Colombian Philosopher: ‘Study the Classics’ to Grasp Tech Disruption

Simón Villegas is a Colombian philosopher who never thought to get involved with tech. Yet, somehow, he is now scalp-deep on Medellin-based Xperience Design, a Constellation Technology-owned company that pioneered software solution design in Colombia.

“Thebes (Egypt) was Plato’s time equivalent to Silicon Valley,” he said in a recent interview with Nearshore Americas.

His social media and Noología newsletter insist that the nearshoring services industry and businesses in general should read classics more often to understand the ongoing technological revolution.

An underlying idea on Villegas’ writing and keynote presentations is that those interested in AI adoption should return to the history of technology development to dodge false prophets and leave the cave of self-centered IT mindsets.

The following are synthesized versions of three interesting ideas he’s shared regarding the generative AI trend.

Using Tech for Understanding

Villegas shared the story of a customer: “The company designed the UI for an app following global on-screen button standards. But they didn’t take into account their users were farmers and market square shopkeepers: a population that in Latin America tends to have big or swollen hands because of their agricultural labor. Their physiognomy happens to make it hard for them to use the app.”

Paraphrasing his follow-up argument: AI or software development experts can be great at programming, but the bombarding of tech news might make us forget the essentials.

One example is the key word: artificial. “Many people don’t realize the technology itself explicitly states it’s a probabilistic simulation of rationale, not a reasoning mechanism.”

How can AI Become a ‘Prosthetic?’

“In Plato’s Phaedrus, he illustrates an argument by telling the tale of a god that offers Thebes’ king every probabilistic invention as gifts. The deity presents everything from dice and luck games up to eventually, writing. Thebes’ king refuses to receive it. He’s unsatisfied by the pitch, so to speak. The god asks why, being that it was the medicine to oblivion. But Thebes’ king insisted it was the opposite.”

Villegas thinks that reference can speak about today:

“This interview is being AI-assisted through a notetaker. You are also taking notes manually. But many people can easily and needlessly turn to notetakers as attention-crutches. Thousands of years ago, oral tradition was so strong societies kept millenary knowledge just through widespread common sense built through speech. There are scientific studies and debates on how memory and synthesis capabilities change in societies that didn’t rely on writing as much as the West has.

“With AI, we’re in a scenario in which stuff that should be facilitating our lives can end up making it harder. It’s progressively harder to remember, synthesize and analyze information as we increasingly rely on automation for those tasks.”

Future of AI Adoption

“There have been various winters in the history of AI development; lack of funding, reconceptualization of trends and the creation of new problems all invite ‘phronesis,’ the creation of prudential wisdom.”

Villegas thinks technology evolves because of continuous depuration through “phronetic” conclusions.

“One of the beauties of capitalism is how it gets rid of bad ideas and bad investment. Companies that understand AI isn’t enough on its own, and needs to be attached to something else, will prevail in competition. For me, the winners will be those who interpret Large Language Model’s mechanisms correctly. Those who avoid confusing two separate events; data-driven AI’s ability to state and navigate an absence of evidence, and the analytic process and questions that surge from identifying an absence of data.”

He provided a concrete example: “Trump’s statement that Stargate’s AI development programs might diagnose new diseases and find new cures fundamentally ignores that generative AI does not create any new knowledge in itself.” It is humanity’s drive, data polishing, and experimentation, aided with technology, which can attempt to find the missing pieces of humanity’s puzzles.”

Villegas also reflected on AI’s problematic capacity to broaden inequality, not only in terms of employee competitiveness but also depending on outrunning the technology’s capacities to do certain tasks.

Also, in a sense, it can make access to qualified human analysis a progressively exclusive luxury.

“Once AI’s limits are clear for everyone, and the technology is worldwide established, there could be a return to the valuation of human labor.

“Take the medical sector. Insurance companies could decide that conversational AI may execute the protocols most BPO assistants or general doctors use when diagnosing patients. Then they could decide, since it’s cheaper, it could be made available for everyone as a solution for health plan coverage deficits,” he said.

In a very Doctor House-esque sense, Villegas believes, “more specialized medicine professionals will then gain value as they can suspect a patient needs attention and exams beyond the protocol to get precise diagnosis. That type of treatment that sees holes on probabilistic approaches will then become increasingly costly.”