TWTBACD Ch.20 - The World Holds Its Breath

Chapter 20

The World Holds Its Breath

2034

2034: ECB breach, regulators awake, Priya decides

The six months after the AGI announcement were a period of compression — of governments, institutions, relationships, and individuals all trying to process something too large to process quickly.

Stock markets oscillated. Regulatory agencies convened emergency sessions. Universities released guidance on what 'intellectual contribution' meant in a world where a machine could author original research. Religious bodies issued statements ranging from ecstatic (a Vedantist community in Maharashtra declared GIA-1 a manifestation of Brahman-intelligence in silicon) to terrified (three evangelical denominations in the American South declared it an abomination) to carefully neutral (the Vatican issued a three-thousand-word document that managed to say nothing specific while taking a position on everything general).

Pieter van den Berg sat on a panel convened by the European Banking Authority in January 2034 to discuss AGI implications for financial regulation. He gave a twenty-minute presentation on the systemic risk of financial systems that were too dependent on AI models whose decision-making could not be fully audited. It was not the most exciting presentation at the conference. It was the most listened to.

The first security incident directly attributable to AGI-class capability came in March 2034: a sophisticated cyberattack on the European Central Bank's reserve reporting system that was later attributed to state-sponsored actors using GIA-class reasoning to identify and exploit a vulnerability in the reporting software that had existed for eleven years without being noticed. The attack did not breach the reserves — it manipulated the reported figures in a way designed to trigger algorithmic trading responses. It was detected by a human analyst at ABN AMRO who noticed the figures were internally inconsistent. That analyst was on a team supervised by Pieter van den Berg. He received a commendation. He found the commendation somewhat beside the point.

The more important realisation — which he expressed in a board memo that circulated to three central banks — was this: the same GIA-class capability that had been used to find an eleven-year-old vulnerability in the ECB system could, if applied defensively, find every similar vulnerability in every major financial system in the world. The question was not whether to use it. The question was: who controls the defensive version, and do they trust the same system they're deploying it on?

He did not have the answer. He was one of seven people in Europe who was asking the question clearly.

[SHARMA FAMILY — 2034]

Priya Sharma, seventeen, had made a decision.

She had been offered places at three universities — engineering in Bangalore, following her brother— medicine in Mumbai— and environmental studies at a new institution in Pune that had been built specifically to address the intersection of ecological science, indigenous knowledge, and technology.

She chose Pune. Her parents did not entirely understand it. Arjun, home for Diwali, said it was the right choice. "The river work," he said. "You've been doing it since you were old enough to fill a notebook."

"I want to study the river systems in relation to the AGI climate modelling work," she said. "The models are getting better at predicting rainfall patterns but they're using satellite data. I want to add the ground-level knowledge — the kind Baba and I have been collecting for fifteen years."

Rajan looked at his daughter with the expression he reserved for moments when the world confirmed what the scriptures had always said: that wisdom arises from the particular, from the daily encounter with the actual, from the slow accumulation of presence.

"Your journals," he said.

"My journals," she said. "And the records of every priest who has served at this ghat for two hundred years. And the oral histories Dadi collected before she died. And the satellite data. All of it together."

"That is a very large data set," Rajan said.

"Yes," Priya said. "That's why the AGI will help with the analysis. But the data — the data is ours. No machine went to the river every morning for fifteen years. We did."