AI and Our Human Values

Guest Opinion

By Isiah Smith, Jr. | July 18, 2026

Whenever I sit down to write, whether it’s an essay or a text to one of my grandchildren, AI tries to intervene. Auto-correction and suggested rewrites pop up so often that I sometimes forget what I intended to say. I may not be James Baldwin or Proust, but I’ve grown accustomed to my pace, style, and intelligence (or lack thereof).

The concept of intelligence is central to our idea of who we are. That is one reason we refer to ourselves as homo sapiens, or “wise ones.” Irrespective of how much or how consistently we use our wisdom, we like to think of ourselves as intelligent beings. That conceit is one reason I, and I suspect other intelligent people, spend so much time thinking about artificial intelligence (AI).

The development and proliferation of artificial intelligence is nothing short of an examination and autopsy of our values as humans. One of the most uncomfortable truths in human history is the near certainty that whatever technological advances humans make can be used for evil as readily as for good. Think: Nuclear energy.

In his book, Human Compatible, Stuart Russell argues that the core AI challenge isn’t building intelligence but ensuring it remains aligned with human values. Russell says the classic approach, fixing a goal and optimizing hard objectives, can be dangerous because the system doesn’t truly grasp what we want. His proposal is that AI should operate under specified goals, because systems that rigidly optimize a target can cause massive unintended harm.

Russell proposes building machines that remain uncertain about what people truly want to learn from human behavior and defer when unsure. The book reframes AI as an alignment problem and calls for urgent research to keep powerful systems reliably beneficial.

I guess he’s talking about alignment with human values, or alignment with the human values we profess to have, but with a twist. He doesn’t want machines to assume they know our values—he wants them to stay uncertain and keep checking back with us for alignment with humanity. Basically, the system should treat human preferences as something to be learned, not declared.

My interest in artificial intelligence was piqued after reading Dr. Fei-Fei Li’s book The World I See. Dr. Li, best known for establishing ImageNet, the dataset that enabled rapid advances in computer vision in the 2010s, highlighted a fundamental problem in artificial intelligence: the underrepresentation of women and minorities in the emerging field. She recounted how a facial recognition program misidentified Black faces as apes. That monstrosity may not bother you until you remember that the president of the United States posted an AI-generated video online depicting President Obama and his wife as dancing apes.

We should not be surprised, therefore, that AI’s generative power, in the wrong hands, can do great harm.

So clearly, any artificial intelligence we develop and employ will operate within the limitations of our prejudices and short-sightedness.

Any technology that can be used for good may also be used for evil. When one considers the United States and its current values, it seems impossible not to conclude that one of our most prominent values is predatory behavior. This behavior leads us to attack smaller countries that lack nuclear weapons, and our leader muses about taking Greenland, a Danish territory, by “whatever means necessary, one way or another.” This predatory behavior has already driven our leadership to invade Venezuela and to promise that “Cuba is next.”

Dr. Li, affectionately named “the godmother of AI,” warns that decisions regarding AI will profoundly impact our healthcare, our employment, and every other aspect of our future. So, it is imperative that we develop clarity and understanding about what is happening and develop the tools to help shape what comes next.

Another book, Supremacy: AI, ChatGPT, and the Race That Will Change the World, by Parmy Olson, warns us of the real threat of AI that its top creators seem to be ignoring: the predatory profit motive is allowing the spread of flawed and biased technology across all aspects of our lives, including education and media. Olson has extensive experience covering and writing about technology, and she illustrates how the exploitation of the greatest inventions in human history has affected humankind.

Everyone who is interested in AI and its potential impact on our lives should read the books mentioned here. Knowledge is power. So much fearmongering is going on, much of it based on misinformation and incomplete understanding, that it prevents us from fully appreciating both the upsides and downsides of AI and how it can benefit and harm us.

What should concern us the most is AI’s potentially economic and social impacts. This technological change is occurring within the current environment of extreme wealth and political inequality. And this pre-existing condition will continue: The haves will continue to have more, and the have-nots will continue to have less.

Isiah Smith, Jr. is a retired government attorney.