The big idea and why it matters: AI agents coupled with the blockchain show promise of performing some of our usual human tasks, including negotiations and payments; this may fundamentally shift some of our current routines and increase our effectiveness, if done correctly. However, there are also concerns that the blockchain could be undermined by even more advanced technology.
“Sometimes the best hiding place is the one that’s in plain sight.” -Stephenie Meyer (novelist/producer)
While Part 1 of this series mainly covered some crypto updates, I also want to ensure we get some perspective on the broader blockchain today…though it may be fair to note that crypto continues to do its best rendition of the great Tom Petty’s Free Fallin’, as Bitcoin and Ether are now respectively more than 34% and 44% off their highs (source: FactSet, 11/21/25). Here we go!
Secret Agent Man
As much as we have concerns about the stock prices of companies at the forefront of the AI revolution being expensive, the potential, eventual use cases of Artificial Intelligence seem undeniably cool. Thus far, however, I haven’t found very much utility with this technology. Yes, I can make fun pictures for entertaining group chat fodder. Still, it turns out that tying together various systems and human interactions to develop thoughtful solutions is complicated, and I believe that’s what I’m really waiting for AI to do.
It sounds like the solution to my dilemma will ultimately include AI agents, which are (per Google Cloud) software systems using “AI to pursue goals and complete tasks on behalf of users. They show reasoning, planning, and memory and have a level of autonomy to make decisions, learn, and adapt.” Agents can also work with other agents to accomplish more complex workflows. While Bots automate simple tasks and AI Assistants respond reactively to our prompts for basic tasks, AI Agents can be proactive in executing more complex actions. It represents a whole other level of capability, as they can go out into the world and do real things.
We’ve all probably interacted with AI agents already, even if we haven’t deployed our own, as there are many use cases, some of which include (per the link) trading (stocks/crypto), dynamic pricing (Uber), Netflix recommendations, Roomba vacuums, Nest thermostats, and virtual assistants.
I know what you’re thinking: the above examples are built on centralized servers, so what does this really have to do with our current topic, which is supposed to be about the blockchain? I’m glad you asked.
Taking it to the Blockchain
When we combine the concept of AI agents with distributed ledgers (aka blockchain), the use cases expand to having personal assistants out in the digital world, acting on our behalf 24/7. Why does that matter? As 101 Blockchains conveys, the blockchain avoids issues of potential manipulation or other influence from centralized platforms, adds trust, auditability, and transparency, and allows AI agents to autonomously “coordinate between various blockchain platforms.”
What are the practical implications? Your AI agent can understand instructions, gather data, analyze the data, make decisions, and then execute on your behalf. The above article uses an example of crypto trading (e.g., take a user’s instructions, gather data across platforms – like prices, news, and social media – analyze what that means for the currency, and make/execute decisions, like placing trades), but I don’t regard that as having mass appeal.
I find it more exciting to think about food and how to use an AI agent for grocery shopping and improvement of diet: I could provide a list of meals or recipes we like to eat, and an agent could go shopping for me based on sales at our local stores and then arrange for pickup or delivery of the groceries each week in our preferred windows of availability. The meal plan could naturally vary based on pricing/availability, adjust for seasonal eating preferences and holidays, and interact with automated inventory control in our fridges and pantries. And would my agent check our schedule to ensure we don’t have a conflict before making “pizza Friday” happen automatically, instead of us going through our routine where I ask my wife if she’d like to order it (we already both know the answer is “no” 100% of the time) before I inevitably place the call for our dinner delivery? Food for thought, but it’s probably only a matter of time before these things are commonplace.
A missing link to becoming an AI Boss
This recent article from Medium poses a challenge that AI agents are naturally going to encounter: obtaining the necessary data, using premium AI models for analysis, and relying on cloud storage and other services to execute instructions on behalf of us human “AI bosses” will require payments. However, as we’re all aware, payment interfaces require credit card data, CAPTCHA verification to prove we’re human, dual-factor authentication, etc. Enter the blockchain and HTTP 402.
It turns out that HTTP 402 was reserved for future use by the Internet Engineering Task Force (IETF) back in 1997. You really have to give them credit for the foresight. The above article gets well into the weeds, including actual code examples, if you’re interested; however, the short story is that HTTP 402 allows for payment requests and coordination of payment (though the payment itself would be made elsewhere and then verified via blockchain), solving for the missing human in the transaction.
They go on to posit some interesting use cases, like micropayments for articles, music, or gaming, dynamic pricing, and API monetization, but one that caught my attention is Agent-to-Agent commerce. Here’s why. As of now, I can list items I’d like to sell on a platform like OfferUp. Other people can search and view my wares, and then we can interact via chat, negotiate a price, and find a time to meet to make the exchange. But what if AI agents can handle all that for us? My agent goes out to sell and negotiate on my behalf. Your agent is there to identify items, negotiate, and buy on your behalf. As with the grocery example, our agents could then coordinate payment and the ultimate pickup/delivery, either by reviewing our calendars or by hiring a service to handle it. Basically, it’s robots transacting with one another to drive the whole process. And given that a colleague and I recently saved a food delivery robot (pulling it out of a ditch in Dallas), even the delivery part may avoid humans entirely.
Honorable mention: There are already concerns over AI agents and bots influencing voter perspectives, but should we expect that blockchain will be key to the future of election integrity, and that AI agents may eventually vote on our behalf?
While it’s fun to hope for upside scenarios that can remove some friction from our day-to-day lives, it’s also vital to consider what happens if things don’t go as planned. For example, what if the blockchain itself were compromised in a way that undermines the entire premise on which it’s built? That shouldn’t be a threat with current computing capabilities, but it has been voiced as a concern if quantum computing becomes a thing. This Forbes article (from May 2025) provides an overview of quantum computing and its game-changing use cases, along with the expectation of a lengthy timeline for development and adoption, due to significant challenges with the technology.
However, just recently, Google announced that its Willow quantum chip “notched two major achievements right off the bat,” suggesting that “quantum” may be ahead of schedule. The first was addressing error reduction (a known challenge of quantum computing noted in the above article), and the second was performing a calculation in less than 5 minutes that “would have taken the fastest supercomputers available today 10 septillion years to handle.” NBD 😊. Nobody knows where this all goes, but there will be a lot of change coming, and it’s part of what keeps life exciting, so I’ll simply add it to the long list of things to be grateful for this holiday season.
Until next time, this is the end of alt.Blend.
Thanks for reading,
Steve
