3 min read
Before debating whether artificial intelligence is a threat or just a tool, we need to understand what artificial intelligence is. Is it something of the future or does it already exist?
What is Artificial Intelligence?
Intelligence describes the ability of learning and problem-solving. Natural intelligence is wired into all living things for survival. Similarly, when a man-made device is able to perceive its surroundings and take appropriate action, it is said to have intelligence. However, since that intelligence is a product of human intervention and programming, it’s called Artificial Intelligence.
Since its birth in the 1960s, AI has faced success and failure. Today, there are many technologies that are products of continuous AI work over the last few decades. Technologies like voice recognition, self-driving vehicles, virtual reality and of course, smartphones with all their features. An ever-evolving field, AI draws from various disciplines. While computer science is foremost, other fields include psychology, neuroscience, linguistics, cognitive science, probability, operations research and logic.
Why use Artificial Intelligence?
There are some strong advantages of why artificial intelligence, if employed correctly, is a boon.
There is no debating that computers are and always will be faster than humans. The speed at which a computer can compute multiple problems simultaneously is one of its biggest advantages.
In technical jobs like construction, the accuracy provided by a machine far supersedes the human ability of computation. This is where AI becomes important with its promise of precision.
- Reduced Biases:
Human intelligence struggles with decision making because of inherent biases. AI being devoid of emotions and biases can make a decision based on logic and facts. While biases cannot be completely removed, AI is, after all, a product of human intelligence and can only make decisions based on the facts given to it, it is as close to removing biases as possible.
Though it’s true that computers, and therefore AI, could get slower and possibly outdated over time without regular updates, unlike humans, computers don’t need breaks.
Drawbacks of Artificial Intelligence
Interestingly, the driving principle of AI is also its biggest weakness. AI was born to mimic and then outdo the capabilities of the human brain. However, the performance and capability of any artificial intelligence agent is only as good as the team of humans who programmed it and continue to maintain and develop it. Today, all the AI in the world is only able to perform tasks that we can perceive or pre-empt. By their very existence, given that these agents don’t have intelligence of their own, they are incapable of creating and inventing beyond the capabilities that we bestow upon them. This is not to say that AI has not made our lives easier and more convenient.
Does this Convenience Come at a Cost?
AI is both a blessing as well as a curse. On the one hand, AI is being used in cyber security to detect malware and suspicious activity online, especially on the dark net. AI’s ability to track patterns and anomalies makes the task easier and more efficient. It’s also beneficial in improving security and safety algorithms with its ability to analyse data and churn out possible solutions. Unlike human resources, AI can work 24/7/365. Artificial intelligence-based cyber security solutions are designed to work around the clock and respond within milliseconds.
On the other hand, it can serve as a deadly weapon in the arsenal of cybercriminals who may use AI to improve their cyberattacks. Another big factor that works in favour of cybercriminals is the cost of developing and implementing AI systems. Most businesses prefer using human power, rather than costly and complicated AI, to protect their systems. Cybercriminals, however, don’t require overly sophisticated systems to run their attacks.
We’ve established that AI takes decisions on the input given to it. Adversarial Artificial Intelligence is the term given to AI that is pushed by cybercriminals to misinterpret inputs and behave favourably to the cybercriminal. Take self-driving cars. Self-driving cars have sensors that pick up cues like obstacles and road signs and accordingly move the vehicle. If a cybercriminal hacks the AI in the car and blocks the cue for a red light, the car is likely to jump the signal, thereby causing an accident and endangering life and property.
Without the right defences in place, the applications of AI for cybercriminals are truly limitless. However, cyber security developers recognise the risks and continuously work at upgrading and improving AI systems to not only fight malicious activity but also get ahead of it.
The advantages of successful artificial intelligence agents far outweigh the disadvantages. That is why this field of technology continues to thrive today. After all, the purpose of machines and technology was, is and always will be to make human life easier.