When APIs and AIs converge, solving for PI (aka π)
The number π is an irrational number, meaning that it cannot be expressed exactly as a ratio of two integers, closely approximated 22/7. It is a transcendental number, meaning that it cannot be a solution of an algebraic equation involving only finite sums, products, powers and integers. This transcendence of π implies that it is impossible to solve the ancient challenge of squaring the circle with a compass and straightedge. As math geeks celebrate 3/14, March 14 each year as the day of PI, we are living
through unprecedented times both geopolitically and technologically. Only very recently, agentic artificial intelligence (AI) alongside large language models (LLMs) are rapidly upending traditional business models and abstracting linear progressions. So, how can we compete and thrive during this period of exponential change?
Before we go any further - how long has AI been around, what is agentic AI and what are LLM? AI can trace its roots back to the 1920’s when a Czech playwright called Karel Čapek wrote a science fiction play named “Rossum’s Universal Robots”, where the concept of artificial people was introduced for robots. However, it took until 1956 when John McCarthy was given the title of Father of Artificial Intelligence. In other words, AI has been around for a lot longer than ChapGPT - which itself really thrust this technology on to the world making AI accessible for all. Meanwhile, Google defines AI as: “Artificial intelligence is a field of science concerned with building computers and machines that can reason, learn and act in such a way that would normally require human intelligence or that involves data whose scale exceeds what humans can analyse.” Agentic AI, or AI agents, refers to AI systems that are able to make independent decisions and exhibit autonomous behaviour. Agentic AI can reason, plan and perform actions, adapting in real-time learning as they go to achieve specific actions. But, unlike traditional automation that follows predetermined programming, agentic AI doesn't rely on a fixed set of instructions. Instead, it uses learned and identified patterns to determine the best strategy to achieve a task. In January 2025, Gartner carried out some research and found that 64% of respondents plan to deploy agentic AI in their business within the next year. Examples of agentic AI would be self-driving cars, virtual assistants and in cyber security where it is deployed to monitor IT systems and detect threats, hacks or nefarious actors. LLMs SAP define as: “A type of machine learning/deep learning model that can perform a variety of natural language processing (NLP) and analysis tasks, including translating, classifying, and generating text; answering questions in a conversational manner; and identifying data patterns.”
Examples of LLMs
Source: Google
I envision a future where these technologies are harnessed in a very user-friendly manner to create programmable, reasoning identity wallets that can automatically make payments in real time without human intervention. Liquidity provisioning and sufficient capital reserves will remain vital, as this is not another stablecoin for trading or investing. Mid-sized corporate treasuries and even governments with complex supply chains will soon manage accounts payables so much more efficiently and accurately using AI and then execute payments in a more intelligent automated fashion 24/7. Since the Satoshi paper, which introduced the world to “Bitcoin: A Peer-to-Peer Electronic Cash System” and then a cornucopia of many other cryptocurrencies and decentralised finance (DeFi), traditional financial institutions (TradFi) keep trying to maintain old systems using legacy fiat money making payments that for cross border transactions still can take two to three working days. Digital programmable payments can happen in almost real time at a fraction of the cost. But what if we shift to programmable payments with embedded intelligence? Real-time value exchange has arrived and is now being deployed alongside agentic AI workflows. Many software companies such as UIPath, MineralTree and n8n are attempting to bring software solutions that integrate and streamline core business processes such as finance, HR, manufacturing, supply chain, sales and procurement systems in sync with workflow automation and AI tooling. It should be of no surprise that large global banks from JP Morgan to Bank of America are experimenting with multiple use cases for their corporate and capital markets divisions. In Brazil, Unibanco is looking at issuing its own stablecoin given all the attention these digital programmable products are receiving. Here in the UK, given the Bank of England’s blessing, creating an intelligent wallet processing payments is entirely doable today. In essence you just need a four-step process with an ‘A team’ of experts:
· design and mapping (architecture)
· proof of concept (real data in a lab)
· minimal viable product (test with 3 midmarket companies)
· launch (UK)
In addition to investor, regulatory and licensing support, an initial project team would require expertise across AI, data, identity and banking. Now is the time, and the UK is an ideal market to build an identity wallet with embedded AI. London is home to a thriving FinTech community of over 1,600 companies and has the added benefit of being the home of English and Welsh law, which is widely used and respected globally. As the Trump administration champions dollar-based stablecoins - backed by powerful fintech figures and geopolitical ambition - Europe finds itself at a critical crossroads. The digital payments war is no longer hypothetical; it’s happening in real time. With the dominance of US infrastructure in the form of Mastercard and Visa in EU transactions and the creeping threat of “digital dollarisation.” There are some that are concerned Europe risks losing control over its monetary sovereignty and economic direction if we see stablecoins dominated by use companies. Philip Lane, the ECB’s chief economist, recently said: “It is imperative for the ECB to introduce a digital euro.” The push for a digital € isn’t just about catching up with technological trends - it’s about defending geopolitical autonomy, reclaiming payment sovereignty and initiating a new tech-driven financial era. Furthermore, surely now is the time for the UK government to instruct regulators to provide clear, concise guidance and so allow the UK to innovate in the field of programmable payments.
Agentic AI and large language models (LLMs) aren’t just tools - they’re reshaping how value moves across the world. As legacy systems groan under the weight of slow, centralised processes, a new era is emerging where intelligent wallets can think, decide and act - automating payments, managing supply chains and optimising liquidity without human friction. The bigger question isn’t whether we’ll adopt these systems - but who will build them first, and whose rules will govern them? Will the UK seize this moment to lead with a programmable, AI-driven financial infrastructure? Can stablecoins and intelligent wallets offer more than speculation - perhaps becoming a new backbone for national economies and trade?
This article first appeared in Digital Bytes (15th of April, 2025), a weekly newsletter by Jonny Fry of Team Blockchain.