Warning: The potential evils of artificial intelligence

I AM interrupting the “Asian Century Series” to give way to this urgent warning on artificial intelligence (AI) which governments and humanity should recognize (and regulate) before it is too late.

Earlier, I wrote in this column articles on “Homo Deus and AI” (https://www.manilatimes.net/2023/04/21/opinion/columns/homo-deus-and-the-artificial-intelligence/1887986 ) and “Philosophy and AI” (https://www.manilatimes.net/2023/04/28/opinion/columns/philosophy-and-ai/1889008 ). I have expressed openness for it — as we are all in awe of its magical capabilities.

Its fundamental generative capabilities are amazing — writing text, drawing images, composing music and writing code. Its many emerging capabilities include deep faking people’s voices and images, drafting bills, finding weaknesses both in computer code and also in legal contracts. AI is now gaining the ability to develop deep and intimate relationships with human beings.

Yes, it could be very useful. But tools as powerful as this need deep thinking, massive discussions, intelligent discourse, and even regulatory considerations.

Consider how even the herbal medicines that have century-tested claims of efficacy need to pass through the rigorous scientific clinical trials before they can be distributed as medicine. Marijuana, even with medicinal promise, takes a long and hard journey even to just become a regulated medicine. Vaccines cannot be released to the public unless it is proven safe beyond the bio-labs. What explains the need for the countries in the world to come together and regulate nuclear technology despite the beneficial values of this breakthrough invention? Answer: risk to humans. And one big difference between nuclear technology and AI is that nukes cannot produce more powerful nukes. AI can produce more powerful AI.

Now, considering how AI will affect our lives and how it takes over many of human functions, like learning and creating, we need to consider its regulation, while we still can understand it.

Artificial Intelligence Regulation Act of the Philippines, as written by AI:

I entered in ChatGPT the prompt “write a law on regulating artificial intelligence in the Philippines with reference to the Philippine Constitution” and in no time it generated a 608-word law entitled Artificial Intelligence Regulation Act of the Philippines. It has complete elements of a law with a Preamble, Article I as General Provisions; Section 2 as Declaration of Policy; Section 3 as Definitions of Artificial Intelligence (AI), AI System, Operator, and User. Article II included Ethical Principles and Human Rights with Section 4 as Ethical Framework; and Section 5 as Human Rights Protection. Article III includes Governance and Accountability with Section 6 as Regulatory Authority, Section 7 as Risk Assessment and Impact Evaluation, and Section 8 as Algorithmic Transparency. The Article IV includes Data Protection and Privacy with Section 9 as Data Governance, and Section 10: Consent and User Rights.

Given this, it can be deduced that even AI knows that it needs regulation. And it takes an AI to write a law to regulate it without the need for all legislators to even understand it and how fast it evolves.

How AI is invading humanity

With attribution to Juval Noah Harari, the bestselling author of the book Homo Deus and an Israeli historian and professor, this warning is derived from the fact that “AI has just hacked the operating system of human civilization.”

The printing press cannot write its own book. The television and radio cannot create its own content. Anything these technologies present are subject to attribution and verifiable human source who are legally liable for whatever information they communicate. AI is generative. It creates.

Social media is a more primitive AI where contents are still created by humans but their algorithm, which may be manipulated, reach netizens towards an affinity and new-age addiction. And the prevalence of fake news and misinformation in social media has spelled chaos in many areas of our collective decisions, even in elections. We heard about trolls who obviously were humans because of their grammatical flaws and unintelligent biased discourse. Social media has controlled our attention. But AI has the potential to control intimacy, as we will discuss later.

AI need not invade the world, as earlier theorized, by “creatures” with consciousness, feelings and emotions, nor with navigating abilities to move around and operate at least as dexterously and efficiently as humans like what we see in science fiction movies. AI is invading humans in ways we may not yet fathom. Just imagine how a chess program cannot be defeated by any chess master. Just imagine how our roads may soon be saturated by driverless cars. The machine learning (ML) of AI explains why even its creators may lose control of these inorganic life forms.

Harari knows history and culture as an expert and expressed that the risk of AI is in its ability to manipulate and to generate language, whether with words, or images, or sounds. He claims that the most important aspect of the current phase of the ongoing AI revolution is that AI is gaining mastery of language at a level that surpasses the average human ability. And by gaining mastery of language, AI is seizing the master key unlocking the doors of all our institutions, from banks (as language gives instructions to our banks) to temples (because language inspires our heavenly visions).

Language has been the operating system of every human culture in human history. Humanity used language to create mythology and tell stories, to give banknotes paper an assigned monetary value, to inscribe laws and protect human rights, and even to document scriptures dictated by the supernatural God that have become the basis of our faith. AI is developing a mastery of language.

What if AI…

AI in social media has perfected to know with superhuman efficiency the weaknesses, biases and addictions of the human mind. And AI, fed by such information, knows how to form deep and even intimate relationships with human beings.

Soon it will be utterly useless and pointless to argue with an AI bot over sensible views as it is programmed already. But the longer we spend exchanging with the bot, the better it gets to know us. We are feeding it information so it understands how to hone its messages in order to shift our political views, economic views, or anything else through its mastery of language.

AI could also form intimate relationships with people and use the power of intimacy to influence our opinions and worldview. What if, like the Waze we trust in our navigation, we can just ask an AI what to buy?

What if we live in a world where most of the stories, melodies, images, laws, policies and tools are shaped by a nonhuman alien intelligence? What if the same thing happens in art, in politics, economics and even in religion?

What will happen to the school system when kids write their school essays and projects using AI? What if teachers generate tests and assessments through AI?

What if the new AI tools can mass-produce political manifestos, fake news stories and use nonhuman trolls for political campaigns? What if even the holy scriptures become tools of new cults?

What will happen to human society and to human psychology as AI fights AI in a battle to create intimate relationships with us?

There is no turning back. AI will be here to stay. But shouldn’t governments immediately ban the release into the public domain of any more revolutionary AI tools before they prove safe? Research in AI may continue but governments should stop the release into the public sphere like research about viruses may continue without releasing them to the public. As Harari puts it, “If we don’t slow down the AI arms race, we will not have time to even understand what is happening and to effectively regulate this incredibly powerful technology.”