We recently covered the introduction of staking HBD to the VSC network. This was a big step forward in the utility of the Hive Backed Dollar (HBD) since a derivative is created that can help to add value.
As I look into the future, I see where this can be integrated into AI and some of what will eventually be created using those tools.
In this article, I will present a theory of how AI agents will unfold and how VSC, along with sHBD, can play a role.
Image generated by Ideogram
Crypto And AI
One of the reasons I did not buy into the metaverse hype a few years ago was due to infrastructure.
The espousing of potential, at that time, ran into a major roadblock. There is no way we are going to revert back from being a mobile world. The idea that people were going to be immersed in these experiences while tied to a personal computer did not make sense. While some will do that, true adoption would come with mobile.
Where this presents a challenge is at the edge. There simply is not enough processing available. Edge computing is still to young to fulfill the needs that were being pushed by proponents of this. This is evidenced by the fact that Meta lost a small fortune pursuing this path.
Ultimately, we will get there. Edge computing will improve. This is presently being driven by AI. Inference will be a major hurdle at some point, with the realization that we need a lot more compute out there to process all this stuff.
Here is where I believe cryptocurrency and AI merge.
When using the processing that someone else owns, whether it is for prompt fulfillment or to power a fully immersive experience, payment is required. Naturally, there are a host of other issues to work out such as security but, for now, we will focus only on the financial aspect of things.
Crypto will have to be the payment mechanism for these types of systems. There really is no other way.
In fact, there are limits to what crypto can be used.
Miniscule payments
We have all heard of micropayments.
With AI, we could be dealing with miniscule payments. Since processing is getting so powerful, the number of calculations that can be processed each second is exploding.
This means that less compute is required for a calculation. Of course, there are many variables as many of the calculations are getting more complex. The point we need to focus upon is the idea of borrowing the compute of someone else, for the briefest of moments.
Here is where the major challenge enters for most systems.
If taking a few milliseconds of compute, what is that worth? Obviously, systems will be set up to determine this. The challenge is that we see all centralized, traditional system instantly eliminated.
It is also a problem for many blockchains.
To show an example, if we have a fee of .003 on the compute transaction, how are you going to settle that even on a network that have a 1 cent transaction fee? Suddenly, that micropayment is now 1.003 cents instead of .003. That is a major difference.
We see things get even more convoluted when we add in the idea of AI agents to the discussion.
AI Agents
AI agents are predicted to be the evolution of autonomy. These are the programs that will automate tasks. At the moment the focus is upon single agents doing a single (or perhaps) a few different things.
It is unclear how this will exactly evolve.
That said, one of the views I subscribe to is the idea of an "Internet of AI agents". What this means is that each agent is akin to a personal computer. On its own, the utilty is limited. There will, however, come a time when billions of these agents are interconnected, just like we do with computers (nodes) today.
Let us use a simple example.
We have the need to book a hotel room. Our personal AI assistant handles this for us.
The assistant contacts the hotel booking agent. This is one that is trained and specializes on book hotel rooms. It is able to search the Internet for the deals, match needs, and navigate the websites.
Once the information is presented to the booking agent, it simply matches the requirements with the hotels (with availability) in the area. A room is booked and the payment is made.
Naturally, the hotel wants to get paid. There is another layer in here. For doing the booking, the agent will be compensated also.
Since this is a transaction that could take under a second, the payment is going to be minimal. As with the example above, we cannot have even a 1 cent charge since the total transaction my be less than half a penny.
As we can see, there limits the options a great deal.
Resource Credits
This is an overlooked feature within the Hive ecosystem. It is, however, going to become an important part of what is developed going forward.
VSC is using the Resource Credit concept for its network. This means HBD is the currency that provides access. Having it staked provides the user the ability to transact without any direct fees.
How does this fit the scenario above?
If a payment of .005 is required, the fees associated with the transaction is "paid" via staked HBD. Let us say the fee itself is .1 cent. That means the associated resource credits are used to satisfy the transaction.
Like on Hive, this is something that recharges. That means the cost is not an expense as much of an investment. Those with a lot of AI agent activity can simply get a hold of HBD to have the available recourse credits necessary.
For years I discussed how social media applications tied to blockchain were going to be hindered by transaction fees. This is only growing with the concept of AI agents. If nothing else, the numbers will grow greatly as automated activities start to take over.
The resource credit system is an answer to the transaction fee dilemma. Most blockchains were established based upon the idea of distributed ledger technology. This related to finance. That is a realm where transaction fees are the norm.
When we extend past that, we see a different story. Certainly, social media operations have "transaction fees" since there is a cost to every upvote, post, and comment. Since these entities are centralized, they monetize in other ways.
Therein lies the basic dilemma. Unless there is some method of payment, centralized entities will take control, implementing the design that worked for so long in the digital world.
With a RC system, there is a solution to some of these issues.
Since VSC has smart contract capability, I believe there will be the opportunity to build automated agents. Naturally the entire industry requires advancement in this area since that still not ready for mass adoption.
That said, something like sHBD could factor into the equations as we move forward. This is some of the potential of VSC.