Here’s a headscratcher for you.
Does a bigger brain make you smarter?
The answer is…probably not. Bit of a letdown I know.
As a species we have long held the belief that our massive noggins make us brighter. After all, what sets us apart from animals is our superior cognitive abilities. So, naturally we must have bigger brains, right?
Whales and elephants have much bigger brains than we do. And while I can say that I’ve seen an elephant paint, it certainly didn’t convince me that it was smarter than a human being.
Ah, but surely this is simply just a matter of context. The immense size of these creatures is obviously the cause of confusion. Their brains may be bigger, but so too are their bodies.
Therefore, surely humans must have the biggest brain-to-body ratio…right?
Wrong again I’m afraid.
Our relative brain size is in fact comparable to mice. And while some people may enjoy being lab rats, I’d wager that most of us are probably a tad smarter.
Size simply doesn’t matter. When it comes to the old IQ, what really counts is connections. The more neural connections or synapses you’ve got, the more thinking capacity you’ll have.
And the best way to create more synapses is to think. Turns out, thinking has been the secret to getting smarter all along…
But, imagine the possibilities if we could create synapses outside the body. Imagine if we could create an artificial brain.
Before we go any further, let me make one thing clear. When I refer to an artificial brain, it’s not a like-for-like comparison of artificial intelligence (AI).
A brain is the device or organ that does the thinking, intelligence is a measure of how complex that thinking is. That’s how it works in people, and that’s how it works for machines.
The challenge for researchers is to create an artificial brain which can produce superior AI. And the obvious place to start is by replicating real brains. The kinds found in you me and almost every other organism.
These artificial brains are more commonly referred to as neural networks. Complex systems of mathematical models inspired by our own brain. Though we haven’t managed to make them as smart as us just yet.
See, when everything is said and done, the best computers in the world are actually us. Sure, we may not be able to compute massive sums at the drop of hat, but our decision making abilities are far more diverse than that. As Time so eloquently puts it:
‘We perform tasks, make decisions, and solve problems based not just on our intelligence but on our massively parallel processing wetware — in abstract, what we like to call our instincts, our common sense, and perhaps most importantly, our life experiences. Computers can be programmed with vast libraries of information, but they can’t experience life the way we do.
‘Humans possess traits we sometimes refer to (again, in the abstract) as creativity, imagination and inspiration. A person can write a poem, compose and play music, sing a song, create a painting or dream up a new invention. Computers can be programmed to replicate some of those tasks, but they don’t possess the innate ability to create the way humans do.’
This is what makes us special and smarter than machines. But, this divide is rapidly closing. Machines are learning far faster than we are, and soon they will almost certainly overtake us.
The secret ingredient to making a perfect neural network could be biological rather than technical. The key might not be more silicon or diodes, it could be DNA.
A research team at the California Institute of Technology recently made a breakthrough. They created a neural network made out of DNA. Instead of plastic and metal, the circuits are biochemically derived.
But they still didn’t know if this DNA computer would really work. They had to test whether or not it could show signs of superior intelligence. So, they set it a challenge, and this is when the real breakthrough happened…
The research team set this DNA computer the task of identifying human handwriting. That may sound rudimentary, but for a machine it’s actually incredibly difficult.
The way people write by hand is incredibly varied. And if you’ve ever had a hard time deciphering someone’s scribbles, well that’s your brain performing complex computational processing.
That’s why identifying handwriting is such a common test for neural networks. They have to be taught how to recognise the complexities that go into handwriting. They need examples to recall on and look for patterns.
Trouble is, this DNA computer can only work with molecules. So, instead of human handwriting researchers used mother nature’s handwriting — or ‘molecular handwriting’.
Instead of using the lines and circles we draw for letters and numbers, molecules ‘handwrite’ with DNA. They use random strands of DNA made up from over 100 molecules. Which, as the researchers note, may not convey a number or a letter but something more abstract:
‘…for example, a mixture of unique odor molecules comprises a smell,’
I realise this is all getting a little technical. What matters though is that the DNA computer passed with flying colours. In one test it was right 100% of the time.
It brings us one step closer to true neural networks and superior AI. These biological computers could one day possess creativity and imagination that surpasses even our greatest human artists.
Just don’t hold your breath waiting for it to happen anytime soon.
The first ‘artificial’ Shakespeare is still a long way away from being born just yet.
We’ll just have to make do with our regular old human brains until then.
Editor, Tech Insider
PS: If you’re interested in new technology, you might want to check out our Free report ‘Could ‘The Next Bitcoin’ Make You Rich?’. In just eight years the crypto market has exploded from nothing…to a $500 billion dollar industry. That’s an insane amount of growth. If you’re interested in learning more, download this free report here.