ARTIFICIAL INTELLIGENCE REQUIRES OUR FULL ATTENTION

 

For my fellow traditionalists who love print newspapers, writing notes in cursive, and reading books made of paper, I have a warning and some thoughts. Artificial Intelligence (AI) is taking over the world. Now is not the time to complain about its disruptions or to ignore its presence. That reaction is exactly what the AI developers, venture capitalists, and certain government officials with an authoritarian bent hope you will do.

Each of us should study AI’s potential and its dangers. We should be knowledgeable enough to take a firm position on its further development and deployment.

For those who think AI is an unnecessary intrusion into our sedate, predictable lives, consider the 1811 Luddite movement of British textile workers. As part of this protest against industrialization, machinery was destroyed and riots ensued. Within five years, the Luddites were a distant memory and the industrial revolution took over the western world.  The United States was transformed from an agrarian society into an urban, industrialized one.

Skipping forward to the development of the internet, in 1992 a unique means of communication utilized by a few researchers was transformed into the “Information Superhighway.” Under a program championed by then Senator, Al Gore, seed money was provided to link computer networks in universities, governments, and industries around the world.

At first, the messy joint effort to develop the internet was encouraging. It was spearheaded by an early idealism that the new technology would empower individuals and unleash a wave of creativity for the benefit of mankind. The infant internet was praised for being decentralized and democratic. It was called a television station without producers and a newspaper without editors.

However, the internet took a dark turn that few saw coming. The erosion of privacy and the inability to protect personal data is now pervasive. The emergence of giant corporations (Google, Meta, Amazon), larger than nation states, has created a new class of oligarchs in America. The collapse of “old media” (newspapers, magazines, non-partisan news shows) has produced a society based on fake news and conspiracy theories rather than hard facts. The growth of tribal politics and the loss of a consensus reality is encouraged by social media.

The book to read on the downward spiral of the internet is Enshittification: Why Everything Suddenly Got Worse and What to Do About It by journalist Corey Doctorow. Doctorow believes that “The once-glorious internet has degenerated into “platforms” that rose to dominance because they delivered convenient and delightful services efficiently and reliably. But once we were locked in to those services, the tech bosses turned on us, relying on our dependency to keep us using the services.”

In spite of this sad state of affairs, Doctorow believes that there is a way to take the internet back from the tech giants. Each of us must learn about the algorithms that lock us into platforms and how to avoid their “siren call.” Governments must have the fortitude to use antitrust laws and regulations to break the monopolies, prevent fraud, and protect privacy.

While we deal with the internet’s failures, now is not the time to ignore AI or hand the keys over to the developers and financers without controls. AI poses more serious consequences to society and deserves our full attention.

The lightning speed with which AI has developed was not predicted by those who study its transformation. Early neural networks (the kind that underpin ChatGPT and Google’s Gemini) for decades did not work as well as more mainstream machine-learning techniques. It turned out to be a question of scale.

Building bigger neural networks was not easy, and no one knew whether it would work. Once these networks were fed an abundance of computer capacity to perform large numbers of computations, everything changed.

In 2012 a program running on only two Nvidia graphics chips in a developer’s bedroom left all of the conventional models in the dust. Fourfold annual growth led to sixteen-fold over two years, 64-fold over three years. This is how we progressed from a promising bedroom experiment to data centers the size of Manhattan. The models’ predictions continued to get better as their scale increased.

Bigger is expensive. Citigroup estimates that total AI investment globally will be at least $7.8 trillion between 2025 and 2030. No one knows when the rate of improvement will diminish or disappear. The ultimate result could be to reach “superintelligence,” a hypothetical form of AI that surpasses human intelligence across all fields, including scientific creativity, general wisdom, and social skills.

The dangers of superintelligence have been the subject of many science fiction novels. The developers lose control over the technology leading AI to pursue goals harmful to humanity. To prevent such an outcome international regulation must be developed to ensure that AI improvements are gradual and safe.

AI is not the internet. The internet shares information. The little understood AI technology independently interprets and acts on the information. We cannot afford to let AI develop without oversight.

The best solution may be the one advanced by the Economist in a recent Special Report, “Stop Panicking and Start Preparing.” Use AI to improve all aspects of your life. Stay on top of the latest developments and support controls and regulations. The Economist concludes that each of us needs to be “a cynical optimist,” as the world changes before our eyes.