AI Is Changing Everything. But Is It Making Us More Intelligent?

“The important thing is not to stop questioning.” — Albert Einstein

Artificial intelligence is having its iPhone moment.

OpenAI is worth hundreds of billions of dollars. Anthropic has become one of the fastest-growing software companies in history. Every major technology company is racing to build AI products, while governments increasingly view AI as critical national infrastructure. The excitement is understandable. Today’s AI can write code, summarize documents, answer questions, translate languages, generate images, and automate countless repetitive tasks. Those are remarkable engineering achievements. But amid all the excitement, one question is rarely asked.

Is AI making us more intelligent, or simply more efficient? Those are not the same thing.

Intelligence Is More Than Having Answers

Most AI systems are trained on one thing:

The past. Books. Research papers. Websites. Code. Images. Videos.

In other words, AI learns from what humans have already created. That makes AI incredibly good at organizing knowledge. But organizing knowledge is different from creating new understanding. As computer scientist François Chollet puts it,

“Intelligence is skill-acquisition efficiency.”

True intelligence is not remembering more facts. It is learning something new from very little information.

Data Tells You What Happened. Not Why.

Imagine watching the scoreboard after a football match. You know the final score. You still don’t know why one team won. The same is true for data. Stock prices tell you what happened. Medical records tell you who became sick. Economic indicators tell you what changed. None of them automatically explain the causes.

As AI pioneer Judea Pearl argues,

“Data are profoundly dumb.”

Without understanding cause and effect, prediction alone has limits.

More Data Doesn’t Guarantee More Truth

Modern AI has improved largely because researchers discovered that bigger models trained on more data tend to perform better. This is one of the biggest discoveries in machine learning. But better prediction does not automatically mean better understanding. History is full of ideas that were accepted for centuries before being overturned. The Earth was once believed to be the center of the universe. Newton’s laws explained physics—until Einstein showed they were incomplete. Science advances because humans question accepted knowledge. Not because they memorize more of it.

Knowledge Is Different From Intelligence

There’s an old quote often attributed to Einstein:

“Information is not knowledge.”

Whether or not he said those exact words, the idea is correct. Knowledge is accumulated information. Intelligence is deciding when that information is wrong. Every major scientific breakthrough required someone to challenge what everyone else believed. Darwin. Einstein. Curie. Turing. None simply summarized existing knowledge. They changed it.

Nature Optimizes Adaptation

Evolution never optimized humans to predict the next word. It optimized us to survive. To adapt. To solve problems we had never encountered before. That distinction matters. Current AI models are exceptional at recognizing patterns. Humans are exceptional at creating new ones. As psychologist Jean Piaget observed,

“Intelligence is what you use when you don’t know what to do.”

That remains one of the hardest capabilities to reproduce in machines.

The Missing Ingredient: Causality

One of AI’s biggest weaknesses today is explaining “why” something happens. A language model can often tell you what is likely to happen next. It is much less reliable at explaining the underlying mechanisms. This is why causal reasoning has become one of the fastest-growing research areas in AI. Prediction is useful. Understanding is transformative.

AI Is an Incredible Tool

None of this means AI is overhyped. Quite the opposite. AI already saves millions of hours every day. It helps developers write software. Doctors summarize medical literature. Lawyers review contracts. Students learn faster. Businesses automate repetitive work. These are genuine advances.

The mistake is assuming productivity equals intelligence. A calculator performs arithmetic better than any human. Nobody claims it understands mathematics.

The Bigger Risk

The greatest risk may not be AI becoming smarter than humans. The greater risk is humans becoming less curious because AI is available. When answers become effortless, questions become rarer. And throughout history, progress has always started with better questions—not faster answers. As Nobel Prize-winning physicist Richard Feynman said,

“I would rather have questions that can’t be answered than answers that can’t be questioned.”

That mindset built modern science. It should also guide modern AI.

Intelligence Still Belongs to Humans

AI is extraordinarily good at compressing humanity’s existing knowledge. Human intelligence is extraordinary because it expands humanity’s knowledge. Those are different objectives. One predicts. The other discovers. One explains what is already known. The other changes what is possible. Perhaps future AI will bridge that gap. Perhaps it won’t. But until then, we should be careful not to confuse statistical prediction with scientific discovery.

Where TickerTruth Fits

At TickerTruth, we don’t believe the future of investing belongs to people who consume more information.

We believe it belongs to people who ask better questions.

Financial markets already have an abundance of data.

They have an abundance of news.

They have an abundance of opinions.

What they lack is a systematic way to transform information into genuine research.

Our goal is not to build another chatbot that summarizes earnings calls or explains yesterday’s market move.

Our goal is to build a research platform that helps investors discover relationships they didn’t already know existed.

That means moving beyond dashboards and AI summaries toward:

• Hypothesis generation, not just search.
• Causal investigation, not just historical correlation.
• Evidence-first research, not opinion-first content.
• Repeatable factor discovery, not one-off stock tips.
• Transparent reasoning, where every conclusion can be traced back to its underlying evidence.

The best investors have never won because they possessed more data.

They won because they interpreted the same data differently.

AI should amplify that process—not replace it.

If the next generation of AI is about producing answers, TickerTruth’s mission is to help investors ask better questions.

Because in markets, as in science, enduring edge comes not from knowing what everyone else knows—but from discovering what everyone else has overlooked.

References:

[1] Judea Pearl & Dana Mackenzie (2018). The Book of Why: The New Science of Cause and Effect. https://www.penguinrandomhouse.com/books/547672/the-book-of-why-by-judea-pearl-and-dana-mackenzie/

[2] Jared Kaplan et al. (2020). Scaling Laws for Neural Language Models. OpenAI. https://openai.com/index/scaling-laws-for-neural-language-models/

[3] François Chollet (2019). On the Measure of Intelligence. https://arxiv.org/abs/1911.01547

[4] Karl Popper (1959). The Logic of Scientific Discovery. https://www.routledge.com/The-Logic-of-Scientific-Discovery/Popper/p/book/9780415278447

[5] Richard P. Feynman (1965). The Character of Physical Law. https://mitpress.mit.edu/9780262560030/the-character-of-physical-law/