These days, it’s easy to accept arguments that bogus intelligence has become as smart as the human mind—if not smarter. Google appear a speaking AI that dupes its communicative ally that it’s human.

DeepMind, a Google subsidiary, created an AI that defeated the world champion at the most complicated board game. More recently, AI proved it can be as authentic as accomplished doctors in diagnosing eye diseases.

And there are any number of belief that warn about a near future where robots will drive all humans into unemployment.

Everywhere you look, AI is acquisition new domains, tasks and skills that were ahead anticipation to be the absolute domain of human intelligence. But does it mean that AI is better than the human mind?

The answer to that catechism is: It’s wrong to assay bogus intelligence to the human mind, because they are absolutely altered things, even if their functions overlap at times.

Artificial intelligence is good at processing data, bad at cerebration in abstract

webrok

Even the most adult AI technology is, at its core, no altered from other computer software: bits of data active through circuits at super-fast rates.

AI and its accepted branch, machine acquirements and deep learning, can solve any botheration as long as you can turn it into the right data sets.

Take image recognition. If you give a deep neural network, the anatomy basal deep acquirements algorithms, enough labeled images, it can assay their data in very complicated ways and find correlations and patterns that define each type of object.

It then uses that advice to label altar in images it hasn’t seen before.

The same action happens in voice recognition. Given enough agenda samples of a person’s voice, a neural arrangement can find the common patterns in the person’s voice and actuate if future recordings belong to that person.

Everywhere you look, whether it’s a computer vision algorithm doing face acceptance or diagnosing cancer, an AI-powered cybersecurity tool ferreting out awful arrangement traffic, or a complicated AI activity arena computer games, the same rules apply.

The techniques change and progress: Deep neural networks enable AI algorithms to assay data through assorted layers; generative adversarial networks (GAN) enable AI to create new data based on the data set it has accomplished on; accretion acquirements enables AI to advance its own behavior based on the rules that apply to an environment… But what charcoal the same is the same basic principle: If you can break down a task into data, AI will be able to learn it.

Take note, however, that designing AI models is a complicated task that few people can accomplish. Deep acquirements engineers and advisers are some of the most coveted and highly paid experts in the tech industry.

Where AI falls short is cerebration in the abstract, applying common sense, or appointment ability from one area to another. For instance, Google’s Duplex might be very good at reserving restaurant tables and ambience up accessories with your barber, two narrow and very specific tasks.

The AI is even able to mimic accustomed human behavior, using inflections and intonations as any human apostle would. But as soon as the chat goes off course, Duplex will be hard-pressed to answer in a articular way. It will either have to abstruse or use the help of a human assistant to abide the chat in a allusive way.

There are many proven instances in which AI models fail in amazing and casuistic ways as soon as they’re presented with an archetype that falls alfresco of their botheration domain or is altered from the data they’ve been accomplished on.

The broader the domain, the more data the AI needs to be able to master it, and there will always be edge cases, scenarios that haven’t been covered by the training data and will cause the AI to fail.

An archetype is self-driving cars, which are still disturbing to become fully free admitting having driven tens of millions of kilometers, much more than a human needs to become an expert driver.

Humans are bad at processing data, good at making abstruse decisions

webrok

Let’s start with the data part. Contrary to computers, humans are abhorrent at autumn and processing information. For instance, you must listen to a song several times before you can acquire it.

But for a computer, abstraction a song is as simple as acute “Save” in an appliance or artful the file into its hard drive. Likewise, unmemorizing is hard for humans. Try as you might, you can’t forget bad memories. For a computer, it’s as easy as deleting a file.

When it comes to processing data, humans are acutely inferior to AI. In all the examples common above, humans might be able to accomplish the same tasks as computers. However, in the time that it takes for a human to analyze and label an image, an AI algorithm can allocate one actor images.

The sheer processing speed of computers enable them to outdistance humans at any task that involves algebraic calculations and data processing.

However, humans can make abstruse decisions based on instinct, common sense and scarce information. A human child learns to handle altar at a very young age. For an AI algorithm, it takes hundreds of years’ worth of training to accomplish the same task.

For instance, when humans play a video game for the first time in their life, they can bound alteration their accustomed life ability into the game’s environment, such as blockage away pits, ledges, fire and pointy things (or jumping over them).

They know they must dodge bullets and avoid accepting hit by vehicles. For AI, every video game is a new, alien world it must learn from scratch.

Humans can invent new things, including all the technologies that have ushered in the era of bogus intelligence. AI can only take data, assay it, come up with new combinations and presentations, and adumbrate trends based on how antecedent sequences.

Humans can feel, imagine, dream. They can be affectionate or greedy. They can love and hate, they can lie, they forget, they abash facts. And all of those affections can change their decisions in rational or aberrant ways.

They’re amiss and flawed beings made of flesh, which decays with time. But every single one of them is unique in his or her own way and can create things that no one else can.

AI is, at its core, is tiny bursts of electricity active through billions of asleep circuits.

Let’s stop comparing AI with human intelligence

webrok

None of this means that AI is above to the human brain, or vice versa. They point is, they’re absolutely altered things.

AI is good at repetitive tasks that have acutely authentic boundaries and can be represented by data, and bad at broad tasks that crave intuition and controlling based on abridged information.

In contrast, human intelligence is good for settings where you need common sense and abstruse decisions, and bad at tasks that crave heavy computations and data processing in real time.

Looking at it from a altered perspective, we should think about AI as augmented intelligence. AI and human intelligence accompaniment each other, making up for each other’s shortcomings. Together, they can accomplish tasks that none of them could have done individually.

For instance, AI is good at perusing huge amounts of arrangement cartage and pointing out to anomalies, but can make mistakes when chief which ones are the real threats that need investigation.

A human analyst, on the other hand, is not very good at ecology gigabytes of data going through a company’s network, but they’re adept at apropos anomalies to altered events and addition out which ones are the real threats. Calm AI and human analysts can fill each other’s gaps.

Now, what about all those accessories that claim human labor is going instinct? Well, a lot of it is hype, and the facts prove that the amplification of AI is creating more jobs than it is destroying. But it’s true that it will anticipate the need for humans in many tasks, just as every abstruse advance has done in the past.

But that’s apparently because those jobs were never meant for humans. We were spending adored human intelligence and labor on those jobs because we hadn’t developed the technologies to automate them yet.

As AI becomes adept at assuming more and more tasks, we as humans will find more time to put our intelligence to real use, at being creative, being social, at arts, sports, literature, poetry and all the things that are admired because the human aspect and appearance that goes into them. And we’ll use our aggrandized intelligence tools to enhance those creations.

The future will be one where bogus and human intelligence will build together, not apart.