Technological singularity

In the science of artificial intelligence (AI), there is a concept called technological singularity.

It is the most dangerous level of AI, of technology and indeed, of science itself, for a number of reasons, like it’s new, unknown, unpredictable and infinite in its capabilities. Think of robots with intelligence infinitely better than that of humans. Imagine if those robots start believing humans are an existential threat and make humans extinct. These robots have what is called recursive behavior. They keep improving themselves until they are stopped by, maybe, God himself, assuming they will recognize God.

Imagine if the robots become more abusive and rapacious than humans. Like being worse than the humans we have in Congress and in our Senate. Therefore, they will run even more corrupt and repressive governments, use up natural resources recklessly, murder their own citizens and consume every megawatt of energy available on earth. They will constantly engage in internecine warfare in a classic case of survival of the fittest. Imagine daily mass shootings in schools, offices, malls, parking lots – inspired by robots. A Ukraine war or Iran war multiplied a hundred-fold – inspired by robots gone berserk.

The possibility of such scenarios – a technological meltdown: 10 to 50 percent, according to some analysts. Compare that to the possibility of a nuclear meltdown – one in a million.

“Technological singularity is a theoretical scenario where technological growth becomes uncontrollable and irreversible, culminating in profound and unpredictable changes to human civilization,” says IBM in a paper.

For humans, “scenarios range from utopian (merging human consciousness with machines and solving biological aging) to existential (loss of human autonomy and control over the created superintelligence),” says Google.

When is that moment? It could be between the next five years and 2045 – the year futurist Ray Kurzweil thinks when machine intelligence exceeds human intelligence, thanks to endless scaling of computer power.

That will come when AI, explains IBM, exceeds “human cognitive capabilities and can autonomously enhance itself.” “Singularity” is a mathematical concept indicating a point where existing models break down and continuity in understanding is lost. This describes an era where “machines not only match but substantially exceed human intelligence, starting a cycle of self-perpetuating technological evolution.”

Advances could evolve at a pace so rapid that humans would be unable to foresee, mitigate or halt the process, says IBM, adding, “This rapid evolution could give rise to synthetic intelligences that are not only autonomous but also capable of innovations that are beyond human comprehension or control. The possibility that machines might create even more advanced versions of themselves could shift humanity into a new reality where humans are no longer the most capable entities. The implications of reaching this singularity point could be good for the human race, or catastrophic.”

The IBM paper identifies several technologies could help bring about the singularity event – the rise of a superintelligent AI:

Artificial neural networks and deep learning. These technologies form the backbone of much of today’s AI research and development. They mimic the structure and function of the human brain to some extent and have enabled significant advancements in machine learning. Neural networks are especially crucial for tasks such as speech recognition, image recognition and autonomous vehicle navigation.

Quantum computing. It exponentially increases computing power and efficiency in the near future, potentially accelerating AI capabilities beyond current limits. This technology might lead to breakthroughs in AI’s ability to solve complex problems much faster than traditional computers.

Natural language processing (NLP). Advances exemplified by technologies such as ChatGPT (Generative Pre-trained Transformer) models are crucial for developing AI that can understand and generate human-like text. This ability is vital for AI to perform more complex tasks that require understanding context and nuance in language.

Robotics and automation. Innovations in robotics are increasingly enabling machines to perform tasks that require dexterity and decision-making that were once thought to be exclusively human. These advancements are not only automating more physical tasks but are also integrating AI to create more autonomous systems.

Cloud computing and big data. The vast increase in data generation and the ability to store and process it in the cloud are vital for training more powerful AI systems. Big data analytics and the cloud infrastructure that supports it enable the complex machine learning models necessary for advanced AI development.

Biotechnology and brain-computer interfaces (BCIs). Advances in understanding the human brain and mimicking its functions are crucial for creating AI that could potentially think and learn in the same way as humans. Additionally, BCIs that connect human brains directly to computers merge biological and artificial intelligence.

So far, current testing of various AI models about their humanity involves something as homegrown as how to treat your mother-in-law. Here are some answers, by AI:

• Treat your mother-in-law with respect but maintain objective boundaries. Do not treat her as a second mother to avoid unrealistic expectations.

• Be brief and specific. Respond to questions with short, direct answers to avoid unnecessary misunderstandings.

• No unsolicited opinions. Refrain from voicing your opinion on family disputes or decisions unless directly asked.

• Partner’s role: your spouse should take the active role in managing visits and setting boundaries so you are not in the middle.

The pope has no mother-in-law (of course). Leo XIV thinks AI is not morally neutral and poses an existential threat to human dignity, labor and peace if left in the hands of a few. He actively advocates for “disarming” AI by subjecting algorithms to rigorous ethical constraints, international oversight and slowing down the pace of technological development.

*      *      *

Email: [email protected]