“AI is likely to be either the best or worst thing to happen to humanity.” Stephen Hawking

As we kick off the unstoppable train toward automation: do we think an AI will always align with the interests of a DAO?

There is an apparent synergy between Artificial Intelligence (AI) and decentralized autonomous organizations (DAOs). Much like the middle word in the latter suggests, DAOs have technology at the center of their operations, and many are likely beginning to set their eyes on AI.

That makes sense. Groups relying on smart contracts to help make collective decisions, a form of AI one might argue, are the logical experimenters for new types of technology. But, although we are beginning to see encouraging use cases pop up, is AI everything it is touted to be?

Central to this essay is the answer as to whether AI (a technology that has been presented as a force for good) is a boon or bane for DAOs. My goal, as you will soon find out, was to highlight the potential dangers AI poses to these organizations, from rogue AIs that could act against the best interests of the group to certain types of workers needing to urgently upskill to meet the challenge. We will examine the various angles through which AI may impact DAOs, emphasizing both the threats as well as the opportunities they present.

Admittedly, this post began as an all-encompassing exploration of the dangers of AI to DAOs. It aimed to dive into how this technology could upend and finally overtake the humanity of these organizations, causing existential threats and leaving the door open for their impending doom. I sought, merely out of intellectual curiosity, to sow doubt and encourage debate as to where AI could help, and how it could cause harm.

I ran into a wall, however. The case for AI in many DAOs is strong, much stronger than I expected. Sure, there are dangers, and we will touch upon them in due time. Still, interestingly, as conversations with enthusiasts and experts ensued, my mind wandered from unwavering skepticism to open-minded acceptance that there is more than simply “we will become useless.”

For the sake of reading, this article won’t expand on what AI is, or what a DAO is. Surely, if you are reading this, you’ve run across the former, and have explored the latter to some degree of depth.

Could a DAO like Content Guild be under threat?

The purpose of Content Guild is to create content, as the name very much suggests. One could argue, in that case, that it is a media outlet of sorts. So, as conversations ebb and flow about expanding beyond the written form, with video or audio production looming on the horizon, CG could see itself in the throes of an AI-driven interruption. Many of you ChatGPT buffs probably know that the tool is making some of us wordcels sweat, as will other sites geared for alternative forms of content.

Personally, I’m seeing this happen today. The editorial for which I worked is beginning to expand into AI to create their posts, leaving journalists on the sidelines as the all-powerful algorithm takes over. And this is occurring across the industry, with some of the largest research centers in crypto also walking this path. “Powered by MessariAI” is one that immediately comes to mind. News stories sourced by AI, offering readers a quick blurb of what’s going on, leaving those truly interested to expand via alternative media outlets. But as less readers read, and more people scan, the future seems (hopefully not!) evermore artificial.

I find these moves odd but understandable. Although I somewhat refuse to believe in a world where readers look at the byline of an AI-written post and get giddy to read what it has to say–even if it has been reviewed and edited by a person–it’s coming to fruition and things need to be done. Journalism, and original reporting (humans speaking to humans for insights and deeper dives on topics) will prevail, especially in a world of deep fakes, and artificially created stories. At least I hope so.

Writers that survive the AI onslaught will be those that hone their styles (and interests), make their writing a little more pristine, and eventually challenge themselves to write in a manner that informs and engages in a way that an algorithm cannot.

The first battles have been won, so it seems. But let’s keep in mind that knowledge workers will be the first required to upskill, or die. It’s really that simple. Nevertheless, and I can’t stress this enough, this shouldn’t become a cause for stress, but rather a moment to buckle up, and learn to improve.

Are DAOs AI-resistant?

Now might be a good moment to stop and say, yes, there is an existential threat inherent with AI. Like the quote at the beginning, Steven Hawking and people much smarter than me have been sounding the alarms on the escape velocity of the technology, and what it could do to humanity along the way. However, that doesn’t mean it will happen, nor that it will “attack” in one specific way. If anything, it will override our systems in myriad and unexpected ways, forcing us to remain agile for whatever gets thrown in our direction.

Planning for the “singularity” (the moment when artificial intelligence surpasses human intelligence, and human control) can be hard, but I reckon that DAOs might be uniquely prepared for what’s to come. Their decentralized nature means that there isn’t a single central point of attack–despite a certain hierarchy that does crop up in some organizations–and their automated decision-making allows for a kind of imperviousness to reign.

How so? DAOs are more resilient to centralized manipulation, have a layered decision-making structure, and without single-point gatekeepers, make it difficult for an AI to overrun their operations. A second feature is that open and public blockchains are transparent by nature, providing valuable defense against data manipulation or concealment of data by nefarious implementations of AI.

While the future, especially one with an emerging technology like AI is uncertain, there are plentiful ways to implement safeguards against the inexorable command of an AI–led future. As I wrote above, the fact there are multiple layers in a DAO for decision making makes it inherently well positioned against an AI that could act out against the group.

On the other hand, I think first and foremost would be to include members of DAOs in the implementation process. Educating members on how an AI would work, what it’s limitations and reach is, and having those affected involved in what gets developed and inserted would go a long way. This would help quell any FUD that might arise, present alternative perspectives to what’s being added, and harness the power of groups to collectively understand what is being introduced, how it can be used for good, and what eventually might go wrong. This could involve Q&A with technical folks, easy-to-understand blog posts or video descriptions, or even interactive games to “test” how these technologies will affect the everyday activities of DAOs.

Later, there could be filters for anything that involves an AI, whether those are checkpoints, fact checking by a dedicated task force (some might argue that this invalidates the automation bit of an AI, but it doesn’t need to be anything exhaustive), have a voting process, or a smart contract that verifies what an AI is doing. I think the inherent transparency of a blockchain also goes a long way, and allows for us humans to keep tabs on the AI’s doings.

A financial use-case

Plenty has been written about how DAOs might benefit from implementing AI. Today, I’d like to begin with one particular use case–which lies at the intersection of the crypto narrative and coordination.

Funding is one of the main concerns that DAOs face. Hell, it’s one of the issues that every layer of society has to deal with. Given that the majority of capital that DAOs are using today–Content Guild being a prime example–is the US dollar, how does an inflationary monetary policy affect its functioning and its members?

For us living in the Global South, a controversial concept that refers to developing countries, having access to dollars is a godsent. Whether that’s physical benjamins or stablecoins, they help us survive in the face of endless monetary debasement and corrupt leaders printing our livelihoods to hell. That said, I’m aware, as the majority of crypto-native people also know, that the dollar is subject to discretionary policies that ultimately inflate the money supply and devour our purchasing power.

How about an AI to quell the effects? What if the AI were to make investments on behalf of the DAO? In split seconds it could plug into the internet, and not only find opportunities but also find threats. The threat of conflict in the Middle East could drive up the price of gold, so that an AI could immediately and instantly buy gold futures on the London Stock Exchange, leaving the organization’s sometimes depleted treasuries a bit more guarded.

I know what many of you are thinking: this is endorsing speculation and could leave us ultimately without funds. Both are true, and both are risks. But where lies risk lies reward, and it’s not necessary to allocate all the funds to an AI investment branch. Try it out with a small percentage and see how it goes. A hedge against the impending bust of the greenback or other fiat currencies currently reign supreme.

Some might roll their eyes at the difficulty–if not impossibility–of training an AI for this purpose. But the above funding concerns, and inherent issues with the dollar, are real, and it could be time to think outside the box as to how to protect our treasuries, and even increase our monetary resources.

Help DAOs do things

We are seeing, all across the internet, people begin to leverage AI to write code for them. This could also become an area of interest for DAOs, which would unburden a lot of the administrative work, along with reducing development costs, freeing up time and capital resources to allocate for other purposes.

Allowing AI to write code for DAOs could eventually harness the power of human members to create, experiment and innovate even more–now a little less unrestricted by the typical constraints that come from having to complete sometimes menial tasks that are waiting for an algorithm to automatically get cracking at. This could, going a bit further, be the inflection point for things to happen. Sometimes, our organizations are troubled by the aforementioned limitations–there are only 24 hours in a day, and only so much in a treasury–which could greatly benefit from the automatization of certain processes.

Admittedly, this might not be something new, but it’s a use case for DAOs to find a way to foster their middle initial.

And my last point, perhaps a bit controversial, is that this could give certain DAOs an edge over others, simply because of their ability to try new things or just to get things done. Let’s be honest: even though it might be crude, funding will usually go to projects, platforms, and people that ship products, artwork or services. Sometimes this isn’t the case, but DAOs that can deliver will have an edge. Offloading the things holding DAOs back from following through in greater numbers might be an effort that organizations should be making.

While I don’t consider myself sold on AI, I’ve made some strides toward approaching the technology and its apparent benefits. Although there is a fair bit of fetishizing AI, especially as the hype continues to mount, the use cases are far and beyond, and the threats slim and between–for now. Admittedly, I have changed my mind time and time again as I wrote these words, but I find myself holding two different but seemingly coherent thoughts.

Technology such as AI holds some auspicious rewards, and these could compound for those willing to take the risks. But, there are still threats looming on the horizon, some of which we still haven’t even discovered, which begs me to end on a final question that I still don’t have the answer to.

I sincerely appreciate the provocative food for thought offered by Francisco Díaz, Trent McGonaghy and Kenneth Francis for this post.