Are You Smarter Than An LLM?

Chris Warkentin
3 min readMay 17, 2024

--

Man talking to robot.

I wonder if we will look back on this AI revolution and be embarrassed for our arrogance in assuming that our much vaunted intelligence was special. Detractors of these large language models call it things like ‘spicy autocomplete’ and state with conviction that machines will never match us. They dogmatically claim that the creative ‘spark’ possessed by humans is something that could never be captured by an algorithm, no matter how sophisticated.

The crux of the argument usually centers around how these various generative processes work. Built with massive amounts of data ingested from the internet, these programs create very convincing output by predicting the next part of a picture or a sentence based on statistically similar examples. Critics cry, “They’re just fancy predictive pattern matchers. All they do is put together a chain of statistically likely outputs and regurgitate them with no thought or reasoning.”

Consider, however, that the output is really good. It’s not perfect and sometimes suffers from errors or ‘hallucinations’ but for many purposes it’s incredibly useful. Often convincing enough that a normal person will have a difficulty telling if it was produced by man or machine. For such a seemingly simple process, it works fantastically well.

We can look down on generative AI because we know, more or less, how it works. The details are clouded in a lot of math and computation but the basic gist of it is readily apparent to almost anyone. Again the skeptics chime in. “How could something like that be intelligent? It’s just data extraction on steroids. How could it match the amazing creativity and reasoning that humans are capable of?”

When it comes to our brains, we don’t really know “how the sausage is made.” There are centuries of research and we know a great deal about the functioning of various parts of the brain but the things we call consciousness or self or creativity are still very much black boxes. Humans often deploy mysticism to explain that which they fail to comprehend and talk about things like ‘souls’ and ‘sparks’.

But how do we know that we aren’t also just fancy pattern matchers? Almost any creative person will list their influences and inspirations. Sir Isaac Newton famously stated, ‘If I have seen further than others, it is by standing upon the shoulders of giants.’ Every one of us is a conglomerate of our experiences, ingesting data all our lives like an AI model being trained. We too make mistakes and are subject to ‘hallucinations’ in the form of false memories, mental illnesses, errors of logic and other fallacies. Are we really so much better than these machines or are we just marginally better and convincing ourselves otherwise?

I believe that we stand at the delivery table of a new species that will expose the lie that we have convinced ourselves of. That we are somehow special and unique. That our intelligence is unmatchable and that some special part of us exists outside the bounds of scientific explanation. We’ll finally have the answer to the question of whether we are truly alone in the universe. I hope that it goes well for us.

--

--

No responses yet