The latest new trend that we’ve been hearing about everywhere is AI, or Artificial Intelligence. It has become the most trending word in the IT field today. Especially with the introduction of AI robots such as Sophia. It seems almost every IT company is on the quest to integrate AI into their products for the future. It includes household devices such as thermostats, security systems, even fridges.
It is no surprise that our devices are definitely getting smarter, we now have phones that cannot only connect you to the internet, but can also recognize your voice and even your face. That’s not all; it can now take voice requests and perform tasks based on that as well.
And we haven’t even reached the tip of the iceberg. The potential for Artificial Intelligence deep learning technology, today and in the future, is vast. It offers a lot of different ways for us to upgrade our technology and make lives easier for us in the future.
According to Forrester, the potential for AI is so significant that investment in AI increased by 300% in 2017 alone. And, PwC estimates that productivity gains driven by AI will add $15.7 trillion to the global economy by 2030.
While, AI can offer a lot of benefits – it does leave people with worry. If movies and shows like The Terminator and I, Robot have taught us something, it’s that the future of AI makes the future of humans look extremely bleak. It won’t be long that AI will become smarter than the humans and then resort to overthrow us and try to take over the world. There are many movies and shows that are dedicated to the apocalyptic future which will the result from smarter machines. This has resulted in a lot of people becoming worried about the growth of AI.
However, this is one of the many different misconceptions that exist when it comes to AI.
7 common misconceptions when it comes to AI include:
1) AI has a conscious
The idea that AIs can think for themselves and actually make decisions is one of the biggest misconceptions that people have regarding AIs, including in movies like The Matrix. According to Chief Architect at technology company AirPR, Patrick Liang, AI machines cannot think for themselves and the input into the machines are completely controlled by humans. The machines only perform objectives that the humans assign them. The AI algorithms are written, controlled and structured by the humans. So, essentially, the humans control the specifications that the machines work on.
While, we do have autonomous drones that are designed for security that detect and prevent threats, these require specific instructions that are given to them by the humans. The idea that these can develop self-preservation instincts and start considering humans a threat, is at this stage, completely unlikely. With the technology we have today, it is highly impossible for AI machines to grow a conscious and take over the world!
2) Future systems will not need humans
This misconception adds to the one before it. The idea that smart systems will become more autonomous and self-sufficient and will no longer need us is not at something that we should believe could happen. As stated in the previous misconception, AI machines still need substantial human support, including for input, structure and algorithms. It is highly dependent on human guidance. These machines can’t think for themselves, and hence they simply can’t perform unless humans provide them with the problem that they need to work on.
3) AI will take over human jobs
One of the biggest worries in regard to AI is that the machines might take over more human jobs. According to research conducted by PwC, 30% of UK jobs in several sectors, such as retail, administration, and logistics, are predicted to be at risk due to AI in the next 15 years. However, here’s the truth. The jobs are not exactly being usurped by AI machines, instead, they are being changed. AI is changing our jobs, where we won’t be required to perform the menial tasks, instead there will be an influx of new positions where human expertise is going to be needed to manage AI machines. So, we will need more humans with AI expertise in the future.
4) AI systems will become dangerous
Earlier in 2017, Facebook reported that it shut down two of its AI chatbots for creating their own language for communication, which we couldn’t decipher. This got a lot of people worried that AI machines could become autonomous. Additionally, with AI usage becoming prominent in security systems and drones, the fear that AI systems could be dangerous to humans wasn’t that far off. However, as stated in the first point, AI-based systems are dependent on humans and they reflect the value-systems of the designers of these systems rather than going rogue in an autonomous manner. At this moment, AI technology is nowhere close to Artificial General Intelligence (AGI) required to become autonomous entities with their own value systems.
5) AI systems are smarter than the humans who built them
This misconception stems from looking at knowledge in a linear scale, where animals are always on the lowest scale while humans are somewhere on the higher end, and AI machines are considered to be at the top of the scale. However, this isn’t how knowledge should be calculated. Instead, intelligence is measured in different dimensions. In terms of calculations, machines may outsmart humans, while in terms of creativity, humans are smarter than machines. As Albert Einstein so brilliantly puts it, “Everybody is a genius. But if you judge a fish by its ability to climb a tree, it will live its whole life believing that it is stupid.” So, AI machines aren’t always smarter than us.
6) AI is fast-paced
We might think that AI is just around the corner. However, AI adoption isn’t going to be that fast-paced. It will actually take another 10-15 years before AI is fully integrated into many of the technology that we use today. And although, we already see snippets of AI in some of the technologies we use, wide-spread usage of AI is still far way.
7) AI needs a robot body
The last and most important misconception is that AI is supposed to be life-like and it needs a robot body. However, that is not at all true. You can find machine learning algorithms and AI algorithms in a lot of devices around us. Even your iPhone includes machine learning code that helps Siri understand you better. Machine learning and AI algorithms work on a processor level to help make your devices better and more personalized to your needs, and it for sure does not require a robot body to perform its task. Adding a more human and realistic body is just for fun!
These are just a few of the many misconceptions that are commonly associated with Artificial Intelligence. However, with AI slowly becoming more important in our everyday use, it is very important for developers and programmers to get on the bandwagon for building more machine learning and AI friendly development. There are a number of different artificial intelligence tutorials and machine learning tutorials available online, to help you get started.
Even through all the misconceptions, Artificial Intelligence technology is definitely one that is here to stay.