Artificial Intelligence Will Get Dumber as an Age of Improbability and Surprise Begins

by Dan Verhaeghe

Number-crunching machines are great with numbers, naturally, and they are able to understand the exponential growth of technology. Futurist Ray Kurzweil, who was in Toronto last week, says that humans are terrible at seeing the future—mostly because we think linearly instead of exponentially.

Humans are also terrible at thinking logically, according to Kurzweil. Very few people can think logically and exponentially, he argues. Kurzweil believes intuition is on the way out—but I argue against that in this post. explains intuitive learners this way: “Intuitive learners tend to focus more on the world of possibility. Unlike sensing learners who are interested in the here and now, intuitive learners enjoy considering ideas, possibilities, and potential outcomes. These learners like abstract thinking, daydreaming, and imagining the future.”

Based on my technological knowledge, even without knowing a line of code, I can see how things would work in the future to an increasing degree. It’s called an imagination, but machines and mostly numerical-thinking people usually have less of one.

Programming is also mostly mathematical and scientific so people who build the Internet think very numerically as well. That might be why Ray Kurzweil could not understand why young people understand technology so much better than older people.

It’s because we grew up with technology, and in doing so we were able to learn its ways through more than just technology knowledge as it was in our hands. We could touch and feel it and as a result became intuitive as well with technology on a very powerful level.

You could say we became “wired” unlike our parents. Every digital device we use also sparks different creativity much like every type of pen or pencil and artist picks up.

Consider that the tech industry is largely dominated by men. Most men don’t use their intuitive abilities as much as women. Therefore they don’t feel the world around them as well, so like Kurzweil they rely on statistics, Moore’s Law, concrete evidence, and numbers.

That’s one of the reasons why feminism will more likely prevail in the coming years in technology: because they’ll be able to intuitively see through the numbers artificial intelligence will be able to deliver on data generalizations easier than men. Forward-thinking machines are based on probability, but as you may have noticed with IBM’s Watson last year, they struggle with the improbable.

My theory is that as the world becomes more technologically advanced, more intangibles that we can’t quite explain will occur, and thus machines won’t be able to adapt very well. An intangible or improbability might represent the increasing diversity of the way we think, feel, act, and are as an increasingly multicultural race.

Technology is also adversely affecting what was once believed to be common psychology. That will only lead to more improbability which will probably mess up machines like Watson that can only understand the probable.

You may remember Watson was never even close to being perfect. Just because a single machine could know the whole Internet may make it factually smarter than you, but you’ll still have an advantage when it comes to analyzing the facts and comparing it to human culture to see perhaps increasing deviations.

That in turn will create an increasing element of surprise due to the increasing inaccuracy of artificial intelligence. That’s much like we see in survey marketing where it is only 95% accurate, or when sometimes we’re surprised by the end result in an election, for example.

After all, you can’t really predict sports, and there aren't even that many rules compared to what governs the world. And it’s a constrained environment. Imagine trying to predict something much larger like a city, a province, a country, or the world on a daily basis.

You would think the best team on paper would win all 82 games of an NHL season but they never do. So how can machines become smarter than man?

I don’t believe that when we have more connected devices than the 13 billion neurons each human brain has will the singularity have been reached. That’s when an artificial intelligence entity will have the capacity to be a “super digital brain” more powerful than a human brain.

I actually think it will go in the opposite direction for machines. I think the stats computers generate will actually eventually become less relevant with even more data to work with. That's because they will become increasingly general and especially as the connected device paradigm expands to tens of billions.

We as humans will become increasingly diverse with more intangibles that machines can’t account for. And quantifying big data will only work for certain fields or subjects that are truly numerical and come into contact with little to no intangibles.

blog comments powered by Disqus

Dan Verhaeghe

Dan Verhaeghe

Dan Verhaeghe focuses on marketing, mobile, major technology players, entertainment, and new media. Dan has a dozen years of online experience that dates back to the turn of the millennium where he dominated a now non-existent online RPG game for a couple of years at the age of 15. He would eventually become a Toronto Blue Jays blogger who earned his way into Toronto's CP24 studios six years... more

Who's Hiring

Recent Comments

Powered by Disqus