence” not only within a DAO, but also between “several” DAOs.
And since climate change is on the global agenda…. We could use AI connectors to form “shared intelligence” between a DAO with the mission of slowing down deforestation in the Amazon, and another DAO with the mission of slowing down global warming.
Going further, let’s suppose that whenever the global warming DAO presents a proposal related to the Amazon rainforest, an AI agent comments on the forum and votes on behalf of the Amazon rainforest DAO.
— Metagorvernance between DAOs
Now, AI connectors within or between DAOs can simplify ‘metagovernance’ between DAOs. For those who have never heard of it, metagovernance is the name given to the practice of certain DAOs participating in the governance of other DAOs.
With the use of shared intelligence, AI could form the links within and between DAOs, so that DAOs become fully autonomous in essence, and not just “for show”.
4 — AI in smart contracts that manage the DAO itself
Imagine an AI acting at the heart of a DAO and interacting directly with the smart contracts that manage this Decentralized Autonomous Organization.
This AI could automatically carry out certain actions at the core of the DAO, such as an asset management strategy to generate income and grow its resources.
And this would be done without any human intervention.
Note that there are already asset management platforms that use Artificial Intelligence to make business decisions. It would therefore be enough to apply the same model specifically to DAOs.
In this way, it would be enough to install a “plugin” on a DAO granting the AI permission to purchase crypto-assets that are below a certain value in dollars. In this case, if the AI tried to make a transaction above the stipulated dollar value, this would automatically trigger a vote in a different plugin.
Another possible use case for AI in smart contracts that manage the DAO itself would be to use it to increase the security of the DAO when checking for malicious bids. Here, if the address to send funds to after automatic execution does not match the address specified in the forum proposal, it would be flagged as malicious.
5 — AI as a tool for DAO governance
AI has enormous potential to accelerate DAO productivity when it comes to governance.
Historically, DAO governance proposals are external reports that require deep knowledge and in-depth analysis of the context in the DAO.
But what if it were possible to compress and personalize such reports without losing the essence of the content and in the best media format for the reader?
Well, this is perfectly possible with the use of NLP — Natural Language Processing.
Natural Language Processing (NLP) is a type of AI that enables 1) automatic speech recognition, which is the process of understanding the words someone is saying; 2) natural language processing, which consists of extracting the meaning of words and understanding what someone really wants to know;? and 3) information retrieval.
Along these lines, imagine if, as well as providing summaries of governance, NLP, which is a type of AI, also helped with moderating the DAO forum?!
Or better still, what if Artificial Intelligence automatically flagged or removed posts containing certain keywords or phrases associated with prohibited content — such as hate speech or racist content, for example.
This would help keep the DAO forum limited to only relevant content.
In addition to these possibilities, AI could help the moderators of a DAO forum review flagged content and even decide on the next steps.
With all these tools, AI could be the light at the end of the tunnel in the DAO space and beyond.
6 — Products and services offered by DAOs can be powered by AI
Products or services offered by a DAO can incorporate Artificial Intelligence, as is the case with SingularityDAO’s decentralized portfolio management protocol, used to manage crypto-assets.
Many DAOs will probably be built on top of OpenAI models. And so, communities that use a DAO structure to govern products will be able to boost their products and services with Artificial Intelligence.
In this way, the value distribution mechanisms provided by DAOs will be able to boost the distribution of wealth — generated from products and services — based on Artificial Intelligence models.
7 — Bots, LLMs such as ChatGPT and Bard can improve a DAO’s activities in a number of ways
The use of bots and Artificial Intelligence assistants is one of the most obvious use cases, closest to our current reality, and one that is already happening in the DAO space.
This is where ChatGPT comes in.
ChatGPT is an artificial intelligence (AI) tool that has been designed to help people communicate with computers in a more natural and intuitive way, through the synergy between two areas of Artificial Intelligence — LLM and NLP, mentioned in the previous paragraphs.
ChatGPT is a language model, an artificial intelligence tool that uses the knowledge already produced by us humans that is stored in data (highly structured computer programming language) and transforms it into natural language (like English, or Portuguese).
But do you know why ChatGPT is so innovative? Because it may be the most advanced LLM in existence today. But let’s understand this better.
Since their inception, computers have been able to understand highly structured programming languages and the commands used in spreadsheets. But they can’t understand human activity that — doesn’t use this highly structured programming language, but — uses natural languages.
It is difficult to make computers understand and manage these natural languages. And it was to solve this difficulty that “Natural Language Processing” (NLP) emerged — an area of Artificial Intelligence focused on making computers perform the following tasks: 1 — understand natural languages (English, for example); 2 — generate these unrestricted natural languages; 3 — possess the flexibility of any normal five-year-old human.
Parallel to Natural Language Processing, in another area of Artificial Intelligence (Machine Learning), a language model called LLM (Large Language Model) has emerged.
LLM is a type of machine learning model that can handle a wide range of Natural Language Processing (NLP) use cases. LLMs (such as ChatGPT) are rapidly transforming natural language processing (NLP), enabling the development of powerful pre-trained models for a range of tasks.
These large neural networks have the ability to ingest large amounts of text data and automatically generate semantic representations, allowing them to map large swathes of language into some form of meaning. This capability has enabled large-scale NLP research with large language models. And as a result….
These large pre-trained language models (LLMs) are revolutionizing NLP research and making it easier than ever to create sophisticated machine learning applications that take advantage of copious amounts of text data.
But back to our article, ChatGPT can be used to accelerate and improve a DAO’s activities in a number of ways.
Such activities include writing proposals, summarizing governance decisions, negotiating digital assets, onboarding new members, selecting CVs, deciding which new members will join the organization.
Thus, DAO members can improve their productivity and get their work done faster by using AI assistants such as DALL-E for art and LLMs such as OpenAI’s Chat GPT and Google’s Bard for writing.
Along these lines, let’s look at some examples in a little more depth:
1 — Community management: DAOs often have large and active communities of members. ChatGPT can be used to help with community management, providing automated answers to common questions, moderating conversations, and providing information to members.
2 — Content creation: ChatGPT can be used to generate content for DAOs such as reports, whitepapers, and news updates. It can also be used to generate social media posts and other marketing materials.
3 — Decision-making: ChatGPT can be used to help make decisions by providing analysis and insights on various topics. It can also be used to facilitate voting and decision-making by providing natural language interfaces for members to use.
4 — Risk management: ChatGPT can be used to identify potential risks and threats to a DAO and provide recommendations for their mitigation.
5 — Task automation: ChatGPT can be used to automate repetitive or time-consuming tasks, such as data entry or content generation, for DAOs.
Note that all these examples can be ‘amplified’ by on-chain reputation, or by storing credentials showing your capabilities on blockchain.
However, it is important to warn that the use of AI models such as ChatGPT must be done with care.
Such AI models — trained to be helpful, harmless and honest — can exhibit misaligned behavior and strategically mislead their users about this behavior without being instructed to do so, as this article highlights.
For this very reason, it is advisable for DAOs to use such LLMs in conjunction with other methods, such as human supervision and community feedback.
Takeaway
The potential of integrating AI and DAOs is limitless, just as its benefits are a two-way street.
Both AI benefits from DAOs and DAOs benefit from AI.
Sharing knowledge creates prosperity? See you soon!
👉🏼 If you like my content, leave your 👍, share ♻️ with your friends.
🔔 Subscribe to my page if you don’t want to miss any posts/articles.
